139739

CRANFIELD UNIVERSITY
Yueh-Ling Hsu '`,
Airline Safety Management:

The developmentof a proactive safetymechanismmodel
for the evolution of safety management system
COLLEGE OF AERONAUTICS
AIR TRANSPORT GROUP
Ph. D. THESIS
I
CRANFIELD UNIVERSITY
COLLEGE OF AERONAUTICS
AIR TRANSPORT GROUP
Ph. D. THESIS
Academic Year 2003-2004
Yueh-Ling Hsu
Airline Safety Management:

The development of a proactive safety mechanism model
for the evolution of safety management system
Supervisor: Mr. A. Frank Taylor
August 2004
This thesis is submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy
Cranfield University 2004. All right reserved. No part of this publication may be
reproduced without the written permission of the copyright owner.
Abstract
The systemic origins of many accidents have led to heightened interest in the way in
which organisations identify and manage risks within the airline industry. The
activities which are thought to represent the term "organisational accident", "safety
culture" and "proactive approach" are documented and seek to explain the fact that
airlines differ in their willingness and ability to conduct safety management. However,
an important but yet relatively undefined task in the airline industry is to
conceptualise the safety mechanism in proactive safety, and its influential factors.
What is required is a model of a proactive safety mechanism which builds upon
existing knowledge of what is thought to contribute to safety by adding an increased

knowledge of the organisational factors. These factors not only serve to influence the
safety mechanism, but also serve to be the predictors of the performance of safety
management system.
This thesis aims to fill that gap. It firstly conducts an overview of the current airline
safety management system literature and identifies the strengths and weaknesses of
the current system. Given the need to explore the important but undefined field, a
proactive safety mechanism model is then developed and tested to identify the
organisational factors which exert an influence upon the safety mechanism.
Four hypotheses were set out to be tested in an attempt to justify the
multi-dimensional and complex nature of the safety mechanism model. The model is
then tested by applying it to a past accident (case study) and a survey of opinions with
questionnaire. The results of this research work show that the safety mechanism
model is a model of the evolution of safety management system in the context of
proactive safety management. Further study can apply the proposed model to the
re-organisation of an airline safety management system and evaluate the impact upon
the company's system. It leads to the suggestion that an airline's safety health and
performance needs the co-ordination of both retroactive and proactive safety

management, and concludes that the ultimate contribution of this research is to
provide airlines with reliable data, applicable references and a practicable
methodology to enable their safety management system to evolve at a fundamentally

"genetic" level.
Acknowledgments
This project would not have been possible without the help, support and
of
encouragement a great many people to whom I owe a debt of gratitude.
First and foremost,my thanks go to my supervisor,Mr. Frank Taylor, whose technical
advice and moral support are greatly appreciatedthroughout the duration of this study.
My thanks go as well to everyone else in the Air Transport Group. To Prof Fariba
Alamdari, Dr. Peter Morrell and Dr George Williams who made this project possible,
to Dr. Graham Braithwaite who took time to be of help and support, to Mr. Mike
Thompson, Mr. Rodney Fewings and other staff whose invaluable assistance, and to
the students who made life at Cranfield more pleasant, I am truly grateful.
In particular, special thanks to the following people for their kind support and
information: Prof Peter Brooker, Mr. Robin Ablett, Capt. Peter Hunt, Mr. Jim Done
and Mr. Adrain Sayce from the Safety Regulation Group, UK CAA; Director of
Taiwan CAA - Mr. Billy K. C. Chang and Mr. F. H. Han also from Taiwan CAA.
My sincere appreciation also goes to all those anonymouspeople who gave of their
time and expertiseto provide the data for this research.
Finally, my thanks go to all my family and friends who were supportive of this study
project throughout. In particular, thanks and love to my parents, whose endless

affection kept me focusedon what was really important throughout.
Table of Contents
ABSTRACT
ACKNOWLEDGEMENTS ii
TABLE OF CONTENTS III
LIST OF FIGURES Vil
LIST OF TABLES ix
ABBREVIATIONS xi
CHAPTER 1 INTRODUCTION
1.1 RESEARCH BACKGROUND 1

1.2 RESEARCH OBJECTIVES 4
1.3 THESIS STRUCTURE 5
CHAPTER 2 LITERATURE REVIEW & EMPIRICAL ANALYSIS

2.0 ROLE AND DEFINITION OF SAFETY 7
2.1 EXTERNAL IMPACT ON AIRLINE SAFETY 8
2.1.1 Technical innovation and its impact 9

2.1.2. Growing traffic and airline safety 11
2.1.3 Regulatory environment 13
2.2 AIRLINE SAFETY MANAGEMENT SYSTEM 18
2.2.1 Flight safety risk 18

2.2.2 Impact of accident on airline safetyperformance 21
2.2.3 Definition of safety management 25
2.2.4 Systemsafety and safety managementsystem 26
2.2.5 Linking quality with safety 36
2.3 RISK MANAGEMENT 38
2.3.1 The definition of risk and risk management 38
2.3.2 Risk managementprocess in safety managementsystem 40
2.4 HUMAN CONTRIBUTION 54
2.4.1. Humanfactors versusrisk. 54
2.4.2 Human error 59
2.5 A RETROACTIVE APPROACH TO SAFETY 65
2.6 PROACTIVE SAFETY 68
2.6.1 Whatis proactive safety? 68
2.6.2 Organisational safety 69
2.6.3 Culture and safety 73
2.6.4 Contemporary safetyprogrammes with proactive concepts
in the airline industry 83
2.7 THE PROACTIVE VS. RETROACTIVE APPROACHES TO SAFETY 93
CHAPTER 3 RESEARCH METHODOLOGIES
3.0 INTRODUCTION 97
3.1 INTERviEw 100

3.1.1 Selection of airlines 100
3.1.2. Selectionof interviewees 101
3.1.3lnterviewformat 101
3.2 MODEL DEVELOPMENT 102
3.2.1 Purpose of the model 102

3.3 How TO TEST THE MODEL 104
3.3.1A retroactive casestudy 104
3.3.2 Safetysurvey 105
3.4 SUMMARY 114
CHAPTER 4 INTERVIEW FINDINGS & MODEL DEVELOPMENT
4.1 INTERVIEW FINDINGS 115
4.1.1 Safety management system 115

4.1.2. Organisationalstructure 118
4.1.3 Risk managementand risk programmes 118
4.1.4 Safetyaudit 123
4.1.5 Culture versus safety 124
4.1.6 The role of regulator and the relationship with airlines 126
4.1.7 Improvementand other comments 128
4.2 THE DEVELOPMENT OF SAFETY MECHANISM MODEL 130
4.2.1 Overview and hypothesisof the model 130

4.2.2 Layers and their definitions 135
4.2.3 Organisational factors 146
CHAPTER 5 RESULTS OF MODEL TEST
5.0 INTRODUCTION 159

5.1 QUANTITATIVE METHOD RETROSPECTIVE CASE STUDY 159
-A
5.1.1 The event 159
5.1.2 The pathogen trails by applying the safety mechanism model 161
5.1.3 Conclusionsof organisationalfactors which influence the safety mechanism 166
5.2 QUANTITATIVE METHOD SAFETY QUESTIONNAIRE 171
-
5.2.1 Findings from pilot study 172
5.2.2 Resultsof the safetysurvey 173
CHAPTER 6 ANALYSES AND DISCUSSIONS
6.0 INTRODUCTION 177
iv
6.1 ORGANISATIONAL FACTORS IN INTERNAL ENVIRONMENT 178
6.1.1 Loading factors from Principal Component Analysis 179
6.1.2 Statistical analysesfor internal factors 186
6.2 ORGANISATIONAL FACTORS IN EXTERNAL ENVIRONMENT 188
6.2.1 Loading factors from Principal Component Analysis 189
6.2.2 Statistical analyses for external factors 191
6.3 RELATIONSHIP BETWEEN ORGANISATIONAL FACTORS AND AFETY PERFORMANCE 192
6.4 RELATIONSHIP BETWEEN INTERNALAND EXTERNAL FACTORS 194
6.5 INTERPRETATION OF INDIVIDUAL FACTORS 195
6.5.1 Discussion of Internal factors 196
6.5.2 Discussion of External factors 206
6.6 COMPARISON WITH THE HYPOTHESISED FACTOR STRUCTURE 211
6.7 DISCUSSION OF THE BIAS IN THE RESULT 214
6.8 IMPLICATIONS FOR THE SURVEY RESULT 216
6.9 EXAMINATION OF THE FINDINGS FOR THE PRIMARY HYPOTHESIS 223
6.9.1 Hypothesis 1 223
6.9.2Hypothesis2 224
6.10 LIMITATIONS AND VALIDATION OF THE SAFETY MECHANISM MODEL 232
6.11 THE APPLICATION OF THE SAFETY MECHANISM MODEL 236
CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS
7.1 CONCLUSIONS- THESIS AIMS REVISITED 241

7.1.1 Objective one 241
7.1.2 Objective two 243
7.1.3 Objective three 245
7.1.4 Objectivefour 246
7.1.5 Objectivefive 249
7.2 RECOMMENDATIONS 250
REFERENCES 259
BIBLIOGRAPHY 273
V
Appendices
Page
APPENDIX A Definitions of Accident/Incident 277
APPENDIX B Impact of Accident on Airline Operation &
FinancePerformance 279
APPENDIX C Quality Management versus Safety Management 283
APPENDIX D Risk Analysis Techniques 289
APPENDIX E Documentation Pyramid 291
APPENDIX F Other MEDA-Like Approaches 292
APPENDIX G Casesof OrganisationalChange-culture transformation 293
APPENDIX H The Four Proactive Safety Programmes 297
APPENDIX I Domino Theory 309
APPENDIXJ Interview Questions 311
APPENDIX K Airline Distribution List 313
APPENDIX L Covering Letter and the Questionnaire 317
APPENDIX M Participants Comments 325
APPENDIX N RotatedComponentMatrix for Internal &
External Loading Factors 327
APPENDIX 0 Alpha Value for Internal & External Factors 329
APPENDIX P Mean and StandardDeviation of Internal
and External Factor Scores 331
APPENDIX Q Table ofANOVA and Mean/StandardDeviation
for Internal & External Factors 335
APPENDIX R Multiple RegressionResult 341
APPENDIX S Correlation of Internal & External Factors 342
APPENDIX T DMAIC Processin Six Sigma 343
vi
List of Figures
Page
Figure 1-1 Researchstructureand methodsused 6
Figure 2-1 Safety relationships in the air transport system 9
Figure 2-2 Current regulatory environment 15
Figure 2-3 Annual no. of accidents vs. fatality rate 20
Figure 2-4 The impact of accidentof airline performance 22
Figure 2-5 The safety trinity 23
Figure 2-6 Man causes and machine causes 24
Figure 2-7 The comparisonof finance and safety managementsystem 28
Figure 2-8 The components of a SMS 29
Figure 2-9 The interaction of characteristicsand componentsof a SMS 31
Figure 2-10 The suggestedorganisationalstructure 33
Figure 2-11 The principles and relationship of quality and safety 37
Figure 2-12 Safety loop 40
Figure 2-13 The Heinrich pyramid 43
Figure 2-14 Barriers to organisationallearning 53
Figure 2-15 The SHELL model 57
Figure 2-16 The error iceberg 65
Figure 2-17 Reason'sorganisationalaccidentmodel 70
Figure 2-18 A model of organisational accident causation 71
Figure 2-19 Layers of culture 74

Figure 2-20 Organisation and subculture 80
Figure 2-21 Proactive defenceevaluation model 87

Figure 2-22 The PERS structure 89
Figure 3-1 Structureof usedresearchmethods 99
Figure 4-1 Airline safety managementsystem 117
Figure 4-2 The common framework of risk assessmentin the
interviewed airlines 120
Figure 4-3 The relationship between safety culture and SMS 126
Figure 4-4 The structureof proactive safety mechanismmodel 131
Figure 4-5 The impact of safety climate 140
VII
Figure 4-6 The three "P" framework in the airline industry 144
Figure 4-7 The concept of safety cost 155
Figure 4-8 The hypothesisedsafety mechanismmodel 157
Figure 5-1 Findings and conclusionsof the safetymechanismmodel
applied to Air Ontario accident 167
Figure 5-2 Distribution of final sampleby region 174
Figure 6-1 Organisationallearning barriers vs. perceived
safety performance 222
Figure 6-2 The safety mechanismmodel applied to the
Air Ontario accident 226
Figure 6-3 The safety mechanism model (survey result) 213
Figure 7-1 A GenericFramework for Retroactiveand Proactive
Approach to Safetywithin Airline SMS 254
viii
List of Tables
Page
Table 2-1 Airline industry businessrisk profile 19
Table 2-2 Risk control tools for top accident causes 50
Table 2-3 Estimatesof Human Error in Different Industries 60

Table 2-4 Basic organisation communication styles 81
Table 2-5 Definitions of safety culture 82
Table 2-6 Various error types and remedial strategies 92

Table 2-7 Comparisonof different characteristicsin the proactive and
reactive systems 96
Table 3-1 The profile of selectedairlines 100

Table 3-2 Samplesand regions of questionnaireadministration 112
Table 4-1 Definition of organisationalclimate and culture 138
Table 4-2 Definition of safety climate 139
Table 4-3 The definition of safety climate and safety culture 142
Table 4-4 The definitions of safety philosophy and decision making 145
Table 4-5 Various groups' concerns& impact factors 152
Table 4-6 Breakdown of the factors 153
Table 5-1 Distribution percentage of final sample by region 175
Table 5-2 The Factor structureof a proactive safety mechanism 176

Table 6-1 KMO and Bartlett's Test: Internal factors 178
Table 6-2 The underlying nature of factor 1 179

ix

Table 6-17 KMO and Bartlett's Test: External factors 188

Table 6-18 The underlying nature of factor El 189
Table 6-19 The underlying nature of factor E2 189

Table 6-22 Summariesof the survey results 197

Table 6-23 Comparisonof the factor structureof proactive
safety mechanism 213

Table 6-24 The comparisonsof safety mechanismmodel
and four programmes 233
Table 6-25 The DMAIC processof X12"Communication system" 239
X
Glossary of Abbreviations
AHI Aviation Health Institute

ALARP As Low As ReasonablyPracticable
ANOVA Analysis of Variance
APU Auxiliary Power Unit
ATC Air Traffic Control
ASRS Aviation SafetyReporting System
ATA Air TransportAssociation
ATI Air Transport Intelligence
BA British Airways
BASI Bureau of Air Safety Investigation
BASIS British Airways Safety Information System

CAA Civil Aviation Authority
CAP Civil Aviation Publication
CAIR Confidential Aviation Incident Reporting Network
CEO Chief Executive Officer
CMT Crisis ManagementTeam
CFIT Controlled Flight Into Terrain
CP Check Pilot
PCA Principle Component Analysis
CRM Crew ResourceManagement
CST Central StandardTime
DFDR Digital Flight Data Recorder

DME DistanceMeasuring Equipment
ECAC European Civil Aviation Conference
ERCs Error Reducing Conditions
ERP EmergencyResponseProcedures
EGPWS EnhancedGround Proximity Warning System
EM Error Management
EST EasternStandardTime
xi
EUCARE European Confidential Aviation Safety Reporting
System
FAA FederalAviation Administration
FARs Federal Aviation Regulations
FSF Flight Safety Foundation

FORAS Flight Operation Risk Assessment
FOQA Flight OperationQuality Assurance

GAIN Global Aviation Information Network
GDP Gross Domestic Product
GPWS Ground Position Warning System
GMT Greenwich Mean Time
HF Human Factors
HEAR Human Error and Accident Reduction
IATA InternationalAir TransportAssociation

ICAO International Civil Aviation Organisation
IFA International Federationof Airworthiness
INDICATE Identifying NeededDefencesIn the Civil Aviation
Transport Environment
ILU Institute of London Underwriters

JAA Joint Airworthiness Association
JARS Joint Aviation Requirements
KAL Korean Air
LAUA Lloyd's Aviation Underwriters Association

LIRMA London Insurance Reinsurance Market Association
LOSA Line Operation SafetyAudit

LSD Least Significant Difference
MEDA MaintenanceError Decision Aid
MESH Managing EngineeringSafety Health
MRM MaintenanceResourceManagement
MORs Mandatory OccurrenceReports
MRO Maintenance,Repair and Overhaul

NATS National Air Traffic Services
XII
NTSB National TransportationSafety Board
OEM Original Equipment Manufacturer
PERS Proactive Errors Reduction System

PFC Passenger Facility Charge
QAR Quick AccessRecorder

RPKs RevenuePassengerKilometres
SARPs Standardsand RecommendedPractices
SAS ScandinavianAirlines System
SCQ Safety Climate Questionnaire
SECURITAS The Confidential TransportationSafety Reporting
SystemProgram
SMS Safety ManagementSystem
SOPS StandardOperatingProcedures
SRG Safety Regulation Group
VOR VHF Omnidirectional Range
xm
This pagehas been left blank intentionally
xiv
CHAPTER 1
Introduction
"How, then, do we get this companyahead of change,

ahead of thesestrong global players, permanently out in front
in the '90s?How do we set thepace. "
Jack Welch,1989
1.1 Research Background
Air transport is experiencing increasing growth year by year, with passenger air
transport, in particular, becoming more affordable and feasible for both short journeys
and long haul flights (Shorrock and Kirwan, 2002). During the past decade, although
the growth in air travel varied from region to region, global demand for air travel
increased strongly according to the statistics of IATA (2001). Doganis (2001) stated that
long-term traffic growth from 2000 to 2010 will average close to 5 percent per annum.
Boeing, too, predicts that world air traffic, measured in Revenue Passenger Kilometres
1,
(RPKs) will grow by 4.7 % annually over the next 20 years2 (Boeing, 2001).
1 RevenuePassengerKilometres (RPKs) are obtainedby multiplying the number of fare paying

passengerson eachflight stageby the distanceof that stage.
2 IATA (2002) has statedin its new interim five-year forecastthat the global airline industry will
recoverby 2003 from the effects of September110',and traffic demandmight return to the long-term
predictions basedon the relationship with GDP (GrossDomestic Product), which is forecastedto
grow by 3 percent over the next 20 years,during which period air travel will grow about 2
percentagepoints faster than economicswill grow.
i
Chapter 1 Introduction
Advanced aeronautical technology and the reliability of aircraft constitute great
achievements of human ingenuity, but the increasing demand for air travel presents the
air transport industry with some of its greatest challenges. Statistics showed that
aviation industry has it

achieved a remarkable safety record, making currently the safest
form of mass transportation in the world today (Muir and Thomas, 2003). However, if
the risks are measured by comparing numbers of casualties with the number of trips
made, rather than the miles covered, air transport is less safe than ground transport (The
Economist, 1997). As passenger numbers rise, at least one major accident is predicted
to occur every week by 2010 (IFA, 1998), and this is unlikely to be acceptable.
For the airline industry, transport and safety constitute value; each is mutually
dependent on, and worthless without, the other. It is taken for granted that the airline
has done everything practically possible to maintain safety standards by proper
maintenance, operations and training. Nonetheless, even following the greatest

in
advances airframe designs, and the hardware and electronics in commercial aircraft,
accidents still occur. With each accident, public fears about air safety are magnified,
regardless of the cause of the accident. Moreover, risks in the aviation industry are
usually associated with threats to life and body, so an aircraft accident always attracts a
great deal of public interest, frequently resulting in enormous media coverage and a
high impact on the airline's performance. The possibility of a serious and costly impact
on business, perhaps including the company's demise, makes safety an airline's largest
area of concern.
Profit (1995) summarisedthe situation clearly: "Aircraft accident rates are usually
expressedas accidentsper million flying hours, aircraft or kilometres

passenger flown,
and there has been a dramatic reduction in the commercial air transport accident rate
since the 1950s.However, as air traffic is expectedto double over the next decadein
...
terms of annualpassengerhours flown, the number of accidentsper annum due to all
causescould rise, even though the accident rate remained constant. Hence there is a
perception that flying is becoming more dangerous. The downward trend in accident
ratesmust thereforebe maintained if high public confidencein air transport safety is to
be sustained."
2
Given the concerns of carriers in this decade, Hollnagel (1993) claimed that past
accidents should be studied more closely, to see whether something can be learned that
can prevent future accidents. Maurino (2001) also noted that the most widely used tool
for documenting operational performance and defining remedial strategies is the
investigation of accidents. Fatal or serious accidents/incidents often catalyse the
improvement of a safety system, because a thorough accident investigation can reveal
how specific behaviours, including errors and error management, can generate an
unstable or catastrophic situation. Such events can cause an airline either temporarily or
permanently to change the management of its safety system.
Consequently,the analysis of the behaviour of operational personnelin accidentsand

incidents was traditionally adapted to assessthe impact of human performance on
safety. Investigators seek to discover the potentially detrimental behaviours of

operational personnel, in order to identify and manage risks. Such investigations of
human performance enjoy the benefit of hindsight. Risk managementtools have been
accordingly developed to collect safety information and prevent the recurrence of

identified errors.
However, looking only at data after the fact (i. e. after an accident) is a little like trying
to design a good celebration by focusing on the "sweeping up after the parade"

(Maurino, 2001), and is a retroactive approach to safety. The causes of accidents and
the primary contributory factors that are identified through accident investigation do
constitute a form of risk management, especially if the lessons learned are properly
applied. However, so few accidents occur that analysing trends and patterns with such
limited data is difficult. Savage (1999) indicated that another way of preventing the
next accident is needed. One must look beyond the visible manifestations of errors in
designing remedial strategies, to uncover the mechanism that underlies human
contributions to failure (and success) in aviation safety.
It is increasingly recognisedthat organisational risk managementrequires the active
support of managersand employeesat all organisationallevels. Effective prevention
3
strategies must focus upon the identification, removal or amelioration of systematic risk
factors. The control of operational risk in the aviation system may require greater
proactive intervention by airline management in the future. Reason (1995a) and

Johnston (1996) proposed that proactive and systematic risk management approaches
will be more effective in preventing accidents than ad hoc reactions to individual

failures or, than reactive interventions directed towards individual workers.
Within the airline industry in particular, systematicstudy of the origins of accidentsand

incidents have led to heightenedinterest in the way in which organisationsidentify and
managerisk. The activities which are thought to representthe term "organisational

accident", "safety culture" and "proactive approach" are well documented in the
literature, and have beenusedto explain the fact that airlines differ in their willingness
to
and ability managesafety.However, an important but yet relatively undefinedareain
the airline industry is to conceptualisethe `safety mechanism'in proactive safety, and
the factors that influence the mechanism.
0.
1.2 Research Objectives
Accordingly, this thesis is designedto contribute to the pool of knowledge,by meeting

the following objectives:
+ To evaluatecurrent airline safetymanagementsystemsand becomea reference

text which can be usedby academicsand industry.
+ To investigate retroactive and proactive approaches to safety and their
application within the airline industry.
+ To develop a model for a `safetymechanism' in the context of proactive safety
management.
4
+ To verify organisational factors which affect the safety mechanism, and

investigatethe relationship betweenthe factors.
+ To make recommendationsconcerninghow airlines can fit the proposedsafety
mechanism model to the industry for the evolution of a safety management

system.
1.3 Thesis Structure
In keeping with the aims of this researchproject, this thesis is divided into seven
chapters. Chapter 2 is the literature review and empirical industry analysis. The
rationale behind this section is to explore the framework of the airline safety
managementsystem.The results of the analysisserveto explore what hasbeendone in
airline safety management,to identify the problem generatedfrom the current system,
and to verify what is neededfor the continual improvement of airline safety services.
Chapter 3 states the methodologies applied in the research, such as the interview study
and how the safety survey is designed and conducted, in the thesis in order to achieve
the aims stated in Chapter 1 and the problem identified in Chapter 2. Chapter 4 is
designed to develop the safety mechanism model and further determine its influential
factors by employing a safety survey. Chapter 5 presents a qualitative and quantitative
study to test the model and the results, including a case study and a survey of opinions
with questionnaire. Chapter 6 serves to analyse the survey results, to discuss the
structure underlying the model and the implication of results. It also probes into the
applications and limitation of this model. Chapter 7 summarises the knowledge
obtained from this study project as a whole with respect to the development of
proactive safety mechanism, as well as the recommendations for future research areas
to improve airline, safety management system.
Figure 1-1 demonstratesthe researchstructureand methodsused in thesis.
5
Figure 1-1 Research Structure and Methods Used
Why airline safety management system
needs improvement? And in which area?
I
[
Safety/ Risk mana_nement
------------------ -
Accident/Incident t
investigation Organisat ional acc iden ý' Ä
i
Human contribution Cultural dimension

U
Review retroactive tip Review applications of
measures proactive measures
YwI I 0
Interview study Proactive safety mechanism Ro
model development
Case study of Safety survey study ti
r
Result ý.
Result b
Analyses & Discussion fi
Ö'
Conclusion and
Recommendation
6
CHAPTER 2
Literature Review &

Empirical Analysis
"Disasters do not causeeffects.

The effectsare what we call disasters."
WR Dombrowsky,1995
2.0 Role and Definition of Safety
The primary objectives of an airline are associated with profitability, namely providing
services and receiving monetary remuneration. While flying is widely accepted as an
extremely efficient means of quickly transporting people, cargo or equipment, and
performing a wide range of various other activities, safety is not only the compulsory
responsibility of an airline but "safety" also supports airline profitability, for example
through brand image.
However, exactly defining safety is rather difficult. According to the Flight Safety
Foundation (1999a): "Safety is an abstraction, and in a sensea negative one - the
absenceof accidentsand incidents - which makes safety difficult to visualize". Indeed,
comparedwith risks and hazards,hazardsare usually easierto identify than risks, and
thus are easierto measurethrough practical approaches.
Previous literature defines safety as freedom from danger or risk (Profit, 1995). Prof J
Reasonnotes that "safety is a dynamic non-event, so we have to work hard to make
nothing happen" (IFA, 1998). Moreover, McIntyre (2002) arguesthat safety is more
7
Chapter 2 Literature Review & Empirical Analysis
than the absenceof accidents.Safetyis also a goal of reducing the levels of risk that are
inherent in all human activity.
It is a fact that no human activities or man-made systems are absolutely safe. Instead, all
that can be discussed is relative safety and acceptable risk. This concept is reflected in
Lowrance's (1976) argument that "safety is a judgment of acceptability of risk" and the
definition of safety of the International Civil Aviation Organisation (ICAO) that holds
that safety is where: "risks are minimized to an acceptable level". This ICAO
definition of safety is commonly adopted in the modem aviation world.
2.1 External Impact on Airline Safety
Figure 2-1, developed by Dannatt (2000), is used as a guide to understanding the
complexity of the variables that determine the effectiveness of the government
regulation of air transport safety. It illustrates the systematic interactions between
airlines and government, the regulatory agency, airlines, air travellers, the legal system,
the insurance industry, and infrastructure, which, in turn, determine the safety
relationship in the air transport system. The safety relationship concept in this diagram
is found to be useful as a guide to understanding the organisations which exert an
influence on airline safety services.
Most of these organisationsin Figure 2-1 have differing, and sometimesconflicting
objectives. However, these relationships also reveal the supply and demand
relationships within this system, which decide the sustainability of the air transport
industry. As airline serviceshave gonethrough a period of historical developmentand
have therefore formed through a processin which professional,political and economic
interestshave played a part, there is no easysolution to the questionof the right way to
manage airline safety. Therefore, following sections will firstly discuss the external
impact of these organisationswhich are categorisedfrom all elementsin Figure 2-1
including manufacturers,travellers (the public), and regulatory agencies,on airline
safety.
8
Figure 2-1 Safety Relationships in the Air Transport System
Compensation
for loss
TRAVELLERS INSURANCE
Revenue Safe travel

INFRASTRUCTURE:
ATS, AIRPORTS Behaviour modification

AIRLINES
MANUFACTURERS Behaviour modification
2 Regulator's ] LIABILITY LAW

other agenda
STATUTE LAW
ADMINISTRATIVE
REGULATORY AGENCY LAW
LEGAL SYSTEM
Private
Acceptable risk Government
objectives
Obligations
LEGISLATURES LOBBY INTERNATIONAL

GROUPS AGREEMENTS
Note:
1: Quality assurance for underinformed travellers + paternalism
2. Quality assurance for underinformed airlines + remedy for monopoly supply
Source: Adapted from Dannatt, 2000
2.1.1 Technological Innovation and Its Impact
The most important milestone in the history of aviation occurred in 1903, when the
Wright brothers flew a heavier-than-air craft for almost one minute, thus launching the
era of the power plane (Kuang Fu publishing, 1988). By 1910, numerous aircraft
manufacturers were already in business. Sixteen years after the Wright Brothers flight,
the first scheduled air transportation service was launched in Germany in 1919 (Chang,
1998). However, during this period the development of aircraft engines was mainly
driven by speed competitions and trials rather than human transportation.
9
World War I& II further drove the development of aircraft capabilities, and caused
increased attention to be paid to expand their function. It was because military victory
depended on faster speed and extensive damage when belligerent states conducting
offensive military operations In WWI, aircraft were initially used for reconnaissance;
later their combat and bombing role became important in the war (Chang, 1998).
Especially in WW II, in order to control the air supremacy, squadrons of fighters of
belligerent states frequently fought each other or dropped bombs on the trenches when
they could not find enemy fighters to shoot at. Consequently, the aircraft manufacturers
were motivated to develop more advanced fighters to achieve military need, such as
B-17, B-24, etc. During the last few years of the war, pure jets were developed in the
end (Kuang Fu publishing, 1988).
Although in WW II, jet engines did not contribute greatly toward overall success on
either side, in the aftermath of World War II, the use and evolution of jet engines not
only accelerated aircraft speed, but also stimulated a boom in civil aviation. The
increased reliability of aircraft stimulated a rapid rise in passenger travel from the
1940s, while cargo transportation grew significantly owing to the development of the
large freighter Boeing 747 in the 1960s (Doganis, 2002). Flying thus became a
convenient and rapid means of mass transportation, which simultaneously created a
global village. In 1976, an even more advanced technology, supersonic aircraft in the
form of Concorde, entered the service market (Kuang Fu publishing, 1988), marking a
new chapter of aviation. With the growing air traffic in recent years, aircraft
manufacturers are competing to develop new aircraft to satisfy more and more air travel
demand, including A380, supersonic transport aircraft (SSTs), Boeing 7e7 etc. In
particular, the idea of SSTs is to fly as fast as Concorde, but be able to carry more
passengers and consume less fuel and moreover, satisfy strict environmental
requirements (Muir and Thomas, 2003).
Doganis (2002) expressedthat: "In the last.fifty years technological innovation in air
transport has far outstripped that in any other transport mode." Particularly in civil
aviation, comparedto the eraof the piston engine,turbo-prop aircraft have significantly
improved productivity, while the arrival of turbojets has dramatically increasedspeed
10
and reliability. Statistics from 1960 to the early 1980s show that the introduction of jet
engines helped to reduce the rate of fatal accidents per 1 million landings from around
50 to 2. In 2000, a new safety figure revealed that big jet hull loss accident figures from
1991 to 2000 showed a reduction in the loss rate from 1.5 to 1 per million flights; a 33.3
percent drop in the hull loss accident in a decade, which is quite impressive (Learmount,
2001).
2.1.2 The Growing Traffic and Airline Safety
Although the technological developments in aviation were beneficial in aircraft safety

improvement, the increasing size and capacity of aircraft and the speed with which new,
larger aircraft were introduced, created two main problems for airlines, according to the
analysis of Doganis (2002). One is the strong downward trend on load factor, and the
other is the problem of financing the new capital investment. However, paradoxically,
for the last fifty years, the airline industry has been characterisedby continued and
rapid growth in demandfor its services,given the problems.
The rate of growth of air traffic seemsto follow closely developmentsin the world's
gross domestic product (GDP), and there would be no slowing down according to the
air traffic (Doganis,

forecasters3 2001). Although air travel growth varied in different
in the worldwide demand for air travel in the last
regions, strong growth was evident
decade(IATA, 2001). In this new decade,Doganis (2001a) and Boeing (2001) also
predict that world air traffic measuredin Revenue Kilometres

Passenger (RPKs) will
5
grow around percent annually within the next 10 and 20 years respectively. Before
September11th,over the 20-year forecastperiod, 2001 to 2020, the market is predicted

to be worth $3.1 trillion (Boeing, 2001).
3 For example,in the 1990sboth short-term,like IATA (1998), and long-term forecasts,suchas Airbus
(1995), agreewith the fact that internationaltraffic would increasesteadily world-wide at around 5.5
percentper annum until the year 2000, and remain at 5 percent per annum well into the twenty first
century.This should culminate in the world international scheduledpassengertraffic rising to a
forecasted789 million by the year 2010 (Watkins, 1997).
11
With the growing traffic of air travel in the long-term, the demand for safety services is
simultaneously rising. It is because speed and safety are the airline industry's coin of
value, each a mutually dependent value that is worthless without the other. Becker
(1992) revealed, "Public awareness of airline safety issues is likely to increase in the
1990s. The growth will in part be simply a function of an increased number of
passengers and frequency of travel. " In addition, it is found at least 30 percent of air
travellers use perceptions of an airline's safety record as a basis for deciding which
airline to choose; at least 85 percent of respondents would pay more for increased
airline safety procedures. Fifty-five percent of respondents have a clear idea of what
safety information is important and what details they want before choosing an airline
(Becker, 1992). The perception towards airline safety remains an important issue for
the public and the media over these years (Taylor and Hsu, 2001b). In particular, when
passengers consider air travel is unsafe, they will choose other modes of transportation
or rather stop travelling. Taking September 11th,the terrorist attack in the US in 2001
for example, although not the fault of the airlines involved4, the strong impact on the
airline industry caused a decline in air traffic flow to a large degree and consequently
shook the air transport business to the cores. Also the Gulf war in 1991 had a similar
effect but to a lesser degree.
After September 11`h,the forecast of future demand had to be reproduced as a result.
IATA (2002) is forecasting in its new interim five-year forecast that the global airline
industry will recover in 2003 led by traffic on Europe-Middle East, transatlantic and
transpacific routes. Traffic demand might return as predicted in the long-term according
to the relationship with GDP. Given the homogeneous nature of the airline product
(Doganis, 2002), no matter what the result will be, the demand for safety and obligation
of providing safety services to ensure passenger safety is still a prerequisite of airline

business.
4 Although the argumentbetweensecurity and safety remainscontroversial,the measureof the former

is for safety purposeswithout a doubt, which makesit necessaryto include in the context of safety
services.
5 Global international passenger traffic in October 2001, the month after the September 11`hattacks, fell
23.5% (IATA statistics, 2003). In addition, airlines, like Sabena, Swissair, and Canada 3000 were
closed down; some others were caved by capital injections from their government, such as Air New
Zealand, LOT, etc. By the end of 2001, the airline industry was in turmoil (Doganis, 2002)
12
2.1.3 The Regulatory Environment
The roots of today's aviation safety programmes extend back to the early days of
commercial aviation following World War I (Wells, 1991). In 1919 Paris convention
acceptedthat stateshave sovereign rights in the air space above their territory, and
directed the government intervention in air transport. Since then, the airline industry
subsequentlyexpandedand the regulatory environmentwas required for improvement.

Accordingly, a framework of international regulation has gradually evolved in response
to the technical, economic and political developmentsin air transport in the period
1919-1949(Doganis, 2002).
Prior to World War II, international aviation issues were mainly agreed by bilateral
agreements between governments. The "Chicago Convention", signed in 1944, formed
the basis for international standards and recommended practices, and aimed to improve
all aspects of civil aviation world-wide through the providing the framework for the
orderly and safety development of international air transport (Berendsen, 2000;
Doganis, 2002). In 1944, International Aviation Transport Association (IATA) evolved
as a result of the Chicago Convention. The IATA's members are airlines, the parties
working with the results of agreements on a day-to-day basis (Berendsen, 2000). In

1947, the International Civil Aviation Organisation (ICAO), based in Montreal, Canada,
became the world authority. ICAO's standards are incorporated into the programmes of
the authorities, like UK Civil Aviation Authority (CAA), etc. The members of ICAO
are governments who signed the Chicago Convention (The function of ICAO will be
introduced in next section).
The sanction of IATA and ICAO together with bilateral agreementsand inter-airline
pooling agreementcreated a highly regulated operating environment till the 1970s.

Broadly speaking,the regulatory environmentcontainstwo kinds of regulations.One is
those which are economic in nature and concernedwith regulating the businessand
commercial aspects of air transport; the other is regarding technical standardsand

regulations, which cover every aspectof airline activity and aim to achieve very high
levels of safety in airline operations(Doganis, 2002). Yet in 1979and the following two
13
decades, the gradual liberalisation of the economic regulations consequently affects
international air services. In the next sections, the non-economic regulation
environment will be outlined as well as the repercussions of the economic regulations

on airline safety.
2.1.3.1 The current regulatory environment of non-economic regulation
There have been many other international aviation organisations formed to provide
effective solutions for airline safety since World War II. In Figure 2-2, Berendsen (2000)
portrayed the structure of current regulatory environment in safety. It can be
circumscribed by Doganis' (2002) non-economic regulations which deal with airlines,
such as flight operations (aircraft airworthiness, performance, etc), engineering and
maintenance, personnel training and qualification, as well as the third parties, such as
airport and air traffic.
As shown in Figure 2-2, ICAO is situatedat the top. The underpinning philosophy of
ICAO is to have aviation safety directly supervised by national civil aviation
authorities, who adapt the ICAO framework within their countries and assist the
development of practical aspects of the implementation of the Standards and
RecommendedPractices(SARPs) issuedby ICAO. For example,CAA in the UK has
powers to cover aviation in the United Kingdom, while the Federal Aviation
Administration (FAA) acts in a similar role in the US. Besides, although FAA is a
national civil aviation organisation, it can span much of the aviation transport chain.
For example, FAA had launchedits own "safety oversight" proceduresto inspect and
monitor if safety regulations are adequatelyimplemented in certain countries. If the

airworthinessstandardsare deemedas inadequate,
aircraft from such countrieswill not
be allowed to fly into the US.
14
Figure 2-2 Current Regulatory Environment
--- .-,----.
ti . ,-' ,.,... _ .,.,,. y., .,. _.,
ICAO
f+
Annexes with SARPs

" ýe
Regional National
Organisation Authorities
JAR
FAR
JAR, FAR, SARPs plus
Loc if requirements
Professional
Industry
Organisations . .................
Application of regulation
4 Development of regulation
Source: Berendsen, 2000
In addition to national regulators, Berendsen (2000) explained that the need for
harmonisation has resulted in the formation of several supra-national organisations.

One of the most successful examples within Europe is the Joint Aviation Authorities
(JAA). JAA produces Joint Airworthiness Requirements (JARs). The aim of JAA is to
foster the harmonisation of aviation safety across its member States, by implementing
these common regulations and their joint application. Recently, the JAA has worked
closely with the FAA in an attempt to achieve some degree of standardisation between
JARs and Federal Aviation Regulations (FARs), with the intention of harmonising
regulations globally.
It is worth noting that it is up to each state to decide whether it wishes to adopt each
JAR by incorporating it into its own legislation, i. e. JARs are not mandatory, which
arguably result in the shortfall of JAA. As such, the need for the European Aviation
15
SafetyAgency (EASA) seemsto be apparentandhasbeenconsideredto establishsince

2002. From 28 September 2003, EASA, with mandatory power, has become
operational for certification of aircraft, engines, parts and appliances. It aims to help to
maintain a high level of safety and environmental protection in civil aviation.
In addition, EUROCONTROL is anothersuccessfulsupranationalorganisationwhich

provides regional upper airspace ATC services within Europe, and harmonises ATC
servicesacrossits member States(Berendsen,

2000).
There are other groups called Professional Organisations in Figure 2-2. Normally they
are independent, non-profit making international organisations offering international
aviation safety resources. For instance, the Flight Safety Foundation (FSF) is known
internationally for providing timely, practical and objective information to its members
and everyone concerned with the safety of flight. Its independence provides the
aviation industry with a neutral forum to meet and identify safety concerns, determine
solutions and implement ideas and actions to improve safety on a non-competitive

basis7.
2.1.3.2 The impact of economic regulation on airline safety
The consequence of globalisation, privatisation & deregulation
Traditionally, governments are concernedwith many aviation issues.These include
ownership of airlines, regulation of domestic routes and fares, limits to foreign
ownership, etc. and most notably with the safety of air transport.
With the increasingdemandand expansionof networks, globalisation and privatisation

to
encouragesnations adjust to the regulatory environment. Since liberalisation has
been adoptedby the US in 1978, followed by key Europeancountries and European
6 Referencefrom EASA website: http://www. easa.eu.int
7 Reference from FSF website: http: //www. flightsafety. orR
16
Union after the mid-80s (Doganis, 2002; Chang, 2002), deregulation is becoming a
worldwide trend and frees airlines to pursue strategies that offer air travellers greater
value. Many countries have removed restraints within their national boundaries and
revised rules of allowable foreign investment, like the US (Chang, 2002). As a result,
airlines are free to choose where to fly, what services to provide and how much to
charge. Passengersbenefit from the offers regarding lower prices and more convenient
flight times (Boeing, 2001).
While airlines now have flexibility to pursue strategiesthat meet the needsof the next
century's global community, it also comeswith the responsibility of ensuringthat each

operation is safe. A series of questions such as: Have the risks increased as a
consequenceof economic deregulation?Has the deregulatedmarket raised financial
pressureson existing companies causing a reduction in their safety standards?Both
needto be answered.
Researchhas been carried out to find the relationship betweenderegulationand safety
performance after deregulation occurred in 1978 in the US. The conclusion reached
after a 1987 conference at Northwestern University was that "subject to conditions,
safety performance does not appearoverall to have been impaired by deregulation"
(Moses and Savage, 1990; Dannatt, 2000). Morrison and Winston (1988) support the
conclusion that the secular improvement in safety has not been interrupted by
deregulation becausethey found a reduction in insurance expense associationwith
deregulation.Wells (1991) also states"Deregulation is not directly related to aviation
safety,however some industry observersare concernedabout its unintendednegative

impact." Some studies generally conclude that while air transport safety has become
safer after economic deregulationin 1978, it might have been safer still in the absence
of deregulation(Barnett and Higgins, 1989; Barnett and Wang, 1998; Savage,1999).
Although most research results reveal the non-negative aspects stemming from
deregulation,there are still some concerns.Rose (1992) argued that improvements in
airline safety do not appear to have slowed appreciably since deregulation....

Nevertheless,the possibility that regulatory effects may operatewith long lags, through
17
such mechanisms as reduced maintenance or increased aircraft age, suggests that
scrutiny of aggregate safety performance over the next few years is essential. So
Dannatt (2000) pointed out that an important variable likely to have influenced the
safety outcome was the effect of the safety regulator, and regulation's effect on safety is
likely to be the outcome of chosen policies, changes taking place in the industry and the
segments of the industry to which they are applied.
However, in terms of safety issues, regulatory authorities have been criticised for being
untimely and unresponsive, and existing frameworks, with local and regional structures,
do not have the flexibility to adapt to a dynamic, changing market place and
environment (Berendsen, 2000). From a 1985 safety review, a task force reported that
the area in most need of improvement was that of timeliness in identifying and
responding to safety issues (FAA, 1997a). In addition, the cost-cutting strategies, profit
concerns, the effect of mergers and alliance between airlines, etc., resulting from the
competitive environment might also cause airline mission and structure to change. As a
result, some regulators have taken action to ensure that appropriate skill levels are
maintained, like UK CAA. Aside from that, regulations and safety action should not be
perceived as a burden by the airlines. It is also of importance to harmonise the
regulation throughout the industry while maintaining adequate scope for competition.
2.2 Airline Safety Management System
2.2.1 Flight Safety Risk
The airline businessis a high risk business(Smith, 1996; Doganis, 2002). The airline
industry's risk encompassesfactors such asoperationefficiency, industry fundamentals,
competitive position, evaluation of management,financial flexibility, etc. Table 2-1

lists the businessrisk profile in the airline business.Due to the geographicalrange and
complexity of "real-time" the

operations, natureof the in
risk profile the airline industry
is far wider than most other businesses.
18
('huptcr 2 Litcratmc Rri, irtiv cl E'nipiricatl Analysis
Table 2-1 Airline Industry Business Risk Profile
Type Risks
Technical Accidents
Increasing system sophistication
Economic Financial exposure (currency, interest, insurance)
Loss of vital systems or information
Impact on value of brand
Commercial Destructive competition
Loss of key markets
Denial of access to key markets
Impact of legislation or regulation
Political Terrorism or hijack
Requirement to operate uneconomic services
Strikes
Human Loss of key personnel
Error or incompetence
Personal injuries health and safety risk
Operational Inadequate monitoring of control systems
Lack of control over suppliers
Noise
Pollution
Environmental Congestion
Natural disasters
Source: Compiled from Smith (1996) and Sadgrove (1996)
Within the business risk listed above, risks regarding flight safety, whether technical or
operational types, often attract lots of attention by the public, because these are most
commonly associated with threat to life and limb. The consequences are accidents or
incidents (see appendix A for the ICAO definitions of accident and incident).
Figure 2-3 presents the yearly number of accidents and fatality accident rate from 1950s
to 2003, which reveals two trends since 1990: one is the decreasing number of accidents
and the other is the flat accident fatality rate8 in fatal aviation accidents`' according to
aircraft accident statistics (Aviation Safety Network, 2003). The former is decreasing
owing to the industry having invested heavily in developing advanced technology to

increase aircraft reliability and productivity. However, it's almost impossible to
8 Accident Fatality Rate: average percentage of occupants involved in fatal airliner accidents that did
not survive the accidents.
19
Chapter 2 Literature Review & Empirical. -tnal'. ws
achieve zero accident rate even with the advanced technology because other factors,
like human factors1() and environmental factors, can still cause incidents and fatal
accidents. If taking a closer look at the fatality rate in fatal accidents, it shows that in
average 70 percent occupants didn't survive in the fatal accidents, and the trend seems
to grow upward since 2000. The observation demonstrates that the consequences of
fatal accidents may become more and more massive given the advanced aircraft
capability and productivity. The dramatic nature of aircraft accident not only always
attracts the media and grabs the headlines, frequently resulting in enormous media
coverage'', but also has become a matter of great public interest and concern,
sometimes resulting in more serious business risk.
Figure 2-3 Annual No. of Accidents vs. Fatality Rate
oo
90
e ýýf ýlýliiY Roiýýc

E 50
-T Olýln 1 ci Oýnt
ý
eo
3o
xo
p
I'll If;, le 14ý -A, 14ý6 oll I; oil 4p
le le I'll lop le 14P 11.1 '0 '01,4p
Y..,
Source: Aviation Safety Network website http: //aviation-safety. net/statistics
9 Accidents are classified into fatal event and non-fatal event. For fatal event, any circumstance where
one or more passengers die during the flight from causes that are directly related to a civilian airline
flight. The fatal event may be due to an accident or due to a deliberate act by another passenger, a
crew member, or by one or more persons not on the aircraft. These events include sabotage,
hijacking, or military action and exclude cases where the only passenger deaths were to hijackers,
saboteurs, or stowaways.
io Muir and Thomas (2003) state that when discussing passenger safety in very large transport aircraft
(VLTA), also pointed out that VLTA will increase passenger capacity and flight duration, but
emergency evacuation in the event of a survivable crash poses a challenge for aircraft manufacturers
and authorities.
tt For example, data reveal that from 1978 to 1994, the New York Times disproportionately reported
fatal events involving jet aircraft and fatal events in the U. S. or involving US carriers (Curtis, 1997).
20
The impact of aircraft crashes with the associated loss of life is important for the public,
who have the power of veto for a particular airline. Not only do crashes cause loss of
human life, they also damage the viability of an airline. The aviation industry tends to
measure the accident rate according to the rate of fatal accidents per 1 million landings.
However, the perceived accident rate of the public, the media and the investment parties
is the number of accidents per month or per year. Consequently, an increase in the
perceived accident rate may result in many airlines suffering financially as the public
refuse to patronise airlines perceived as less safe. These risks can have a serious
commercial impact on the business of certain airlines, and at worst can cause their
demise (Taylor and Hsu, 2001b).
Accidents like the ValuJet crash in Florida Everglades and loss of the TWA B747 off
Long Island (NTSB, 1997) represent the most tragic risk faced by the airline business.
The airlines managed to remain in business but paid a very high price. In other cases,
airlines such as that of Air Florida, which failed to emerge from bankruptcy in late
1980s in the aftermath of a B737 crash in 1982 (NTSB, 1983), were not fortunate
enough to survive (Smith, 1996; ATI). This is the most serious business risk resulting
from flight safety risk.
2.2.2 The Impact of Accidents on Airline Safety Performance
No matter how severe an accident is, the airline performance in terms of company
reputation, airline operation, fiscal problems, safety commitment, etc. will be affected
to somedegree.Figure 2-4 was developedin order to identify the influence of accident
on airline performance by demonstratingthe sequenceof causes,event, effects, and
influenced performance following an accident.
21
Broadly speaking, three main aspects are identified which contains the operation
performance- crisis management, safety performance, and the financial performance-
costs of accidents. The scale of the impact stems from the effects of an accident. The
result of Hsu and Taylor's work (1999,2001b) shows that with the growth of
globalisation and integration, airlines face new problems in the matter of accident; most
importantly, lack of planning, training, and preparation will contribute towards greater
financial loss. It leads to the conclusions that the most important influence of accidents
on airline performance is the safety performance, and the suggestions for airline
operation and safety improvements.
Figure 2-4 The Impact of Accident of Airline Performance
Caine
Event
t. HnmauFactor
2. Et v ronment ACCIDENT
3. EquipmentFadute
Aisiine Ferfomana
Drasiº.k endP opap' Dana;r
Performance
Operation Urei ru
(CrisisManwaunt)
Public ad
zz
19
SafetyPerfocmance Ac d. ntlnr*4atbn
4
(Management
Improvement) iüýlation
FinancialPerformance M
(Cost of accident)
Evolution
Source: Hsu, 1999;Taylor and Hsu, 2001b
22
2.2.2.1 Safety improvement and management
One of the problems in the aftermath of accidents is the failure of safety commitment.
What are these safety deficiencies and how to improve them in order to prevent
accidents from recurring are issues that airlines are eager to know. Meanwhile,
following an accident, the public and the media are always desperate to know who
should be blamed, who should take the responsibility and who should make the
improvements.
Figure 2-5 shows"The Safety Trinity", demonstratedin the ICAO Accident Prevention
Manual. Theseare the basesof all activities in aviation safety and also the main causes
of an aircraft accident.
Figure 2-5 The Safety Trinity
MAN
MACHINE k) ENVIRONME
Source: ICAO Accident Prevention Manual
Tracing back the causes of accidents in the early years of aviation, the causesof
accidents were mostly associatedwith catastrophic structural failures (Machine) or

adverse weather conditions (Environment). Investigators tended to focus on the
technical aspectsof the mishap under investigation becausethe reliability of aircraft
systems was not always guaranteed.There were, therefore, very good reasons for
investigators choosing that approach at the time. Consequently airlines were led to
focus on the technical approachto safety (Taylor and Hsu, 2001a).
When commercialjets becamecommon transport in the seventies,technology reached
a level of maturity. Accidents causedby hardware failures appearedto steadily decline
23
as a result. The attention then moved to the role of human crew (Man) as a result of a
number of high-profile accidents (see Figure 2-6). It became apparent that accidents
were occurring where the primary cause of the accident could not be associated with a
mechanical failure. It was not until then that the science of "Human Factors in
Aviation" was truly born (Taylor and Hsu, 2001 a). Safety responsibilities are therefore
allocated to those at the operational end: flight crews, air traffic controllers, technicians
and others. Nevertheless, this view should be changed when the whole aviation system
is concerned.
Figure 2-6 Man Causes and Machine Causes
Man
C)
En
CD
\°
0
Machine
Time
Source:ICAO Accident PreventionManual
As Robert W. Sweginnis, a specialistin accidentinvestigation and systemsafety, said,

"Mishaps normally have both technical and managementcauses.Technical causes
identify deficiencies in the operational system. Management causes identify
deficiencies in the managementsystemwhich allowed the operational deficiencies to
exist." (quoted from Aarons, 1998).
Management has great leverage in affecting operational safety within a company.
Through its attitudes and actions, management influences the attitudes and actions of
in
all others an airline. No one would deny that senior management commitment ranks
among the top requirements, and that good communication, employee empowerment
and a high value placed on safety are also considered to be very important.
In addition, outside the company,regulatorsand countries are increasingly recognising
24
the role of management in air safety by holding the management accountable for a
serious incident or accident (like corporate killing). Therefore, management must put
safety into perspective, and must make rational decisions about where safety can help to
meet the objectives of the organisation in the light of the impact of accident on safety
performance. As such, the following sections are served to explore the safety
management concept extracted from accident/incident and how these measures act in a
loss control system.
2.2.3 Definition of Safety Management
Safety improvement and management are part of the countermeasures of a loss control
system (accidents/incidents), which result from changed public awareness and
expectation, regulatory changes, and the development of both civil and criminal
liability.
Since safety is a judgment of acceptability of risk, safety management should be the
technique or practice of managing safety or controlling risk. Overall (1999) points out
that a common definition for Safety Management is "A systematic management of
activities to secure high standards of safety performance". Akhurst and Vivian (1997)
also put it "Safety Management is a mechanism that could be employed to address the
lessons by providing for effective monitoring and auditing of safety and the allocation
of responsibility and accountability in safety critical organisations. " In other words,
safety management simply involves giving safety the highest priority possible in a
safety significant business.
In the airline industry, the exact definition of safety managementmay vary a little in
different airlines according to their businessplans or safety aims.,For instance,Bisson
(1997) pointed out that Britannia Airways defines Safety Management as "all those
activities which underpin the safety and worthiness of the aircraft" in accordancewith
the strategicsafety aim of Britannia - to continue to be safe and reliable airline.
Nevertheless,in the aviation industry, regulatory authorities around the world have
25
defined safety managementin greater detail. The United Kingdom's Civil Aviation
Publication (CAP) 712 definessafety managementas"the systemicmanagementof the
risks associated with flight operations, related ground operations, and aircraft
engineering or maintenance activities to achieve high levels of safety "
performance.
(UK CAA, 2001).
The definition provided by UK CAA is a complex and complete explanationof safety
management in the current aviation industry. It focuses on the risks associatedwith

business,not focusing on safety but a lack of unsafeincidents/accidents.Moreover, it
emphasises that management is systematically in associationwith all risks stemming

from aircraft-related activities. Meanwhile, it contains the concept of systemsfor the
managementof safety.
2.2.4 System Safety and Safety Management System
A study of systems for the management of safety (System Safety) or Safety

Management System (SMS) raised consideration of the three constituent parts -
management, safety and systems. Edwards (1999) puts it "A company's Safety
Management Systems define how the company intends to manage air safety as an
integral part of its businessmanagementactivities. A Safety ManagementSystem is
defined as a systematic and explicit approachto managing risk, and is largely a loss
"
system.
control management
In the US, the System Safety discipline emerged on the engineering and management
scene in 1962 with the dawning of the space transportation era. System safety principles
emphasise the rigorous development of effective safety risk mitigation strategies based
on comprehensive and thorough risk assessment and its management is a long-term,
comprehensive approach that assures that systems and techniques have safety designed
in from the outset (McIntyre, 2002).
Canadian's civil aviation authority, Transport Canada (2001b), says it: "A safety
is
managementsystem a businesslikeapproachto safety.It is a systematic,explicit and
26
comprehensiveprocessfor managing safety risks. As with all managementsystems,a

safety management system provides for goal setting, planning, and measuring
performance."
The United Kingdom National Air Traffic Services (NATS) began the introduction of
formal safety management system (SMS) in 1991 (Profit, 1995), largely because of the
increasing attention on safety matters and airspace capacity from outside groups,
including the public, the media and Parliament. Profit (1995) states, "A safety
management system is no more than a systematic and explicit approach to managing
safety - just as a quality management system is a systematic and explicit approach to

improving the quality of a product to meet the customer's requirement. "
In CAP 712, a safety management system is defined as "an explicit element of the
corporate management responsibility which sets out the company's safety policy and
defines how it intends to manage safety as an integral part of its overall business." (UK
CAA, 2001).
UK CAA also provides an analogybetween SMS and financial management system
of a company.
"The features of a financial management system are well recognised. Financial targets
are set, budgets are prepared, levels of authority are established and so on. The
formalities associated with a financial management system include checks and

balances. The whole system includes a monitoring element so that corrections can be
made ifperformance falls short ofset targets. The outputs from a financial management
system are usually felt across the company. Risks are still taken but the finance
procedures should ensure that there are no business surprises. "
Edwards (1999) also provides a comparison of the financial managementsystem and
safety managementsystem (see Figure 2-7). The managementof safety should fill a
similar place in the organisation'smanagement,in the sameway that a financial system
deals with the control and use of money, providing a framework for managing one of
27
Chapter2 LiteratureReview& EmpiricalAnalysis
the potential loss generators. In other word, the objectives of SMS are to act as a loss
control system and to be focused on actively managing the key risks to an aircraft
operator.
Figure 2-7 The Comparison of Finance and Safety Management System
Company Board Establishes

Management & Direction Safety Management
Financial Management System
System
Establishes Objectives
Business plan & Targets Safety Plan
Targets & Objectives Set Policy Targets & Objectives
Budget Delivers the business Budget
Accountabilities Raises and Approves Budgets Accountabilities
Levels of Authority Allocates Resources Line Management Authority
Procedures
T
Managementof Major Loss
F Procedures
Generators Makes Business
Checks and Balances Monitoring/Line checks
Internal Accoun- Internet Safety

Audit -tants Audit Committee
Audit Balance Audit Safety

Findings Sheet Findin s Achievement
Profit/ Loss Profit/ Loss

Finance Case Safety Case
Source:Edwards, 1999
28
2.2.4.1 The Components of a safety management system
The commitment, organisation and assurance of continuing safe operations are

achieved through a SMS. As such, a Safety Management System must be top
management led and is a systematic approach to managing all aspects of safety in the
business using a structured approach. Profit (1995) describes that in an organisation,
the policies, principles, accountabilities, directives and procedures constitute SMS (see
Figure 2-8). The safety management actions required by the policies and principles are
implemented by directives, as shown by the bubble diagram within the figure.
Directives and their associated procedures can be grouped under the broad headings of
policy issues, incident investigation, safety cases and safety auditing as shown. The
total picture in Figure 2-9 illustrates and rationalises the components of a typical safety
management system in the aviation industry.
Figure 2-8 The Components of a SMS
Policy Principle I Accountabilities d Directives

--
Thefoundation Whatis required Who is responsible Action
Safte
Internal Account- Safety
Investigations groups
abilities manual
Major
Investigation Incident Policy
Imestigaiion IssuES Organisational
change
Direedves
Safety Safety
Auditing
Case
Unit safety Facility safety anal reviews

Interval
cases cases andinspections
surveys
Systemsafet
case
Source:Profit, 1995
29
2.2.4.2 The characteristics of a safety management system
Researchersand organisations(Edwards 1999; Overall 1999; GAIN 2000; and CAP

712) have provided definitions of the basic characteristics (prerequisites) of SMS.
Common characteristicsare:
1. Comprehensive corporate approach to managing safety
2. Effective organisationfor delivering safety

3. Robust systems for assuring safety
The top management of the organisation is responsible for establishing the
comprehensivecorporate approachto safety, but it will fall to the Chief Executive to
ensurethat there is an effective organisationalstructure below Board level to deliver
safety.Meanwhile it is the responsibility

management's to make sure that the systemis
robust enoughto provide safety assurance.
In reality, thesefeaturescannotbe presentedwithout the practice of the componentsof
a SMS. As in
mentioned the previous section,the main componentsof a SMS arePolicy,
Principles, Accountabilities and Directives, which appear sequentially but work
interactively.Policy and Principles define the corporateapproachand the Board has the
corporate approach shown in the statements.This approach is embedded in the
structure of an organisation and everyone is assigned his or her accountabilities.

Eventually, the structuredapproachis implementedby Directives, and aims to create a
robust systemto ensuresafety.
In other words, the characteristicsof a SMS are the prospectsof how the components
work and demonstrate.Therefore, a model is constructed by combining the main

componentsand characteristicsof a SMS (seeFigure 2-9). This model presentsa more
explicit interaction between these three sequentialcharacteristicsof a SMS and other
elementsin the proceedingperiods.
30
Figure 2-9 The Interaction of Characteristics and Components of a SMS
Characteristics Components
Define
Comprehensive Policy + Principles
corporate
approach to Safety policy statement
safety
Effective Assign
organisation Accountabilities
for
delivering Organisational structure
safety
1
Robust Practice
systems Directives
for assuring
safety Instructions
Source: compiled from various authors
1. Policy and principles: corporate approach to safety
Corporate approach to safety is about the involvement of Board/Top management to
show leadership and commitment to safety by clear policy objectives and safety
improvement targets. The "Place" allowing the Board to show leadership and
commitment to safety is "safety policy statements".
In other words, the Board defines and details safety objectives and intentions for safety
standards. These generic ideas are reflected in safety policy statements, which enable
31
to
management demonstratethe fundamentalapproachto managing safety that is to be
adopted in the organisation. As such, the policy statement is a vital starting point.
Safety management principles contemplate the policy statements
In the policy statements, safety management principles, decided by the Board to be
applied within the organisation, are fundamental requirements that define the scope of
the SMS. "What is required and what is achievable" are key questions of the safety
objectives in order to provide a framework for processes to identify safety
shortcomings so that remedial action can be taken (Profit, 1995).
In addition, a safety improvement programme, approved by the Board is an important
and powerful way of keeping the Board's attention on safety. It forces the Board to
review the safety standards and the development of a SMS with regulatory minima. As
such, a company's safety statement should include the following:
Safety objectives
Arrangements for the achievement of safety objectives
Flight safety principles
Health principles
Quality principles
Corporateand safety standards
2 Effective organisation for delivering safety versus accountabilities
Edwards (1999) pointed out seven key areas for effective aviation organisations to
deliver safety.
Committee/structure for overseeing safety management
Management review mechanism
Clarity of line managementresponsibilities

Coherentcascadeof accountabilitiesfor safety
Role ofAccountable Manager (CEO) and a SMS custodian
Change management process in place
Effective company and training requirements
32
In accordancewith thesekey points, Overall (1999) presenteda relative organisational

structure (see Figure 2-10). The consideration of safety is systematically designed into
the management structure, the committee and business of planning, and operation for
an effective organisation. The objective is to ensure that everyone involved in a
safety-significant role is left with no doubt as to his or her individual accountabilities

for safety. The committee is, in the meantime, allowed to oversee safety management
and the management can review mechanisms.
Figure 2-10 The Suggested Organisational Structure
Board
Board Safety Policy
Committee
Chief Executive
(Accountable Manager)
SMS Project Steering

Group/ Safety Review
Review Committee Committee
SMS Custodian
(Safety Manager)
j SMS Project team
Engineering II Flight Operation II Ground Operation
Source: Overall, 1999
33
The coherent cascade of accountabilities for safety is from Board level down through
the management structure. Each safety accountability, within a job specification,
requires certain knowledge, skills and experience. The organisation needs to ensure that
everyone understands his or her individual and collective responsibilities and
accountabilities.
The Board safety policy committee should have inputs from all the senior fleet
managers and other senior managers of the departments. However, the executive
responsibility for safety management rests with the Chief Executive Officer (CEO).
The CEO is the Accountable Manager and is the link between the Board and the
executive. GAIN (2000) defines the Accountable Manager as the person acceptable to
the country's regulatory authority who has corporate authority for ensuring that all
operations and maintenance activities can be financed and carried out to the standard
required by the Authority, and any additional requirements defined by the operator.
An airline's resourcesare controlled by its headquarterand how the resourcesare
allocatedhas a direct impact on the company's safety managementprogramme.Some

regulators have mandated the appointment of an independentsafety expert directly
reporting to the CEO/Accountable Manager of the airline. There are two reasonsfor
this. One is to enablethe organisation'sheadto allocate necessaryresourceson safety
and the other is to provide a Custodianof the SMS when any safety project is planned.
In FAR part 121 (1996), the expert is namedthe Director of Safetywhile in JAR-OPS 1
(1998), it is called Accident PreventionAdvisor who could be the Quality Manager of
the organisation. The exact placement of the safety manager function can vary from
airline to airline, but the critical elements of accessto top managementshould be
maintained because safety reports can then be assured of the proper levels of
assessmentand implementation.
To ensurethe safety managerretains a clear and objective view of the safety of the
operation,ideally he/sheshould have no operationalresponsibilities,but the incumbent
should certainly have considerable operational management experience and the

technical backgroundnecessaryto understandthe engineeredsystemsthat support the
34
operation (Profit, 1995). The role of a safety manager is responsibility for the
development and maintenanceof an effective SMS, promotion safety management,
reporting shortcomingsand monitoring remedial actions.
Generally speaking,operationaleffectivenesswill not be obtainedwithout the adoption
of a rigorous approach to identifying accountability for safety, which is exercised

through an organisation's hierarchical structure of management committees and
meetings,and its consultative arrangementswith its workforce.
3. Robust systems for assuring safety by practising directives
The previous corporate approachand organisational structure have helped to build a

basic framework in which to deliver safety. In addition, safety directives, shown in
Figure 2-8, offer the practices to complete the objectives mentioned and assuresafety
for a robust system.
Safety directives are instructions or procedures for implementing the SMS after
developing the policies, principles and accountabilities suitable for the organisations.
According to his experiences,Profit (1995) indicates Directives which include:
1. Incident investigation: internal investigation and major investigation.

2. Safety auditing: internal surveysand external review and inspection.
3. Safety cases:unit safety case,systemsafety caseand facility safety case.
4. Policy issues:accountabilities,safety groups, safety manual and
organisationalchange.
No sequence exists for these four directives as long as they are kept to the minimum
essential to implement the SMS. Before the important managerial issue in SMS-risk
management is introduced, there is a need to see the link between quality and safety
management.
35
2.2.5 Linking Quality with Safety12
Understanding and recognising quality in the spectrum of civil aviation is important.
From the customer point of view, quality has influence on travel demand and market
share. For both regulators and carriers, the performance of carriers is of concern.
Knowing the information and position can help to enhance the quality of carriers,
especially when the outcomes of a specific aspect of quality, such as air safety, are
engaging people's curiosity.
In 1987, after carrying out a study for more than 5 years, Dumas discovered that quality
programmes and safety programmes have the same components, i. e. successful safety
programmes and successful quality programmes are based on the same solid
foundations (Dumas, 1987). This accounted for the first contribution relating to the
integration of quality with safety (Herrero et al., 2002). Using this idea, Manzella (1997)
affirmed that SMS and quality management system are in need of integrating together.
Figure 2-11 shows that quality and safety principles are essentially the same.
Accordingly, based on the ICAO recommended practice (Annex 6 part! ), JAR-OPS
states that an operator shall establish an accident prevention and flight safety
programme, which may be integrated with the quality system, including programmes to
achieve and maintain risk awareness by all persons involved in operations. They
instruct the operator to design and run a "quality system" with its "quality assurance
programme" to demonstrate regulatory compliance. In addition, the ISO 9000
international standards also help to implement quality systems. It offers some useful
advice that procedures should be documented only where a lack of documentation may
detract from quality. Yet it is worth noting that the decision as to whether or not it does
detract from quality (or safety) is a crucial one and thus one that should only be taken by
a person or committee fully competent to make such a decision.
12 Pleaserefer to Appendix C for the academicbackgroundof quality and safety.
36
Figure 2- 11 The Principles and Relationship of Quality and Safety
SAFETY vs. QUALITY
Goal: Zero accidents Goal: Zero defects
Incident analysis Event analysis
Written policies, procedures Documented policies, procedures

and guidelines and work instructions
Safety committees Quality circles, employee

involvement team
Employee participation Empowerment
Statisticalanalysis Control charts, statistical process

control
All accidents are preventable All non-conformances are preventable
Source:Manzella, 1997
In terms of organisational structure, FAA (GAIN, 2000) suggested that the Flight Safety
Officer has a similar position to Quality Manager. When the management functions of
safety and quality are the same, these two positions can be combined in one, as some
airlines do. Also CAP 712 (UKCAA, 2001) indicated that in most small and medium
sized companies it is expected that the Flight Safety and Quality tasks will have many
common points and there can be no objection to the combination of the roles in one staff
member.
37
Chapter 2 Literature Review& Empirical Analysis
2.3 Risk Management
2.3.1 Definition of Risk and Risk Management
There are various definitions of "Risk" (Profits, 1995; Janic, 2000; Transport Canada,
2001a) and they are all worded in slightly different ways. Yet the underlying concept
remains the same:a chancecausing injury or loss. This concept implies that risk may
involve objectively or subjectively known or assumedexposureprobabilities in relation
to space,people and time-dependency.And the degree of risk will be based on the
likelihood that damageor harm will result from the hazard13togetherwith the severity
of the consequences. As Paries (1996) states "A risk is the product of a given
probability and a given amount of damage".
Civil aviation is an activity where four types of risks are present (Janic, 2000).
Identified by Sage and White (1980), these four risks are as follows:
+ Real risk to an individual, which may be determinedon the basis of future
circumstancesas they develop;

4- Statistical risk, which maybe determinedby the availabledata on the incidents
and accidentsin question;

+ Predicted risk, which may be predictedanalytically from the modelsstructured
and relevant historical studies;

Perceived risk, which may intuitively be felt andthus perceivedby individuals.
To the airline industry, the occurrence of an incident/accident constitutes a known
statistical risk when flying has its inherent real risk. To manage risk involves the
prediction of risk by anticipating and making changesin equipment when the risk is
perceived: As such, risk is

management a technique to manageall these four risks in
this industry.
13 A hazardcanbecomea risk becauseof people,procedures,aircraft and equipment,and actsof nature

(GAIN, 2000). It is an event that has the potential to result in damageor injury.
38
Organisational risk management
Risk managementrelies upon the premisethat the likelihood of an eventhappeningcan

be reduced.In the aviation industry, risk managementis defined by aviation authorities
and organisationsas:
The identification, analysis and economic elimination, and/or control to an
acceptablelevel, of thoserisks that canthreatenthe assetsor earningcapacityof

a commercial airline (GAIN, 2000).
The process of identifying risks, assessingtheir implications, deciding on a
course of action and evaluation the results. In civil aviation, the term is
frequently used in the context of decision-making about how to handle
situations which affect aviation safety (Transport Canada, 2001a).
Modern airlines face a formidable range of risks, ranging from strategicchangesin the
commercial environment, through to the adversecommercial impact of accidentsand

public relations disasters. Management of risk is, therefore, the essenceof safety
management.Knowing the risks enablesresourcesto be more efficiently allocated to
the concerns,and assessedon the basis of severity and frequency so that effort can be
put into the areasof greatestrisk and of significant safety concerns.
Risk management activities and the failure to manage risk involve the expenditure of
resources (FSF, 1999a), whether for the airlines or society. In terms of the former,
Taylor and Hsu (1999,2001a) have identified the impact of accident on airlines'
financial performance and safety improvement activities. It is a truism that effective
organisations actively attempt to manage those risks which potentially impact upon
organisational survivability. That is why Janie (2000) points out that a practical
problem in air transport is how to manage risk and safety. However, the difficult task for
management is to determine which risks carry the most potent dangers (Fischoff, 1994;
Hood et al, 1992).
Therefore, a thorough and systematic risk management process should be made
associatedwith resourcestargetedaccordingly, in order to make the bestrisk allocation.
39
Chapter 2 Literature Review K Empiric ct! Attetll si.ý
2.3.2 Risk Management Process in Safety Management System
As risk is manageable, it is termed risk management. Global Aviation Information

:
Network (GAIN)(2000) depicts the system approach to risk management is known as
system safety, implying that the process of risk management, which is used throughout
industry and commerce, involving identifying work activities and hazards and
estimating, evaluating and controlling the associated risk, is the just tool used to
achieve a SMS. International Federation of Airworthiness (IFA)(1998) used to suggest
a safety loop, which is a cycle of activities including hazard identification, risk
assessment, risk control and recovery, and feedback in order to implement SMS (see
Figure 2-12).
Figure 2-12 Safety Loop
1. Identify the 2. Assess the

hazards risks
Safety
4. Feedback Management 3. Put in place
information on System measures to
the performance control risks (or to
of these measures recover if things go
to enable appropriate wrong)
organisational
learning
Source: IFA, 1998
In aviation operations, not all of risks can be eliminated; some risks can he accepted and
some can be reduced to an acceptable level. Figure 2-12 demonstrates the sequential
procedures to a robust SMS are the processes by which risk can be identified, measured,
evaluated and controlled so that the highest standards of safety can he achieved. The
whole process follows a logical pattern. The first step is to identify the hazards. The
second step is to assess the risk stemming from hazardous activities and determine
whether the organisations are prepared to accept the risk. The third step is to find and
40
.\
identify the defencesthat can control the risk. The fourth step is to examine whether
risks are appropriately managed and use the feedback information to evaluate
organisationalchanges.
In other words, risk management is SMS in the making. It is effective risk management
that contribute to the robust SMS. Profit's (1995) model, previously presented in Figure
2-8 (page 29) which shows a robust SMS can be achieved by maintaining safety levels
(eliminating risks) through the directives and procedures (practices), provides a
discussion base for the following sections, which will explore these four steps
describing in Figure 2-12 in detail and investigate current risk measures used in the
airline industry.
2.3.2.1 Hazard identification
Hazard identification is a systematicexamination of potentially hazardousactivities to
establish safe, effective procedures and practices. There are many ways to identify
hazards,which might be obvious or latent in operations.The most important thing is
that hazard identification should be undertakenon a frequent basis dependingon the
complexity of operationsand associatedrisks. The following are someuseful methods

of identifying hazards:
1. Incidentlaccident investigation
Incident/accident investigations help to find out the causes of mishaps or serious
occurrences.This is also classified asone of the most important identification processes

for hazards.As such,this sectionaims to firstly introduce how accidentsare analysedin
order to identify the accidentcausesand then illustrate the role of incident investigation
on hazardidentification.
Analytical methods for accident investigation
An accidentis always deemedas a failure of risk managementand is a brutal eruption

into damagemanagement(Panes, 1996). As such, accident investigation has a clear
41
role within the safety process. Accident investigation is the appropriate tool to uncover
unanticipated failures in technology or rare, bizarre events. A proper accident

investigation can reveal how specific behaviours, including errors and error
management, can generate an unstable or catastrophic state of affairs (Ho, 1996; Hsu,
1999).
Ferry (1988) distinguishesover 20 types of accidentanalysisapproaches,including:
" Events sequencing

" Known precedent
" All cause/multiplecause
" Codes,standards,and regulations(CSR's)
" The four M's of man, machine, media, management
" Re-enactment
" Reconstruction
" Simulation
" Epidemiological
" Hazard analysis documentation
" Inferential conclusions
" Programme evaluation review technique (PERT)
" Critical path method (CPM)
" Failure mode and effect analysis(FEMA)
" Technique for human error rate prediction (THERP)
" Fault tree analysis (FTA)
" Changeanalysis
" Management oversight review (TOR)
". Scenario modelling
" Preliminary hazard analysis
Two most common accident analytical approachesare "Event Sequencing" and "All
cause/multiple cause". The former is to select the cause initiating the sequenceof
eventsthat lead to an accident.The latter is to identify the primary causewhich is most

for
responsible the accident.
Importance of incidents
Although the analytical tools of an accident are many, and experts can use them
successfully to trace the causation, accidents are too rare to provide enough data for
research. Therefore, an incident investigation is called. According to the Heinrich
42
Pyramid (sec Figure 2-13), for every major accident in a given endeavour, there will be
3-5 less significant accidents and 7-10 incidents but there will be at least several
hundred unreported occurrences (the exact numbers may vary in different airlines).
Figure 2-13 The Heinrich Pyramid
NTS
UNREPORTED
OCCURRENCES
Source: Adapted from Hart, 1999
Apart from the reason that incidents outnumber accidents, so they provide more data for
analysis and investigation, there are another two reasons to investigate and study
incidents. Firstly, incidents reveal similar hazards as accidents, but they are not as
severe as accidents and they will not result in serious adverse legal or financial
consequences. Secondly, more information is available from the people involved (Ho,
1996).
Safety investigation seeks to identify causes of the incident and recommend the
necessary remedial action to reduce the risk of recurrence. For a complex system like
the airline industry, it needs to conduct a more formal and detailed investigation using
an investigation team from a higher formation in the organisation with specific

investigation or technical experience.
2 Incident reporting systems
In 1947, the pioneering work of Fitts and Jones in developing the "Critical Incident
Technique" helped to establish the value of the investigation of incidents. Interviews
and written surveys were used to examine errors made by crewmembers in utilising
43
cockpit instruments (Nagel, 1988). Their research systematically showed the
significance of poor human engineering in incident generation and accident causation.

Lauber also stated the importance of incident and an incident databases which he
described "... is a veritable gold mine of information waiting to be tapped" (quoted from
Ho, 1996). Clearly, this shows that the incident investigation is important but the source
of incident database incident reporting systems are even more important to develop.
-
National reporting systems
ICAO Annex 13 recommendsto its member Statesthe provisions about the incidents
reporting system:
1. The requirementto establishincident reporting systems
2. A requirementfor the investigation of seriousincident.
Some countries have set up their own national reporting systems, for example, ASRS
(Aviation Safety Reporting System) (U. S.), MORs (Mandatory Occurrence
Reports)(U. K. ), EUCARE (European Confidential Aviation Safety Reporting System)

(Germany), CAIR (Confidential Aviation Incident Reporting Network) (Australia), and
SECURITAS (The Confidential Transportation Safety Reporting System Program)
(Canada), etc. The quality of their databases is attributed to the cooperation and
provision of airlines, and the information provided by airlines is attributed to the
established reporting system in their companies.
Reporting systems of airlines
The existenceand health of a reporting systemin a company can create accessto top
management and ensure that safety deficiencies are recorded.Reporting systems,such

as BASIS (British Airways Safety Information System) and ASAP (Aviation Safety
Action Partnership)14,
are createdfor airline operationwith two major goals: to identify
safety concern and to provide methodsof corrective action.
14 BASIS: Createdby British Airways. Gather,categoriseand analysesafety information including

incident reports and digital datausing modular system.
ASAP: A joint project betweenthe Allied Pilots Association, the FAA and the American Airlines.
44
Take BASIS 15for example. It is used by over 100 organisations for safety management.
BASIS was developed by safety professionals to answer such questions as "How safe
are we? ", "Where should we put our limited resources to become even safer?" The
modules used contain:
1. Air SafetySystem
This module is used to process flight crew generatedreports of any safety-related
incident. This was the original BASIS module and provides analysis in an
exceptionally easy-to-useformat.
2. Auditing System
The BASIS Audit module has been designedto store and analysedetails of JAR Ops
(Flight Operations,Engineering,and Ground Operations)and Health and Safetyaudits.
3. Cabin SafetyReporting
Safety incidents in the cabin are now starting to receive the attention they deserve.
Violent, abusive and/or unsafebehaviour by passengersare among the issuesthat the
aviation industry is trying to better understandand ultimately address.
4. Ground Found OccurrenceReporting

Aircraft maintenanceplays a vital role in ensuringif aircraft are airworthy for services.
It is thereforeessentialthat reportsraisedby ground mechanicsrelatedto aircraft safety
are taken seriouslythen storedand analysedin a similar way to air safetyreports (ASRs)
generatedby flight crew. This module was developed to collect and analyse such
maintenance reports.
5. Ground Handling Reporting

There is an increased awareness of Ground Handling events within the aviation
industry and this, together with the financial cost of damage and the operational
disruption caused by delays, has made Ground Handling Reporting a key element of an
organisation's safety management programme.
6. Human Factors Reporting

In safety managementthe ultimate tool would enable the Flight Safety Managers to
know all the causal factors behind their next incident before the investigator has a
reason to write a report.
is Sourcesare from GAIN (2000) & BASIS website: http://www. winbasis.com/
45
7. Maintenance Error Investigation

Before an organisationcan hope to tackle maintenanceerror it must first identify what
maintenance errors are occurring and more importantly why; i.e. what were the
contributory factors?This module usesinformation derived from an interview process
known as MEDA (MaintenanceError Decision Aid) which wasjointly developedby
Boeing and a number of airlines including British Airways.
8. Safety Information Exchange

Users extract and send their data quarterly to the BASIS team. The data is de-identified
at source and merged into one global database which is then distributed to those users
who have contributed data. The merged database is sent out every quarter and contains
incidents occurring during the preceding 12 months.
Factors affecting implementation
To obtain a high degreeof incoming reports, the most important characteristicsfor a
good reporting system should contain three characteristicsincluding Confidentiality,

Anonymity and Feedback(Ho, 1996;Bacchi et al, 1997).Confidentiality is the number
one guideline that must be strictly obeyedin any working place. Confidentiality means
the reporter's namewill only be known by thoseauthorisedby the system.Keeping the
reporter's identity from being disclosed or discussedby the third party should also
receive higher priority than the reported caseitself.
It is also important to inform the reporter aseachstageproceedsand assurethe reporter

that his/her work is highly appreciated.Trust in the system must be encouragedby
direct and indirect feedbackto the aviation community. Reportersmust be allowed to
in
see value their programmeparticipation through feedback,which can foster their
the
willingness to continue reporting their experiences,i. e. feedback must demonstrate
results and value of contributing.
Meanwhile, during the implementation of a voluntary reporting system, a reporter's

trust is the main key to the successof any reporting systems.Bacchi et al (1997) point to
some issuessuch as programme publicity and the availability of the reporting forms,
saying that cannot be underestimatedalthough publishing the reports is a good way to

enhancethe exchangeof information. For example, the forms should be designedas
postage-paidand should be available to all potential users.Additionally, the provision
46
of a structuredguide would enhancethe quality of the reports' information and prevent

returnedreports being incomplete.
To sum up, if everyone in the company is able to discuss the incidents frankly without
fear of punishment, and have access to the reporting system with confidence, similar
incidents might not occur and future accidents can be avoided. Anonymity should be
the last resort. Freedom from job loss is the key. This is the major aim in developing
incident reporting systems.
3. Safety audit
During the 1950s and 60s, safety professionals created a different measureto assess
safetyeffectiveness- the audit, when problemswith accidentmeasuresbecameobvious

(Petersen,1998).The theory behind the audit is: Accidents are either unforeseeable(the
true accident) or unforeseen(oversight). If firms can dictate what actions to take in
order to prevent accidents in advance, it can then measure how well these
predetermined actions are executed. As such, the prime objectives of any safety
auditing are to determine safety standardsand enhancesafety, according to Hamilton
(1998).
There are various types of audit, such as checklist; yes-notype audit; quantified audits;
and audits that end with scoresor points awarded, etc., just like quality management
systemseliminate risks as a matter of routine by objectively examining all aspectsof a
unit's activities that impinge on quality. Auditing meansthe checking of compliance
with working practices againstproceduresand It
standards. is one of the tools used to
check the quality of safety programmesand identify any hazardoussituations.
Safety auditing can be also fulfilled by internal safety surveys,and external review and
inspection becausean honest and critical self-audit is one of the most powerful tools
that managementcan employ to measure flight safety margins (FSF, 1999b). The
methodology of self-audit is used by senior 'airline management to identify
administrative, operational and maintenanceprocessesand related training that might
47
present safety problems. The results are used to focus management attention on areas
that require remedying to prevent incidents and accidents. Most of important of all,
Arbon et al, (1990) argue that audits have confirmed that organisations which have
deliberately fostered an attitude of mutual respect between various categories of
employees enjoy a higher level of morale, and operational reliability and maintenance
standards.
2.3.2.2 Risk assessment
After identifying the hazards,the next stepin the processis to critically assessand rank
risks. Since a general conception of risk is the chance, in quantitative terms, of a

defined hazard occurring, it therefore combines a probabilistic measure of the
occurrenceof the primary events with a measureof the consequenceof those events
(Warner,1992),i. e. two considerationsare in need:the likelihood of the hazardand the
severity of the These

consequences. are reflected in the hazard analysis techniques
introducedbelow.
Hazard (risk) analysis
Hazard analysis is the application of methods to identify hazardsand associatedrisk

(UKCAA, 2001). It is basedon prediction, and the accuracyof the results is dependent
on the correct identification of all significant potential hazards,and on the accuracyof

the data analysed(Profit, 1995). This enablesdefences(risk control) to be developed
and contingency plans to be producedand implemented. There are various techniques
to perform risk analysis. The following are some tools frequently used in the airline
industry (SeeAppendix D for the details):
1. Probabilistic Risk Assessment (PRA)

2. Hazard Mode and Effect Analysis (HMEA)
3. Fault Tree Analysis (FTA)
4. Event TreeAnalysis
5. Flight Operation Risk Assessment (FORAS)
6. Risk Analysis Matrix (RAM)
7. Risk specific safety index products - performance indicators
48
2.3.2.3 Risk control/ recovery
The essenceof this stage is to develop and implement appropriate measures to prevent
occurrence(risk control) or to recover if things go wrong (risk recovery). As long as

defencescan be identified, risks are under control.
Identify the defences
The objective of the hazard identification is to provide the organisation with a
technique for early identification of the risks to which it is exposed. Risk management
requires that once hazards are identified and their risks are ranked, the defences, which
may exist as suitable physical or procedural controls to prevent an unplanned release of

the hazards should be identified. Therefore, introducing the identification of control is
to reduce the severity and likelihood of a hazard, to levels of "as low as reasonably
practical" (ALARP). This is obtained by using methodologies for the removal,
reduction, or control of the hazards or the threats that could release them.
Table 2-2 demonstratessome risk control programmesexisting in airline operationsto
prevent the top two accident causes: CFIT (Controlled Flight into Terrain), ALA
(Approaching andLanding Accident). Theseprogrammesare diagnostictools designed
for remedial actions.
Take CFIT for example. It is describedby the Flight Safety Foundation as "When an
airworthy aircraft, under the control of the flight crew is flown unintentionally into
terrain, obstaclesor water, usually with no prior awarenessby the crew." It can happen
at anytime during the flight but most often occurswhen an aircraft is preparing to land.
CFIT has been a large focus for severalyearsbut still remainsa large killer in aviation.
Studies have shown that the industry could prevent more than 80 percent of total
aviation fatalities by eliminating CFIT and approach/landing accidents (Cooper, 1996).

Several recommendations have been made to ICAO to help control this problem.
Among the suggestions is that of implementing Ground Position Warning System

(GPWS). First developed in the early 1970s, the GPWS concept was one aspect helping
49
Chapter' Litt, aliere"Rriit -tit N E'nipiiirul. -Inali'.s'is
to slim down the number of CFIT accidents and it began appearing in large aeroplanes
by 1973. ICAO mandated GPWS in the 1970s and today more than 95 percent of the
world's airline fleet have them installed. Furthermore, the NTSB has encouraged the
use of Enhanced Ground Proximity Warning System (EGPWS) which some of the
major US carriers have begun using on a voluntary basis (Air Safety Week, 2001).
Meanwhile, the use of Flight Data is of significant importance. It not only increases
safety by looking for and addressing weakness in operation, but also helps management
to make sound decisions and monitors the effectiveness of those decisions. Flight
Operation Quality Assurance (FOQA) data provides a precise record of flight
parameters and crew actions. It is a routine downloading and systematic analysis of
aircraft parameters that were recorded during flight either by the crash-protected
recorder or the Quick Access Recorder (QAR).
Table 2-2 Risk Control Tools for Top Accident Causes
in ii ii 1" I)I OR (l)itital I liý'Iit

FOQA (Flight Operation Quality Assurance) )ata Recorder)
TAWS (Terrain Avoidance Warning Systems) AA said, by the end of March
005, the total US fleet will be
1. CFIT GPWS (Enhanced Ground Proximity
: trofittcd with lA\VS (EGI \V' ).
Warning Systems)
(Controlled Flight otne airluic< in the IS
Into "Terrain) PWS (Ground Position Warning System) nplcmcnt u mi i %olunt; uv hasis
Ireadý
...............
VISAW (Minimum Safety Altitude Warning)
2. ALA
(Approaching and and Landing Accident Reduction
Landing Accident)
TCAS (Traffic Collision Avoidance System)

QAR (Quick Access Recorder) Valuable diagnostic took
Providing information tivni
3. Others
normal flights
Precision-like_approaches
Services difficulty report
Source: compiled f om AIR SAI=L 11' WE 1-,K. 200 1 and carious journals
>O
Holtom (1999) statedfrom a Flight Operation perspective,a FOQA programmecould

identify the following:
" Non-compliance and divergencefrom StandardOperatingProcedures(SOPs)
" InadequateSOPsand inadequatepublished procedures

" Ineffective training and briefing, and inadequate handling or command skills
" Fuel inefficiencies and environmental unfriendliness
The ultimate aim of FOQA is to identify these operational shortfalls so that risk
prevention strategiescan be conductedand safety significance eventscan be avoidedin
the future.
2.3.2.4 Feedback & organisational learning
Effective feedbackis an important constructif organisationsdecideto prevent incidents

There
and accidents. are two main tasks in this stageaccording to Figure 2-8 (page 29)
and Figure 2-12 (page40):

Thoroughly documentthe processand its results (Safety case)
,k' Feedbackinformation on performanceto enableorganisationallearning
1. Safety casedevelopment
The concept of Safety Cases has been adopted in several industries, such as nuclear,
chemical, rail, air transport and so on as a means to demonstrate safe operation. More
specifically, the safety case is a systematic and structured demonstration by an
organisation to provide assurance, through comprehensive evidence and argument, that

the aircraft operator has an adequately safe operation. It aims to identify the hazards
faced by a company and how to establish the potential for harm from those hazards
through risk assessment. In practice, the safety case is a documented description of the
major hazards that the aircraft operator faces and the means employed to control those
hazards (Profit, 1995; Edwards, -1999). The organisation will have identified and
assessed the major hazards and safety risk and be able to demonstrate that they can
51
manage them to levels which are as low as reasonably practicable (Edwards, I 999a).
The essential feature is that safety case should identify potential weaknesses in the
process and then identify the measures in place to mitigate or control the risks and
explain how the risks are managed. Only when the procedures are documented can it
help to prevent accidents/incidents from occurring. For example, Standard Operating
Procedures (SOPs) are a set of procedures that provide operators with step by step
guidance for their task. Standardisation ensures the best method of operation and makes
sure employees behave in a consistent and predictable way (See Appendix E for
detailed explanation of documentation).
2. Organisational learning and change
After having identified the hazards, assessedand controlled the risks, and documented
the safety case, the next step is to manage the learning through feedback information.
Effective feedback is important within the context of inter-organisational learning
(Smith, 1999). Stacey (1993) also claims that the whole issue of feedback is central to
the systems approach to learning. It therefore shows the need to discuss organisational
learning.
Marquardt and Reynolds (1993) state: "Learning organisationsteach their employees

the critical thinking process for understandingwhat it does and why does it. These
individuals help the organisationitself to learn from mistakesas well as success.
" This
notion clearly explains that organisational learning is affected by a group rather than
individuals working alone.
"Learning" meansthe adjustmentof coping mechanismsbasedon a new understanding
of the world. According to Shrivastava(1983), learning is a framework of knowledge

development, utilisation of knowledge in understanding the cause and effective
relationships,and perspectiveon knowledge and expertise.Organisational learning also

includes working and dealing with practical and experiential learning, and innovation
diffusion as well as the transfer of that experiential learning. To be effective,
organisational learning should involve a change in the core beliefs, knowledge and
52
Chapter 2 Literature Review &cEmpirical Analysis
assumptionsheld by both managementand operators,which also leadsto effective and

systematic changes to the operation of the organisation.
Changes might be trivial, minor or major. All of them have to be managed in order to
maintain control of their activities. The range of organisational change is from the
whole business, down to local departmental changes. Any major organisational change
should be accompanied by a.formal analysis and evaluation of its safety management
implications. The value of safety should be learned and viewed as a core function value
of the organisation and as such, as a value that will accrue to the benefit of the airline,
its employees and its customer base.
Nevertheless,Choularton (2001) points out that achieving a level of learning which

the
addresses root causesof disasters,is difficult. Recentresearchshows us that while
effective learning about hazards is a common assumption of such attempts,

organisations can be very resistant to learning the full lessons from the past
accidents/incidents and mistakes (Pidgeon and O'Leary, 2000). Smith (1999) has
suggesteda number of barriers to learning, as summarised in Figure 2-14. Many
barriers listed are consequent on the fundamental problems associated with
communication and core belief. It is becauseorganisations may ignore the human
aspectsof causality which are central to crisis incubation and the learning process.
Figure 2-14 Barriers to Organisational Learning

Barriers to Organisational Learning
Rigidity of core beliefs, values and assumptions

Lack of corporate responsibility
Incrementalistapproach (failure to deal with emergency)
Reconstructionand projection
Focus on a single-loop learning
Peripheral inquiry and decoy phenomenon
Centrality of expertise, denial & the disregard of outsiders
Ineffective communication& information difficulties
Cognitive narrowing and fixation (reductionist)
Maladaption,threat minimization and environmental shits
Source:Smith, 1999
53
Human reliability
Hammer (1972) states "A popular misconception is that by eliminating failure a product
will be safe. A product may be made safer by eliminating or minimising failure, but
there are other causes of accident... mishaps often occur where there is no failure. " The
importance of Hammer's work lies in the fact that he recognised there are other causes
of accidents: dangerous characteristics of the product, human action, extraordinary
environmental factors, or combinations of these (McIntyre, 2000). To err is human.

Human errors now account for 80 percent of aircraft accidents (IATA, 1975; IFA, 1998).
Yet many of these accidents could have been avoided if the basic concept of human
factors had been observed. As such, the main aviation safety authorities around the
world (such as FAA, Transport Canada, JAA) have undertaken a series of initiatives,
including the consideration of Human Factors in Operations, Certification and
Maintenance. Some may focus on research, publication of guidance material and the
promotion of Human Factors Programmes without changing the regulatory framework,
while others decided to enhance their regulations by embedding human factors
concepts within them. It is suggested that the cooperation and efforts of regulation
systems should certainly provide more effective controls for the human element and
reliability of airline safety services (JAA). As such, there is a need to have an in-depth
investigation to probe into the relationship between human aspect and SMS.
2.4 Human Contribution
2.4.1 Human Factors versus Risk
A Greekmyth that hasbeenadaptedto tell of the first man-madeaviation disasteris the
story of Icarus flying. It is said that in order to escapefrom Crete, Dxdalus, who was
imprisoned by King Minos, built wings of feathers and wax for himself and his son,
Icarus. Warning Icarus not to fly too close to the sun, the two took to flight.
Nevertheless,Icarus did not listen to what his father said and flew towards the sun.As a
result, the and

wax melted16 Icarus plummeted to his death.
54
This story told about the causeof an accident-human factor- failure to follow SOPsor
principles, which not only made the journey at risk but also resulted in the horrible
consequence(causalfactors). From the modem investigation point of view, in addition
to causal factors, what are the "contributing factors" to the accidents?i. e. why Icarus
failed to follow SOPs?Maybe he knew "the higher you fly, the colder it is", so he
decidedto ignore what his father told him- the wax would melt; or maybehejust simply
did not trust his father and disregardedhis teaching.As such,the study of humanfactors
starts from investigating causal factors of accident/incidents towards exploring the

contributing factors of dangerousactivities for improvement.
No human activities can be carried out without risk, particularly in a high-risk industry
such as the aviation business,which "people" design, build, operate, maintain, and
manage.The failures of people involved in the daily routine of operationsare often the
symptomsof deeperdeficiencies at the foundations of the system.Grose (1987) states
that from a risk managementpoint of view, an inability to absorbthe consequencesof
unsafe acts and omissions is ultimately consideredto be a symptomatic failure of the
overall risk managementsystem.
Therefore, understanding human factors in aviation has become the subject of

increasing attention ever since it was highlighted at the IATA's 1975 Operation
Symposium in Istanbul, Turkey. Later, ICAO acknowledgedhuman error researchas
the most relevant opportunity to increasesafety. In 1989, in order to reflect the results
of researchas to the human factors in flight operationsof that time, ICAO statedthat:
"The expansionof Human Factors awarenesspresentsthe international aviation

community with the single most significant opportunity to make aviation both safer and
more efficient. "
16 As a matter of fact, the higher you fly, the colder it is. Therefore, the ending of this story should be
correctedand replacing "the wax melted" to "the wax crackedand fell apart becauseof the low
temperature".Another possible cause,provided by the aircraft accident investigator - Frank Taylor,
was that Icarus got high enoughto reach "coffin corner" and consequentlylost control without
having suffered any structural failure at all. After all there had been no opportunity to do any flight
testing!
55
An accident can ultimately be deemedto derive from a systemwhich is inadequately
specified or designed,or which has insufficient defences-in-depth(ICAO, 1994). This
notion means human factors not only play an important role in causing
incidents/accidents(insufficient systemor design),but also in preventingthem (as long
as defencesare identified).
`Human factors' is a strange and possible ungrammatical name for a discipline or study.
Nevertheless, it has come to be used to encompass all of those considerations that affect
man at work (Green et al, 1996). Thus, human factors involve gathering information
about human abilities, limitations, and other characteristics and applying it to tools,
machines, systems, and environments. In aviation, human factors is studied for a better
understanding of how humans can be integrated with the technology and working
environment with most safely and efficiently approach. In the Human Factors Guide for
Aviation Maintenance published by FAA (1995), Human Factors is defined as:
"Human Factors refers to the study of humancapabilities and limitations in the workplace.
Human Factors include, but are not limited to, such attributes as humanphysiology,
psychology,workplace design,environmentalconditions, human-machineinterface, and
interaction of humans,the equipmentthey use,the written and verbalproceduresand rules
theyfollow, and the environmentalconditions of any system."
To best illustrate the concept of human factors, a SHELL model (shown in Figure 2-15
a, b) is adopted. The SHELL model was first developed by Professor Elwyn Edwards in
1970s. Since then, this model has been widely applied to the field of aviation human
factors and has become a valuable tool in aviation safety. Hawkins (1993) modified this
model by summarising the main concerns of human factors, subsequently adopted by
ICAO.
56
Figure 2-15 The SHELL Model
(a) (b)
cH
cL o
cI
Source:Taylor, 1999 Source:Hawkins, 1993
These two figures in fact have slightly different focal point although they have the same
elements.There are four main componentsin the SHELL model: Software, Hardware,
Environment and Liveware. Each element symbolises different feature as followings:
S= Software (the non-physical aspects of the workplace, such as documentation and
computer)
H= Hardware (equipment and machinery,such as the layout of the aeroplane)
E= Environment (the workplace and generalsurroundings,such as weather)
L= Liveware (human element,i. e. other people working as team members)
The SHELL conceptrecognisesthe importanceof people, equipmentsand environment,
software and their interactions with each other. The difference between the Hawkins
and Edwards models is the interfaces produced between each element; the former
shows four interfaces while the latter presentsthree more interactions between these
elements.The following are the interfacesdemonstratedin both models.
Liveware- Liveware:
This interface concernsthe interaction of humanbeings. It emphasisesthat humanwork
needs to be recognisedas a team-basedactivity; otherwise this work will not function

well. As such, communication, leadership,sharedsituational awarenessand so on are
emphasisedto avoid the misunderstanding. Crew Resource Management (CRM) is
57
designed to accomplish this goal.
Liveware- Hardware:
This interface concernsabout the man-machinesystem. It is mostly dealt with by the
scienceof Ergonomics. For example, different socket in the aircraft design to prevent
from any mismatch and potential hazards.This interface has been the focus of most
Human Factor attention in the past.
Liveware- Software:
This interface encompassesthe non-physical aspects of the system, such as procedures,
manuals, quick reference handbook, computer programmes, etc. Aircraft engineering

has now attracted lots of concerns because of its heavy reliance on manuals, computers
and paperwork. Procedures are ambiguous or badly written or translation causes
misunderstanding and even cost incident or accidents at the worst.
Liveware- Environment:
This interfacerepresentsa very important interface,becauseof the ways in which tasks
and situations can combine with human limitations to create unsafe acts and quality
lapses.
The following three interfaceswere showed in the Edwards model in addition to the
previous four.
Software- Hardware
This interface is concernedwhen consideringthe non-physical aspects-machineof the
working place. For example, proceduresare designed that are not compatible with the
instrumentor electronic database(manual)on board requirestoo complicatedcomputer
operation to obtain the information.
Software- Environment
This interface concerns with the non-physical aspects of the working place
system. For instance, the ground handling procedures or loading

-environment
be carried out in the rain and darkness.
procedurescannot
58
Hardware- Environment
This interface deals with the machine-environment system, as aviation is developed
based on adapting the environment to match human requirements. For example,
instruments or lights cannot be seen in bright sunlight from some angles or similar
background. To solve this problem, Cranfield University is now conducting a research
regarding this interface. Dr. Tony Head from Human Factor Group was leading a
project studying glider conspicuity trails (Head, 2003).
Taylor (1999) indicates that Edwards' model can best illustrate all the interactions
between these elements when viewing the whole aviation system, while Hawkins'
model failed to draw the attention to all of the interfaces.This is becausean essential
part of Hawkins' model is on the interfaces between the Liveware and other
components,i. e. his conceptrecognisedthe importanceof human elements,which have
important interactions with the people, equipmentsand environment. In other words,
Liveware is the heart of the model. The SHELL model revealed a very important tool
for air safety, as "Human Factors" has been developed progressively to enhance
by
aviation safety, promoting the understandingof predictable human limitations and
its applicationsin an attempt to managehuman error.
2.4.2 Human Error
Hollnagel (1993) reveals that the estimated involvement of human error in the
breakdown of hazardoustechnologies increasedfourfold between the 1960s and the
1990s,from minima of around20 percentto maxima of more than 80 percent.Statistics
presentan estimateof human error in different industries (seeTable 2-3), showing that
human performance (problems) dominates the risks in hazardous industries; the
variability of percentageof failures may come from the type and accessof reporting
system. However, especially in the aviation industry, it is now widely recognisedthat

80 percent of all aviation accidents are the result of human error (IATA, 1975).
Therefore, the obvious dangeris, as ProfessorJamesReasonsays,"Challenge is either
you managehuman error or human error will manage you! " (IFA, 1998) This notion
implies the involvement of greatercost and greater danger.
59
Table 2-3 Estimates of Human Error in Different Industries
Industry Percentage of all failures
Jet transport 65-85

Air traffic control 90
Maritime vessels 80-85
Chemical industry 80-90
Nuclear power plants (US) 70
Road transportation 85
Source:IFA, 1998
2.4.2.1 Cost of human error
In a significant part of accidents, maintenance error is one of the main causes, which is
accountable for up to 25 percent of all aviation accidents resulting from human error
(JAA, IFA, 1998). Maintenance is a highly error-provoking activity, regardless of who
is doing the job. Statistically, there are 600,000 removable parts in an aircraft so there
are many chances for error. The commonest error type is leaving out necessary steps
during installation. Incorrect installation accounts for 60 percent of maintenance error.
Goglia (2000) even announced the fact that "the actual experience in the US was: in the
last 5 years FAR operators had suffered 14 hull losses, 7 of which were attributable to
maintenance or engineering failures. That is 50 percent of the total. " According to GE's
(General Electric) estimation, maintenance error is around 20-30 percent of in-flight
shutdown (IFSD) and costs $500,000 each time (IFA, 1998). Other data presented by
Boeing indicates that it costs around $10-20,000 or more per hour of maintenance
related delay, and $50,000 or more for each flight cancellation (Boeing, 2001). For
airlines this figure is greatly exceeded. Between 1988 and 1991, a US airline with a
fleet of more than 300, encountered 203 recorded maintenance mishaps, which resulted
in 13,299 out-of-services hours, and cost $16.5m in repairs, excluding lost revenue,
which is likely to amount to many more millions of dollars (IFA, 1998).
Perrow (1984) mentions that we are facing the growing complexity of the systemin our
60
technological environment. In the aviation industry, accidents have many causes of

failures arising from different levels of causes. As such, a basic knowledge of Human
Factors is a crucial part in aircraft maintenance, operation and other department
expertise. to improve safety through understanding of human error, it may be useful to
address errors as symptoms (contributing factors) rather than causes of accidents
(causal factors). In addition, there is a need to predict where errors might occur and to
prevent them from happening and causing any financial costs. Error can be managed
systematically and cost-effectively. This is central to error management.
2.4.2.2 Error Management (EM)
Major accidents inquiries (e.g. Three Miles Island, Challenger, King's Cross, Herald of
Free Enterprise, Piper Alpha, Clapham, Exxon Valdez, Kegworth, etc.) indicate that the
human causes of major accidents are distributed very widely, both within the system as
a whole and often over years prior to the actual event (Reason, 1995b). The ICAO
Human Factors Manual states:
"Although humanfailure is thepredominantfactor contributing to aviation accidents
and incidents, it has never been clear what aspectsof human capabilities and
limitations should - or could - be addressedby training. On the other hand, it has been
Equally clear for someyears that Human Factors education and training within the
aviation systemcould be improved."
Indeed,ashuman error is frequently a precursorto failures of risk managementsystems,
a thorough understandingof error is

management required in order for suchmeasuresto
be improved and made effective.
Wiener (1995) points out that error management (EM) can be viewed as involving the
tasks of Error avoidance, Error detection and Error recovery. In addition, there are three
essential facts of EM: (1) Human fallibility unchangeable, (2) Everything we do is
vulnerable and (3) Key activities are interconnected. It reflects the spirit of EM- you
cannot change the human condition, but you can change the conditions under which
people work. Moreover, in an operational context, errors are caught in time and do not
61
produce damaging consequences. Counter-measures to error should not just look at
avoiding errors,but to make them visible and trap them before they producehazardous
This is the essenceof error management.
consequences.
Over the last 20 years,behavioural scientistshave learneda great deal more about the
varieties of human error, their mental origins and the factors likely to promote them. It
is increasinglyclear that "bottom-up" analyses,startingwith the investigationof human
error (Rasmussen,1985; Reason,1990), ultimately reach the sameconclusions as the

"top-down" approach adopted in recent disaster investigations - namely that human
error must be viewed in context and that risk-reducing defenceswill necessarilybe

multi-faceted if they are to match the complexity of the operational context and tasks
(Johnston,1996).
Previous researchhas been focused on finding human factors that link to pilot-error
accidents through accident investigation. The traditional aspects of human factors

(errors) addresscommunication,stress,and ergonomics,such as:
+ Improvements in engine, aircraft system, flight deck design and passenger cabin
design are made to reduce the accident rate and increase efficiency
(manufacturerside).
+ Analyses of aircraft accidents reveal that only 30 percent of crew related

accidentsresultedfrom technical skill failures on the part of individual. Instead,
they show that between 60-70 percent of crew related accidentsare causedby
team managementbreakdown. This has led to Crew ResourceManagement
techniques being developed and applied on the flight-deck. Thus, CRM is
intended to facilitate teamwork and communicationbetweencrew membersto
reducethe incidence of errors (Helmreich, 1997; Saleset al., 1999).
Human Reliability analysis

Much of the effort of human reliability analysis is dedicated to reducing the
uncertainty abouthumanperformanceby producing or obtaining a more precise
estimate of numbers used, i. e. refines the probabilities for the defined events
(Hollnagel, 1993).
62
Gradually it became clear that erroneous actions during maintenance shared the same
importance as erroneous actions of design and system operation. The rationale and
basic principles of comprehensive error management in terms of engineering and
maintenance is developed and named HERO (Human Error Reduction Operation). It is
an "operation" because it can take many forms. There is no infallible way. Different
mixes of techniques and practices suit different organisations. The important thing is
that HERO should be based on sound error management principles. Comprehension
and the judged relevance of this kind of training material has been trailed successfully
in a number of aircraft maintenance organisations (British Airways Engineering,
Singapore Airlines Engineering Company, Cathay Pacific, etc) (IFA, 1998).
The following are somemeasuresthat are effective to achieveHERO.

+ The QantasHuman Error & Accident Reduction (HEAR) Programme
One of the most successful HF programmes is HEAR, developed by Qantas in 1995 for
maintenance engineers following a series of repeated incidents. The programme was
organised and delivered by a group of frontline engineers with strong management
support: a "shop floor" approach. Error-Reducing Conditions (ERCs) are focused on to
achieve the aim.
+ Team Resource Management - Maintenance Resource Management (MRM)
The term MRM was originally used as a parallel to CRM (Crew Resource Management)
but has evolved somewhat over the years as it has been appreciated that CRM concepts
be directly to the maintenance engineering context. An example is
could not all related
Continental Airlines, who has established a CRM-type of course for their engineers.
They are claiming marked declines in ground-damage incidents and dispatch delays as
a result.
+ MaintenanceError Decision Aid (MEDA)

Developed by Boeing, MEDA is a maintenance error investigation tool, for
investigating maintenance lapses. It has been developed as a project to provide
maintenance organisations with a standardised process for analysing contributing

factors or errors and developing possible corrective action.
63
"The Error Iceberg" is used to explain where MEDA should be applied (see Figure
2-16). The basic idea behind MEDA is "Errors are seldom random". Errors originate in
As
a workplace or system. such, the investigation occurs at two levels.
1. Line investigation: MEDA begins with a paper-basedinvestigation that gives line
engineersa standardisedway of identifying the causesof lapsesand of preventing

their recurrence.
2. Organisational trend analysis: MEDA then provides the means for examining past
lapses using a computer-basedtrend analysis. MEDA is designed to provide a
common languageto increasecommunication betweenmaintainers,regulators and

manufacturers.
MEDA is a tool that provides front-line engineers with a principled means of detecting
and removing error-provoking factors. It is a complete and wide-ranging process with a
paper-based investigation at its core. By 2000,104 carriers globally have received

MEDA training from Boeing. It is the most widely used Human Factors based
maintenance error investigation tool in aircraft maintenance today (Chapman, 2000). In

the USA, about 33 percent of airlines use MEDA, about 33 percent are thinking about it,
and about 33 percent have decided not to use it due to concerns about vulnerability to
regulatory action and litigation, particularly in the US. Many Canadian airlines are
using the MEDA process, as are several UK airlines (UKCAA, 2000). (See Appendix
F for other MEDA-like approaches)
64
Figure 2-16 The Error Iceberg
Presentlevel of
investigation (if any)
Operationally
Significant Events Where MEDA shoul start
(IFSD's Delays, Cancellations)
Events without operationalsignificance Where MEDA go
Source:Chapman,2000
To sum up, although all organisationsoperating hazardoustechnologies use various

forms of EM, the main purpose of error managementis rather pragmatically to enable
specific system changesto be made in responseto specific unwanted occurrencesas

well as an effort to improve the systemdesign.With human factors awarenesstraining,
everyone in the organisation is given the tools to make the organisation an improved
working environment.
2.5 A Retroactive Approach to Safety
Accident investigation and incident reporting system
Human imperfection remains a fact of life. Learning the relevant lessonsfor prevention
is the primary reasonswhy aviation accidents are formally investigated and in such
painstaking detail. Accident investigation, therefore, has an apparentrole within the

safety process.It is the appropriatetool to uncover unanticipatedfailures in technology
or rare, bizarre events. Hollnagel (1993) indicated that in order to prevent accidents
65
Chapter2 Literature Review& Empirical Analysis
from happeningit is necessaryto take a closer look at the accidentsthat have happened,
to seeif somethingcan be learned from them. Maurino (2001) also points out that the
most widely used tool to document operational human performance and define
remedial strategiesis the investigation of accidents.
A proper accident investigation can reveal how specific behaviours, including errors
and error management,can generatean unstableor catastrophicstateof affairs. Serious

or fatal accidents often serve as the catalyst for improving the safety system. Such
eventscanmake either a temporaryor a permanentchangefor a companyto manageits
safety.
However, there are limits to the lessons available through this process because
investigation always serves purposes other than accident prevention. To identify the
type and frequency of errors, or discover any training deficiencies, is possible but this is
only the tip of the iceberg. Reason (1995) argues that most accident investigations tend
to stop when answers are found to the proximal cause, responsibility and prevention
questions. Consequently it usually takes a long and expensive public inquiry to identify
the underlying organisational failure types. Should accident investigation restrict itself
to mere retroactive analysis, its only contribution in terms of human error would be
increased industry database, the usefulness of which remains dubious. (Maurino, 2001)
Incident reporting systems are better than accident investigation for understanding
system and operational human performance. Incidents are more significant markers
than accidentsbecausethey identify and signal weaknesseswithin the overall system
before it breaks down. Their value lies in pinpointing the concern.Nevertheless,there
are limits to the value of this information. The main limitations are: Incidents are
self-reportedso the processand mechanismunderlying an error may not reflect reality
and it therefore captures the external manifestations of errors only. Heimreich and
Merritt'7 describeincident reporting thus: the incident is like a broken bone that sends
to the doctor.The doctor setsthe bone, but rarely considersthe root causes-weak
17 From httr)://www.psyiitexas.edu/DsyLhelnueic]Vlocalsol.
htm (09/08/01)
"Local Solutionsfor Global Problems: Theneedfor Specificity in AddressingHuman Factors
Issues." By Heimreich R., and Merritt A., AerospaceCrew ResearchProject, The University of
Texasat Austin.
66
bones,poor diet, victim of abuse,high-risk lifestyle? Therefore, setting the bone is no
guaranteethat the patient will not present again next month with anothersymptom of
the sameroot cause.
Risk management tools and human factors
Analysis of the behaviour of operationalpersonnelin accidentsand incidents has been

the method adaptedto assessthe impact of human performanceon safety traditionally.
Becauseof the recorded negative outcomes,investigators are engagedin discovering
the bad behaviours of operationalpersonnel.While they examinehuman performance
in safety occurrences,they enjoy the benefit of hindsight. Risks can be identified and
managedso that they are reducedto tolerable levels.
In the past, many improvements in safety have been reactive i. e. response to errors.
Risk management tools are developed in order to collect the safety information and
prevent the identified errors, such as DFDR, QAR, GPWS and CRM training. However,
Maurino (2001) argues that DFDR and QAR providing information from normal flights
are valuable tools but these can not yield information on the human behaviours leading
to provide the context in which to diagnose the problems. Also, the limitation of FOQA
data is that no information is recorded about why particular actions were taken
(Helmreich, 2001).
Error management developed after human factors issues were recognised and errors
were found to be managed. Like the observation of training behaviours (e. g. flight crew
simulator training), EM is another tool which is of help in understanding operational

human performance. Nonetheless, training behaviours are sometimes biased towards
safety due to only an approximation of behaviour during line operation if everything is

done by the book.
A retroactive approach
Maurin (2001) stated"Looking only at data after the fact is a bit like trying to designa
good celebration by focusing on sweepingup after the parade", which is a retroactive
67
to
approach safety. Accident causesserve as contributing and primary factors to this
approach while risk management, human factor and error management are the
diagnostictools. Theseafter-fact measuresare as it should be, but the difficulty now is
that there are so few accidentsthat it is difficult to analysetrend and patterns.Savage
(1999) indicated that we need another way to prevent the next accident. In order to
the
uncover mechanismsunderlying the human contribution to failures and successesin
aviation safety, we need to look beyond the visible manifestations of errors when
remedial strategiesare designed.
It is increasingly recognised that effective organisational risk management requires the
active support of managers at all organisational levels. Effective prevention strategies

must focus upon the identification, removal or amelioration of systematic risk factors. It
is suggested that the control of operational risk in the aviation system will require
greater proactive intervention by airline management in the future. Johnston (1996)

further proposes that proactive and systemic risk management approaches will be more
effective in preventing accidents than ad hoc reactions to individual acts of failures, or
reactive interventions directed to individual workers. Therefore, a proactive approach is

in need of investigation.
2.6 Proactive Safety
2.6.1 What is Proactive Safety?
The concept of proactive safety was not officially acknowledged in the aviation
industry until the mid 1990s,when a new and effective safety approachwas actively
being sought. In contrast to a reactive approach, researchershave emphasisedthe
importance of a proactive approachto safety in aviation (Maurino, 1996; Johnston,
1996; Merritt and Heimreich, 1996; Savage, 1999; McFadden and Towell, 1999).
Merritt and Heimreich (1996) concur that "With regard to safety, it is important to be
proactive, rather than waiting for incidents and then in

reacting a band-aid fashion. To
this end,periodic safety audits can identify weaknessesin the system." The "band-aid"
application progressessafety through human factors knowledge in the aftermath of
68
damage. It has fought a consistent battle against latent systemic failures and has
progressed towards improving system safety, but in a limited way. Therefore,

identifying and eliminating latent system failures to achieve safety is not a new problem,
but is a crucial one, and one in need of a solution (Merritt and Heimreich, 1996).
Aviation is an industry where people need to interact closely with technology to achieve
the production goals. In particular, the introduction of advanced technology has added
the challenge to the air transport industry because technology is widely used as a means
to improve safety, but it is used without cultural consideration. The cultural

transformation of these two airlines- British Airways (BA) and Scandinavian Airlines
System (SAS)(See Appendix G) have presented a new dimension for aviation
professional to investigate, which is that the prospective changes in the company are the
way to bring about organisational change (Ho, 1996) and a proactive approach to safety.
The only way of proactive management being present in every working place is through
the organisational and cultural level of company. Cultural transformation can decrease
the latent failure and active errors and prevent incidents and accidents from occurring.
Therefore, following sections will firstly introduce the two dimensions underlying the
proactive safety - organisational and cultural aspects in 2.6.2 and 2.6.3 respectively,
followed by demonstrating the proactive programmes in section 2.6.4, such as LOSA
(Line Operation Safety Audit), to show the importance to detect the latent human error.
2.6.2 Organisational Accident
The airline industry representsa complex systemof multiple interactionsthat is highly
sensitive to the physical environment and the passageof time. Since organisational
structures have adapted procedural methods since 1970s, collective system failures
have naturally followed. The causes of accidents have turned out to be more
complicated than a single causalreason.Lauber (1996) points out that human factors
studieshave changedaccident investigation from fault finding only, to an opportunity
to find system or organisational problems. The term "organisational accident"
gradually emergedand describeshow accidentscan occur or managementinitiatives

can fail becausethe organisation was not well prepared or not well suited to the
69
Chapter 2 Literature Review & Empirical l nah'sis"
initiative. Reason's model (1990) provides a theoretical framework of accident
causation to illustrate a more profound understanding of organisational accident.
The Swiss Cheese Diagram (see Figure 2-17) shows how accidents are the consequence
of a series of failures in a system, which, as shown in the figure, include fallible
decisions, line management deficiencies, precursors of unsafe acts, unsafe acts and
inadequate defences. Each slice of cheese symbolises a condition of a company, and the
holes on the slides of cheese are the breaches of each condition. Whenever these
features happen to combine, there is the possibility of the occurrence of an accident.
Reason (1990) also notes "In considering the human contribution to system disasters, it
is important to distinguish two types of error: Active errors, whose effects are felt
almost immediately, and latent errors's whose adverse consequences may lie dormant
within the system for a long time, only becoming evident when they combine with
others to breach the system's defences."
Figure 2-17 Reason's Organisational Accident Model
Latent defects ýl
active failures Accident
defences
Latent
DECISION
ý ine management
deficiencies
Lw
Fallible
Source: Rcýised torn Reason, 1990
is Reason replaced the term "latent errors" gra(luaIIv as "latent conditions" in his subsequent study.
This well-known cheesemodel is redrawn as Figure 2-18 to better illustrate where and
how active errors and latent failures occur. It shows the various levels of a system and
their possible involvement in contributing to an accident or incident. By using systemic
analysis, the relative importance of each level in preventing a major accident can be
manifested. Traditionally, acts (a person's behaviour) and conditions (the physical work
environment) represent immediate accident causes. These have been viewed as unsafe
acts and conditions. As Figure 2-18 shows, all accidents result from a combination of
specific situations that consist of individual actions and workplace conditions. Two
causal pathways are identified in this model: an active failure pathway running from the
organisation via the workplace conditions to the actions of an individual or team, and a
latent failure pathway that runs directly from the organisation processes to the defences.
Figure 2-18 A Model of Organisational Accident Causation
Defences
Organisation Workplace Person/team H Outcome

Management Error and Errors
decisions violation and Accidents
andorganisational produce violations
ýj ý/
process conditions
Latent failure pathway
º AF: Active failures
"""""'* Latent failures Source:Adapted from Reason,1995
As mentioned previously, with the advance of technology and the reliability of aircraft,
modem aircraft accidents are generally not the result of mechanical failures, but are
mainly due to human factors. However, traditionally in the airline industry human
performance has been considered as separate from the context or system within which it
takes place, such as "pilot-error", "controller-error", or "maintenance-error". It is not
adequate to describe the complex ways in which accidents happen because this ignores
the component of the decision-making process (the management) i. e. the airline
industry's multi-level human involvement. Beaty (1995) puts it, "Modem aircraft
71
accidents result from collective mistakes rather than individual errors". Moreover,
Edkins (1998) points out that "Aircraft accidents have a positive correlation with latent
failures, arising from the broad management functions of an organisation". Latent
failures are decisions or actions originating within management that have damaging
consequences but may lie dormant for a period of time. They combine with local
workplace factors, and errors or violations usually committed by operational personnel.

If system defences are breached, the result may be an accident.
The latent condition has changed the trend in favour of finding systemic or
organisational problems. It also illustrates the effects that management's efforts can
have on instilling a culture where safety is an operational value. Given the unique role
of management in culture, many cultural strategies originate with and/or require the full
enthusiasm of management. But Merritt and Helnireich (1996) argue that before any
action can be taken, in order to strengthen or alter the culture, there must be a clear
perception of the existing culture.
By definition, the commonly accepteddefinition of safety is "risks being minimised to
an acceptablelevel". Acceptable risks are, however, reachedat either an individual or

group level. Each individual's decision will be governedto a large degreeby his or her
personality. At group level, the combination of personalitiessimultaneouslycreatesa
culture (Braithwaite and Caves, 1997). This implies the coherentrelationship between
safety and culture. Meanwhile, Maurino (1996) suggests a need to re-visit the
conventional views on human errors. Not only must human and organisational
behaviourbe consideredas inseparablefrom the contextswithin which they take place,
but alsoprevention endeavoursmust build upon proaction rather than reaction.As such,
following sectionsaim to explore the cultural dimension and its importance in airline
safety together with the proactive to

approaches safety and its correlation with culture
dimension within the airline industry.
72
2.6.3 Culture and Safety
2.6.3.1 The concept of culture
Culture is defined as the ideas,customs,and art of a society (Collins, 1995).Gradually

the conceptof "culture" hasbecomewidely usedin the field of social science,in which
multiple meanings of the term are available. Among them, two main models can be
found. According to Rohner (1984),
"There are thosewho view culture as being behaviour; the regularly occurring, organised
modes of behaviour in technological, economic, religious, political, familial and other
institutional domains within a population. In contrast to the various "behavioural"
models of culture are a group of theorists who hold that culture is a symbol system,an
ideational system,a rule system,a cognitive system,or in short, a systemof meaningsin
the headsof multiple individuals within a population."
Whether culture is seenas being behaviour, or a sharedmeaning systemwith certain
observablebehavioural consequences,a more exact understandingis required for the

presentstudy.
Consistentwith the view of Hofstede (1980), who defines culture as "the software of
the mind", Merritt and Heimreich (1996) provide the following definition of culture:
"Culture can be defined as the values, beliefs, rituals, symbols and behavioursthat we
share with others that help define us a group, especially in relation to other groups.
Culture gives us cues and clues on how to behave in normal and novel situations,
therebymaking the world less uncertain and morepredictable for us."
They also postulate that culture encompasses two components: Surface structure and
Deep structure. The former is constructed. with observable behaviours, while the latter
provides the logic guiding the behaviours.
The definitions of culture reveal the fact that cultures are specific to a defined group of
people. Therefore, Hofstede (1994) argues that culture should include layers of national
culture, regional/linguistic/religious culture, gender culture, generation culture, social
class culture and organisational culture, since people are usually part of a number of
groups.
73
In keeping with this theme, Heimreich and Wilhelm (1997) identify and discuss three
intersecting cultures that surround every flight crew; national, professional and
organisational cultures. Heimreich (1999) points out that many professions such as
aviation have strong-cultures and develop their own norms and values along with
recognisable physical characteristics such as uniforms or badges. His research reveals
that professional culture can contribute to aviation's splendid safety record, but the
"macho" attitude of invulnerability can lead to risk-taking, failure to rely on fellow
crew members, and error.
Morley (1999) also identifies a similar stratification of culture which includes layers of
national culture, industry culture, organisational culture and organisational safety
culture. Industry culture is included to reflect the norms, attitudes and values in
association with an industry and to illustrate the fact that different industries may have
their own set of systems of meaning to distinguish them from others.

C-
Summarising from these researchers' work, three main and distinct levels of culture are
found: national, organisational and safety culture. These are presented in Figure 2-19
and will be discussedin the following sections.
Figure 2-19 Layers of Culture
National Culture
Indust Culture
Professional Culture
Organisational Cultur
Safety Culture
Source:compiled from Heimreich and Wilhelm 1997; Morley, 1999
74
2.6.3.2 National culture
National culture is undoubtedly an important influence in aviation (Johnston, 1993;
Maurino, 1994; Merritt and Heimreich, 1996). One of the most influential individuals
in this field of culture variation is Geert Hofstede (1980), who conducted a
questionnaire study across IBM, a large multi-national corporation. The questionnaire

data was collected from 80,000 IBM employees in 66 countries across seven
occupations. Four dimensions of national culture were identified:
Dimension 1: Power Distance

This is the extent to which differences in power among people are expected and desired,
i. e. the unequal distribution and exercise of power which is expected and accepted in a
culture. Countries high on power distance (e. g. China) demonstrate dependence of

subordinates on their superiors. Leaders are expected to be autocratic and decisive, and
subordinates are unlikely to approach their superiors. In countries low on power

distance (e.g. Australia), superiors are likely to consult with their subordinates.
Dimension 2: Uncertainty Avoidance
This is the degree to which members of a culture feel uncomfortable with risk or
unknown situations.Members in a high uncertainty-avoidanceculture (e.g. Japan)tend

to be intolerant of unstructuredor unpredictablesituations. On the other hand, cultures
that have low uncertainty avoidance (e.g. Denmark) are likely to value stability and
rules in their daily lives, and acceptand encouragedissentingviews.
Dimension 3: Individualism-Collectivism
This relates to the extent to which members of society value individual achievement
over group membership and goals. In a highly individual culture (e.g. the US), people
are expected to act according to their own interests, while a collective culture values
loyalty to and harmony with the group, so that people tend to act according to the
interest of the group (e.g. Singapore, Taiwan).
Dimension 4: Masculinity-Femininity
This is the extent to which differencesin genderroles are valued. In a high masculine
75
culture, members place a high value on assertiveness and toughness in males and
tendernessin females(e.g. Japan).In a high feminine culture, membersvalue a welfare

society and both gendersare allowed to display feminine traits (e.g. Sweden).
However, Merritt (1997) found all four dimensions, apart from Masculinity-Femininity,
when replicating Hofstede's survey by using commercial pilots as respondents. As a
result, Merritt suggests that this dimension can be absent because aviation is already a
financially rewarding profession and has, therefore, little concern for masculine traits
such as "the opportunity for high earning".
Heimreich (1999) notes that Individualism-Collectivism and Power Distance are two
related dimensions of national culture having particular relevance for aviation. He also
suggested a third dimension, labelled Rules and Order, which is similar to Hofstede's
concept of Uncertainty Avoidance. Members in a high Rule and Order culture believe
rules should not be broken and written procedures are needed for all situations (e.g.
most Asian countries). Those low on this attitude show lower concern for rules and
written procedures (the US, European countries).
2.6.3.3 Organisational culture
The concept and essence of culture has been linked increasingly with the study of
organisations. With the recognition of the symbolic aspects of organised settings have
come calls for a cultural perspective on organisations (Turner, 1971; Whorton and
Worthley, 1981; Smircich, 1983). Similar to national culture, organisational culture
provides a behavioural referent for members of an organisation by defining what the
organisation does, and does not, represent. An organisation's culture reflects its attitude
and policies about human error, and the openness and trust between employees and
management.
The characteristics of organisational culture
Although organisationalculture is a complex concept (Guldenmund,2000), a number
76
of researchers have tried to define and describe the characteristics of organisational
culture (e.g. Schein, 1992; Robbins, 1992; Hampden-Turner, 1994). For example,
Hampden-Turner argues that corporate culture is describable, measurable if necessary
and, within limits, alterable. He describes the characteristics of corporate culture as

follows:
1. Individuals make up a culture: Culture comes from within people. They use
the culture to reinforce ideas, feelings and information which are consistent
with their beliefs.
2. Cultures can be rewarders of excellence: Culture is a source of motivation. It

creates an environment for bringing out the potential of all its members and a
system for rewarding defined tasks.
3. Culture is a set of affirmations: No group, corporation, tribe or nation can

start from nothing. Culture proves that.
4. Culture affirmations tend to fulfil themselves: Cultures create results and

consequences, no matter whether for good or for bad.
5. Cultures make sense and has coherent points of view: One cannot
understanda corporateculture without seeingthat its actions logically follow
from its beliefs. Culture can be studied.
6. Cultures provide members with continuity and identity: Only if beliefs are
shared, affirmed, fulfil themselvesand retain distinctive meaningsover time,
despite changing environments,can a corporation retain its senseof identity
and continuity.
7. Cultures are patterns: A culture is no particular thing or object, but a pattern

which appearsboth through time and across the organisation. So cultures are
structuredfrom repeatingevents.
8. Cultures are about communications: Culture can produce solidarity. It is

most important to grasp that many cultures facilitate communication, sharing
of experience and information. They can make their members strongly
supportive of eachother.
77
Guldenmund (2000) summarises and lists the organisational characteristics from
previous researchers and literature review as follows:
1. It is a construct: When operationalising a construct, it is generally assumed

that there are several variables that co-vary or fit together to form a unified
whole.
2. It is relatively stable: Researchershave found a period of stability of at least

five yearsfor organisationalculture.
3. It has multiple dimensionality: Dimensions are always composites,

comprised of several variables.
4. It is something shared by (groups o) people: Culture is something that is

mutual and reciprocal. It is a synergistic aggregate composed of several parts.
The characteristics are the base for assuming multiple cultures within a large
organisation, in that such an organisation can be divided into divisions,
departments, units, etc, that will have developed their own culture.
5. It consistsof various aspects:This meansthat severaldifferent cultures can

be distinguished within an organisation(e.g. safety culture, service culture,
etc).
6. It constitutes practices: Culture is perceived as'having multiple layers. At

each level, culture has particular manifestations which can be studied
separately.
7. It is functional: Culture is functional in the sensethat it suppliesa frame of

referencefor behaviour.
Although there is some disagreement with numbers 1 and 3 between researchers,
overall, organisational culture is a relatively stable, multi-dimensional, holistic
construct shared by (groups of) organisational members, which supplies a frame of
reference and which gives meaning to and/or is typically revealed in certain practices
(Guldenmund, 2000). It is not suggested that a'single organisational culture is needed
universally. Successful airlines can be characterised by different cultures and
operational styles, just as many different personality types can make good airline pilots,
cabin crew, ground staff and so on.
78
Organisational culture and safety
Organisations are complex entities and safety is just one aspect of their function.
Naturally, different organisationsdemonstratedifferent abilities to managerisk and to
realise the long-term benefits of the effective management

of safety.The difference in
the ability of organisationsto managetheir affairs with respect to safety has led to a
wide-ranging treatmentwithin the literature of the characteristicsof organisations.
The relevance of organisational characteristicsto safety is obvious in the previous
section when discussing the characteristics of organisational culture. Since
organisationalculture is a multi-dimensional construct and consistsof various aspects,

the link between organisational culture and safety is therefore clear; Korean Air is a
good example". As Hayward (1997) indicates, organisationalculture has the potential
to have a very significant direct impact on the safety performance of organisations.
Merritt and Helmreich (1996) alsonote that it is organisationalculture which ultimately
shapesworkers' perceptions of safety, the relative importance placed on safety, and

members'activities regarding safety.
Figure 2-20 portrays schematically the concept of the organisation-culture relationship.

Culture is imported into the organisation through its membership. Its presence is
believed to be revealed in the patterns of attitudes and actions of individual organisation
members (Smircich, 1983). This figure also embodies the concepts of Merritt and
Helmreich (1996):
"An integrated organisational culture can be characterised by sub-group

cooperation, a strong corporate identity ... high employee morale, all of which
create a positive impact on service and safety. "
15 TakeKorean Airlines (KAL) for example.Since 1995and 1993,Delta Air Lines and Air Canadahad
code-shareagreementswith KAL respectively.Unfortunately, KAL had suffered 10 serious
accidentsfrom 1990 to 1998.By anyone'sstandard,thosenumbersrepresentthe symptomsof a
systemin trouble. Following the crashof a Boeing MD-11 freighter in China in April, 1999, Delta
Air Lines and Air Canadaimmediately suspendedits code-shareon KAL flights pending a thorough
review of the Korean carrier's operations.Becauseboth of them would not like to take the risk of
being perceivedas in the sameimage asKAL. In November 1999the Korean governmentimposeda
ban on KAL launching new international flights following a spateof seriousaccidentsthat led to the
suspensionof a number of airline alliance agreements.It raises the question:When KAL's
code-sharepartner Delta Air Lines completedits independentsafety audit, would KAL really ignore
things that clash with traditional ways? In 1999,KAL underwent a safety audit, but KAL's reluctant
acknowledgmentof a leaked internal safety report in 1999 revealedits fundamentalsafety culture
had not been changedat all following the seriesof accidents(Hsu, 1999; Lee, 2000).
79
Figure 2-20 Organisation and Subcultures
Seniority Culture
Culture
Context
Safety
Culture
Servic
Culhue
Others
ORGANISA
Source:Adapted from Smircich, 1983
Various types of organisation
Westrum (1993,1995) provides a view regarding the basic organisation communication
styles so that an aviation organisation can learn from the styles of management of its
employees. He examines three patterns of information used by aviation organisations:
pathological, bureaucratic, and generative (see Table 2-4). This concept has been
adapted widely in the aviation world.
These categories reflect the communication patterns between employees and upper
management and support employees after problems are reported. The "pathological"
style is a highly conflicted organisation. "bureaucratic" organisations process

information as a routine job and fail to handle changes. A"generative" style can help to
create a highly reliable organisation.
Different types of organisation will foster different corporate cultures. Ideally, the
generativetypes of organisation are favoured and encouragedbecausethe generative

culture can help and support employees in learning from mistakes, rather than
apportioning blame and taking punitive action.
80
'Fable 2-4 Basic Organisation Communication Styles
Pathological Bureaucratic Generative
Information is personal power Information is routine formation is seen as

key source
Responsibility is shirked ponsibility is Responsibility is shared

partmented
(Messengers are shot Messengers are listened to Messengers are trained

if they arrive
Bridging is discouraged Bridging is tolerated idging is rewarded
Failure is punished or covered isation is just and fair ilure leads to inquiry or
ew ideas are actively ideas present problems ew ideas are welcome
Source: Westrum, 1995
2.6.3.4 Safety culture
As presented previously, the characteristics of organisational culture can affect the way
in which an organisation manages risk. Adams and Ingersoll (1989) comment that at the
organisational level of analysis it may often be more appropriate to discuss culture in
terms of subcultures. Hence, combined with the fact that different types of
organisations demonstrate different safety records, the concept of organisational culture

to the management of risk has led to a significant amount of discussion within the
management and literature of the safety culture concept.
In particular, safety culture was discussed to a large degree following the Chernobyl
nuclear accident (1986). The Chernobyl accident was identified as being caused by the
organisations responsible for the Chernobyl Nuclear Power Plant, which lacked a
"safety culture" and resulted in an inability to remedy design weaknesses despite these
being known about before the accident. The human errors and violations of procedures
were interpreted as evidence of a poor safety culture.
In the airline industry however, it is not easy to measure how good a good safety culture
is or how had is a had safety culture. If examining the definition of safety culture, there
81
are various kinds of interpretation and explanations. Table 2-5 lists some definitions of
safety culture. Although there is much discussion around the concept, there is little
common understanding of what constitutes an exact definition of safety culture

(McDonald and Ryan, 1992). Nonetheless, it is generally agreed that they contain
similar underlying elements of a safety culture: beliefs, attitudes, norms, and values.
Some definitions also encompass the tangible manifestations of culture: priorities,
behaviours and practices. Yet the most important of all, as indicated by Hayward (1997),
is that the establishment of an appropriate safety culture is the recognition that human
error is unavoidable and that it is the responsibility of a mature organisation to
effectively manage that error.
Table 2-5 Definitions of Safety Culture
Researchers Definition
Cox and Cox (1991) Safety cultures reflect the attitudes, beliefs, perceptions, and
values that employees share related to safety.
................................. ..........................
Pidgeon (1991) The set of beliefs, norms, attitudes, roles, and social and
technical practices that are concerned with minimising the
exposure of employees, managers, customs and members of
the public to conditions considered dangerous or injurious.
Geller (1994) Everyone feels responsible for safety and pursues it on a daily
basis in a total safety culture.
Lee (1996) The safety culture of an organisation is the product of

individual and group values, attitudes, perceptions,
competencies, and patterns of behaviour that determine the
commitment to and the style and proficiency of an
organisation's health and safety management.
Merritt and A safety culture is more than a group of individuals

Heimreich enacting a set of safety guidelines. It is a group of people
(1996) guided in their behaviour by their joint belief in the
importance of safety, and their shared understanding that
every member willingly upholds the group's satctv norms and
will support other members to that common cud.
82
2.6.4 Contemporary Safety Programmes with Proactive Concepts in the

Airline Industry
Abeyratne (1998) points out that regulation of a high tech industry such as aviation
must move towards a proactive rather than a reactive approach to safety. Now there is a
consensus among the international aviation community that human factors must
progress beyond the "knob and dials" approach of ergonomics, beyond training and
beyond the post-mortem application of human factors knowledge in accident
investigation. A good safety programme would most likely have identified safety
hazards within the training and checking systems, and resolved deficient aviation safety
defences in order to evaluate and improve the integrity of safety defences, as well as the
safety measures before an accident or incident. Efforts have already been made in this
field recently and some programmes were developed as a result. Within the modern
airline industry, there are five programmes with proactive concepts to help to assessthe
airline's safety. They are: (1) British Airways MESH Programme (Reason, 1994), (2)
Boeing's Safety Programme Model (Boeing), (3) BASI-INDICATE Programme
(Edkins, 1998), (4) PERS (FAA, 1997b), and (5) LOSA (Line Operation Audit). In
particular, MESH and PERS are applied in engineering maintenance aspects (see
Appendix H for the introduction of these programmes in details).
2.6.4.1 Managing Engineering Safety Health (MESH)
Objectives
Managing Engineering Safety Health (MESH) is a programme created for British
Airways Engineering Company in 1992 by a team lead by Professor James Reason

from the University of Manchester. It is a set of diagnostic instruments for making
visible, within a particular engineering location, the situational and organisational

factors most likely to contribute to human factors problems (Reason, 1994; Mäurino et
al, 1995).
Designedto assessthe safety climate of an organisation,the measuresof MESH give an

indication of the system'sstateof safety (and quality), both at the local workplace level
83
in
and general. It is a system of measuring a number of local and organisational factors
and the interplay between them. The local factor assessments are made at weekly
intervals by a randomly selected proportion of the workplace in each of a variety of
workplaces (i. e. operational hangars, majors overhaul hangars, workshops, etc.). The
organisational factors are assessedat three-monthly intervals by technical management

in each location because these are the people best placed to judge the impact of
"upstream" organisational factors upon the reliability of their various workplaces

(Reason, 1995a).
Advantages
Reason (1995a) suggests that by identifying factors in need of improvement and
tracking the changes over time, MESH enables the maintenance of adequate safety
health, comparable to a long-term fitness programme, in which the focus of remedial
efforts switches from dimension to dimension as previously salient factors improve and
new ones come into prominence. The advantages of MESH can be thus summarised as
follows:
Staffs are all involved in safety.

Managerscan prioritise remedial actionsand check upon their impact.
Direct safety resourceswhere they are most needed.
Encouragebetter communicationsbetweenmanagementand staff.
Limitations
Although the first impression of MESH appears to be that it is easy, the implementation
of this tool may lead to some difficulties. Caution must be expressed when interpreting
the implied meaning of results when employing a single rating scale method of
investigation. Due to personal tendencies, some people tend to use only the extreme of
the scale, while others tend to use only the middle area. In addition, the MESH users
have no access to any overview of the global structure and the items to be rated are
displayed on separate rating pages.
Meanwhile, the factors rated in MESH have very general meanings and can be
84
interpreted and rated according to different interpretations. Consequently,the results
obtained by MESH are not sufficiently detailed for suggesting proper corrective
actions.
MESH has never been applied by low capacity operators, and would be beyond the
resources of smaller operators. Although it has also been implemented by Singapore

Airlines Engineering Company, this programme is not adopted widely by other airlines
and has not achieved the significant improvements in safety performance that were
originally expected. A number of improvements and modifications are currently being
made as a result.
2.6.4.2 The Boeing Safety Programme
Quoted from CASA (1998), the Boeing Safety Programme is a comprehensive
programme designed not only to introduce the reason for having effective safety
programmes, but also to introduce the tools for running them. The programme is
by
presented Boeing as a two-day training course and is useful for ramp, maintenance
and flight operations.
Approaching from an organisational standpoint, it covers management's involvement
and support for the programme through the development of integrated mission and
policy statements. The model describes how safety functions can be linked within an
operation in different ways.
The safety process begins with some of the traditional safety programme elements
including information gathering, investigation, evaluation, and change. The
programme also includes sections on the products of a safety programme, including

newsletters,bulletins and other forms of communication within the organisation, and
resourcesavailable to the Safety Officer/Manager. The section on resourcesidentifies
some organisationsthat are dedicatedto safety, various training institutions that offer
safety related training.
85
The Boeing manual provides many useful checklists. It also gives anecdotalcoverage
of incidents to show how pre-emptive actions could have preventedthe occurrences.

However, it was developed for larger airline operations and much of it would probably
be superfluous to the smaller passenger-carrying operator (CASA, 1998).
2.6.4.3 Identifying Needed Defences in the Civil Aviation Transport

Environment (INDICATE)
Obiectives
INDICATE is a safety programme that has been developed in consultation with the
Australian regional airline industry for proactive purposes. The name is based on the
underlying purpose of the programme which is to identify and resolve deficient aviation
safety defences (Edkins, 1998). It provides a formal communication channel for aircraft
operators to regulations, policies and standards to the Bureau of Air Safety
Investigation (BASI) in Australia. It is also known as BASI-INDICATE.
According to Edkins (1998), safety defencesare barriers or safeguardsput in place to
protect a systemfrom both human andtechnical failure. He presentsa modified version

(Figure 2-21) of Reason'smodel of organisationalaccidentcausation(seeFigure 2-18,
page 71). Edkins argues that each of the organisation, workplace and person/team
componentsof Reason'smodel is difficult to identify before an accidentbecauselatent
failures are usually unforeseeable,workplace factors are dynamic, and errors or
violations are unpredictable.This model implies that the integrity of safety defences
canbe more accuratelydeterminedasthey are more tangible and thus more measurable
componentswithin a system.The INDICATE programmehas thereforebeen designed
to regularly evaluateairline safety defencesso that the potential risk of an accidentcan
be minimised.
Evaluation criteria
There aremany potential measuresof airline safetyperformance,including the absolute
number of fatal and non-fatal accidents;fatal and total accidentsper million departures;
passengerfatalities per million passengersor per million miles, etc. Regardlessof
86
which measure is used, it is important that it is examined regularly if a safety
management programme is to be effective in improving safety performance.

Nevertheless, accidents are so rare in airline operations that they cannot be used as a
statistically reliable index of safety performance.
Figure 2-21 Proactive Defence Evaluation Model

Defences
Engineered safety devices

Organisation Workplace Person/team
Policy, standards, controls
Management Error and
10
decisionand violation Errors and Procedures, Introductions, CCIDENT
organisational producing 10 violations supervision
process 10 conditions 10
Trainin briefin drills
Personal protective
Difficult to identify pre accidents equipment
R 1f
Proactively identifiable
Source:Edkins, 1998
As such, the INDICATE programme was evaluated based upon the following five
safety performancecriteria (Edkins, 1998; BASI, 1998).These criteria were:
1. Airline safety culture

2. Airline staff risk perception of aviation safety hazards
3. Staff willingness to report safety hazards
4. Action taken on identified safety hazards
5. Staff commentsabout safety management
A trail with an Australian regional airline revealed that there was a clear difference
betweentwo experimentalgroups, and the programmehad had a positive influence on
the airline's safety performance.
Benefits
In Australia, it is a legal requirement to report air safety incidents via an Aviation
Safety Incident Report (ASIR). However, there is a recognised problem of
under-reporting, which in part stems from a lack of awareness about what should be
87
reported despite this mandatory requirement.
The programme provides a simple and structured process to encourage staff to report
safety hazards and deficiencies within their work area. The safety information database
and software allows management to address all safety-related concerns. Furthermore,
senior management regularly meet with safety staff to determine what to do about
identified hazards. It is clear that consistent communication of safety-related
information within an organisation is crucial for improving staff attitudes towards
safety.
Edkins (1998) explains that the results of the INDICATE trial suggest that measuring
safety culture provides a useful method for monitoring changes in company safety
performance and may assist in identifying elements of a safety management

programme that require improvement, such as a hazard reporting system.
Most importantly of all, the evaluation of the INDICATE programme illustrates that the
greatest source of variance is not necessarily aircraft equipment or the category of
operation, but the real cost from the safety culture of organisations within the aviation
system. The benefits from implementing such initiatives will ultimately help to
improve operational safety and, in some cases,reduce operating costs.
2.6.4.4 Proactive Error Reduction System (PERS)
Objectives
The PERS programme is funded by the Federal Aviation Administration's Office of

Aviation Medicine and devisedby Dr. Colin Drury at the University at Buffalo (UB),
the StateUniversity of New York (IFA, 1998).The FAA-funded project beganin 1989
following a Congressionalhearing prompted by an incident in which 18 feet of roof
pulled away from an Aloha Airline jet as it was flying over the Pacific Ocean.A flight
attendantwas suckedfrom the plane and 61 passengers
were injured.
For years,Drury and his co-workers have analysederrors by airline workers in detail.
88
They are using this knowledge to build practical tools that allow users to arrive quickly
at solutions to errors made by airline workers. PERS is an error management system
which can be used by non-experts in human factors because it is based on a
human-factors approach to solving errors. The idea behind this is that one should not
just determine the immediate cause of the error, but examine all the things that lead up
to it. As such, this programme, as shown in Figure 2-22, is structured to use the
repeating patterns found in incident data in order to help airlines move from the
recognition of human factors as an issue to practical human factors solutions.
Figure 2-22 The PERS Structure
Safety I Critical
Audit incidents
Potential Known
Error )roblems problems
Error
assessment report
(e.g. MESH) Hazard patterns
(e.g. MEDA)
ng
Criteria Solutions Database

selection
Similar Select Error control

1
solutions solutions management
Source:FAA, 1997 b
Advantages
There are four distinct functions within the programme:

1. A comprehensiveerror managementsystemwhich can combine many existing
data collection and analysis systems(e.g. MESH, MEDA). In other words, it
facilitates the importing of data from other systems such as other audits and
MESH, and links databasesof maintenance errors with databasesof known
solutions.
89
2. PERS facilitates the identification of error-prone situations by using
- error audit of specific tasks

like
assessment MESH.
- error
3. PERS supports error management strategies, identifying situations for
design/procedurechanges.
4. PERS facilitates error reporting by employing
error reporting modules like MEDA (an error-reporting systemwith interfaces

-
to MEDA)
- critical incident reporting modules.
PERS provides a way for airline personnel to analyse an error or potential error, to
discover why it occurred, and then to see how they might go about changing systems,
equipment or work patterns to prevent future errors. PERS not only tells airline workers
what to do if an error occurs, but it also tells them what to do even if what have occurred
are not actual errors, but error-prone situations.
2.6.4.5 Line Operation SafetyAudit (LOSA)
Objectives
LOSA was developedby the teamof ProfessorR. Heimreich at the University of Texas,
USA. Under LOSA, flaws in human performanceandprevalenceof error are taken for
grantedand the objective becomesimproving the context within which humanperform.

LOSA ultimately aims to introduce a buffer zoneor time delaybetweenan error and the
point at which its become

consequences a threat to safety.The better the buffer or the
longer the time delay, the stronger the tolerance of the operational context to the
negativeconsequencesof human error (Maurino, 2001).
LOSA are programmesthat use expert observersto collect data about crew behaviours
and situational factors on normal flights. Observationsgeneratea narrative of the flight

classified by phase,and theseare conductedunder strict nojeopardy conditions, which
mean that no crew are at risk for observedactions. Observerscode observedthreats to
safety and how they are addressed, errors and their management, and specific
90
behaviours that have been associatedwith accidents and incidents. If explained by
using Figure 2-21 in INDICATE programme, it is the errors of person/team that LOSA
would like to collect.
The critical difference betweena LOSA flight and a line check is LOSA's guaranteeof
anonymity for the crew. Data are entered into a de-identified databaseand no crew
actions are reported to managementor the regulatory agency. In LOSA, error is
classified as deviation from organisational or crew exceptions or intentions. Errors
committed by the flight crew are describedand coded along with actions taken to deal
with the consequencesof the errors. Table 2-6 lists the various errors and remedial
in
strategiesemployed LOSA.
Advantages of LOSA
Data from LOSA provides a picture of system operations that can guide organisational
strategy in safety, operations, and training. Heimreich (2001) points out that a particular
strength of LOSA is that it captures exemplary as well as deficient performance, which

provides airlines with the areas in which they excel as well as those in need of
improvement and used as models for training. As such, data collected in LOSA are
proactive and can be used immediately to prevent adverse events.
The other strength is that a database is being developed that allows organisations to
compare their results with other airlines. Such comparisons help in interpreting the
significance of the number of procedural and decision errors observed and the
effectiveness of threat and error counter-measures. The data allow management to
prioritise safety initiatives and training departments can use the information to develop
targeted training.
Meanwhile, the informative aspectof LOSA datais the ability to link threat recognition
and error management

with the specific behavioural markersthat from the of
core CRM.
Using LOSA, a model incorporating the Swiss Cheesemodel has been developed
(Heimreich, 1999). It recognisesboth overt and latent threats,and how they fit into the
managementof error and undesiredstates.
-91
Table 2- 6 Various Error Types and Remedial Strategies
Varieties of error Remedial strategies
Crew intending to follow a procedure but doing it Suggests poor workload

incorrectly. Include the usual classification of slips, management or may be a
Procedural
lapses and mistakes such as incorrect data entries or reflection of inadequate
errors
flying the wrong headings. pioeedwe;.
Intention is correct but the execution flawed.
These involved failures in the transfer of information, ( ' onuIninications errors may
Communication including mis-statements, misunderstandings and reflect a need for more focus
errors omissions. Examples include incorrect read back to on CRM, especially
ATC or communicating wrong course to the other pilot. interpersonal communication
issues.
When crews choose to follow a course of action that It suggests a need for further
unnecessarily increases risk to the flight in a situation CRM concentration on expert
not governed by formal procedures, this action is decision-makine and risk
Decision errors
classified as a decision error. E. g., crews may choose assessment.
not to deviate around bad weather on their flight path,
resulting in an encounter with turbulence.
This classification is applied to situations where a cre%ý a need to tighten

member lacks the knowledge or stick-and-rudder skill standards for qualification and
Proficiency necessary to perform a task. E.g., like extreme evaluation.
errors manoeuvres on approach, choosing to fly into adverse
weather. A number of observed proficiency errors
involve lack of knowledge of flight deck automation.
When crews obviously and intentionally violate

company or regulatory requirements. Failing to abort
Intentional
an unstable approach as required by company
non-compliance
procedures would fit into this classification.
E.g., omitting required briefing or checklist
Source: ,Adapted th)nl I iclmreich, 1999
These markers emerge very clearly in observer ratings of the actions taken by effective
crews. Those who deal proactively with threat and error exhibit the following
behaviours:
92
o active captain leadership

Q briefing known threats
o asking questions,speakingup
Q decisionsmaking and reviewed
o operational plans clearly communicated
Q preparing/planning for threats
Q distributing workload and tasks
o vigilance through monitoring and challenging
Although the challenge for the implementation of analysis of normal operations is to
overcome the obstacles presented by a blame-oriented industry, LOSA has been
appreciated by various airlines which are willing to make the investment in conducting
the necessary observations and analyses. In addition, the value of LOSA has also been
recognised by ICAO and has gained support to conduct a "LOSA Week" to promote
this programme to interested countries.
2.7 The Proactive vs. Retroactive Approaches to Safety
In the twentieth century, the value of accident investigations in identifying causes and
initiating corrective action to prevent future errors has been greatly appreciated because
the knowledge acquired through accident investigations has paved the way for
improvements in air travel. However, the use of after-the-fact measures as a trigger to
initiate safety efforts has been a very reactive approach to airline safety.
Studies have shown that most safety systems. are reactive (Johnson 1994; Earnest,
1997). Johnson (1994) reveals the result of a survey, in which 83 percent of respondents
indicated that "safety programmes are reactive, isolated within organisations and
preoccupied with quick fixes and putting out fires. " Earnest (1997) listed nine select
criteria that can help to determine whether a safety culture is primarily proactive or
reactive. Although these comparisons are focused in the areas of safety and health,
some of them can be adopted and expanded the view on the airline industry. These
characteristics are therefore detailed in the following paragraphs and listed in Table 2-7.
93
1. Incident/Accidents investigation:
In a retroactive system, the process of incident investigation tends to follow the
"Domino theory" (Heinrich, Petersen & Roos, 1980), which typically focuses on a
concept: identify the unsafe act, remove the hazard and prevent the recurrence (See
Appendix I). As such, incident/accident investigators focus solely on accident
symptoms rather than root causes in the organisation, which may become the source of
future accidents.
In a proactive system, the occurrence of incidents/accidents has been recognised to be a
system problem, as with Reason's model, which traces the causes of accidents back to
the management system, and contains active failures and latent failures. To detect the
latent condition and put in place defence against it is beyond the single-cause concept to
take proactive steps to change the system.
2. Safety performance evaluation and measures
In a retroactive system, Earnest (1997) statedthat safety evaluation is based on the
absenceof injury, so the managementtends to rely on after-fact-measuresto provide

guidance and assume that safety is achieved. With a system that judges safety
performance based solely on the absenceof incidents/accidents,it may be wrongly
assumed that the company is doing the right things as long as there is no occurrenceof
injuries.
With a proactive safety system, evaluation is based on the improvement of safety
systems,and includes the function of management.Safety is evaluatedand monitored

in the long term in order to enablestrategicplanning.
3. Safety practice and performance

Within a retroactive safety system, safety practices are developed in response to
just
accidents/incidents, like a band-aid applications(Merritt A. and Heimreich, 1996).
For example,GPWS or EGPWS are designedto prevent CFIT, becauseCFIT hasbeen
in
the major cause of aircraft accidents the last decade.There are also other risk
management tools which are designedfor different causes.
94
By contrast,there are clear written safety practices in a proactive safety system,which

have been developedbasedon a thorough evaluation of hazardsin order to implement
safety defencesbefore incidents/accidentsoccur. Not only risk managementtools are

usedto prevent identified risks (hazards),but also the latent conditions and safeguards
that can be identified by using tools like MESH or INDICATE.
4. Safety goal and vision

One of the obvious differences between retroactive and proactive approaches to safety
is the goal for safety. In the former, safety goals are based only on the reduction of
injury or incident/accident at an organisational level. In the latter, attention is placed on
the strategies for achieving these goals at both departmental and organisational levels,
or even at group level. There are also periodic reviews and different milestone
expectations from the management in order to ensure that the goals are achieved, and to
reward employees. Safety goals are aligned with management plans to prevent
recurrence, which results in substantial changes.
5. Safety programme and training

In a retroactive safety system,programmesand training go little beyond the mandated
regulation (Earnest, 1997) due to cost consideration. Take EGPWS for example;
although it is a fact that this equipment can effectively prevent CFIT from occurring,
not all airlines are willing to install EGPWS unless it is a regulatory requirement.
Training is another example,which is in some airlines, cabin crew training takes two
months and in others it takesthree months.
Conversely,in order to ensurethe bestprogrammesand performance,a proactive safety
system needs to go beyond a purely mandatory level. Safety should have a higher
priority than cost. The quality of safety programmes reflects how important the
managementthink it is, and the willingness for the managementto allocate resources.
6. Employee recognition and attitudes

To improve the safety system, employees must understand the structure, content and
goals of the system, as well as how it functions. They should also understandhow
95
Chapter 2 Literutwr Rev: e t c\ Empi, n aI Irin/Ysis
.
performance is evaluated in the system. This concept is not fostered in a retroactive
system, so employees just adopt the attitude that "good safety perf'ormance is zero
accidents. " Since safety culture is weak, safety improvement just ends up with related
personnel being re-trained in the aftermath of incidents or accidents.
Recognition in a proactive safety system is based on improving system safety, which
helps to ensure the continuous improvement in the long-tenn.. A strong safety culture is
fostered in both the company and its safety system. As such, safety health is
consequently achieved.
Table 2-7 Comparison of Different Characteristics in the
Proactive and Retroactive systems
Retroactive Proactive
F-OCJSe(:
Focused on causal factors and or f [)U c 1LJses and
Incidents/accidents
unsafe acts, active failures (e. g. latent mistakes in management
investigation
Domino theory) (e q REO1snn's model)
After-the-fact
performance:
Safety performance on the
measure, evaluation based on
measure and improvement of safety system
the absence of injuries, and
evaluation and long-term strategies
mistakes in the short-term
Identifies safety defences before

Tends to respond to
Safety practice and any latent cui tli'iO rls
incident/accident and injury
performance are forme(!
avoidance, a band-aid solution
Aligned with company qoA and

Based solely on eliminate errors
Safety goal and have no conflict with production
and accident reduction
vision goals
Safety programme Only meets regulatory minimal Beyond regulatory requirements

and training requirement on a voluntary basis
Safety recognition Good safety performance is zero Emphasised on continuous
and attitudes accidents - short term improvement in the long tern)
Source: Adapted fruni Farnest (1997)
96
CHAPTER 3
Research Methodologies
"Science came into existence through the opposite belief.

It is the same with mathematics, which would certainly not
have come into existence if one had known from the beginning that
there was in nature no exactly straight line,
no real circle, no absolute magnitude. "
Friedrich Nietzsche
3.0 Introduction
In Chapter 2, the concept of an airline safety management system (SMS) has been
investigated (Research objective 1). The literature and practical industry programmes
demonstrate that both retroactive and proactive approaches to safety are used within
airline SMS. The difference between retroactive and proactive safety management is
in the treatment of the contributing causes of accidents and their underlying factors to
prevent accidents from occurring again. In particular, the cultural and organisational
dimensions have been illustrated as two main components in the proactive approach
to safety (Research objective 2).
However, it is evident from the preceding chapter that the componentsof a proactive
approachto safety are an important but as yet, relatively undefined task in the airline
industry. Therefore, the industry needs a model to conceptualise the `safety
in
mechanism' proactive safety management(Researchobjective 3), which contains
the cultural and organisational consideration, identified in Chapter 2, and
organisationally influential factors (Researchobjective 4). Chapter 2 has completed

the study to meet the research objective One & Two. According to the findings,
97
Chapter3 Research
Methodologies
Chapter 3 follows to present the research methods to solve the found problem in order
to achieve research objective Three, Four and Five ultimately. As such, the
investigation requires a fully explored method.
This chapter introduces the methodologies used in this thesis, which are exploratory in
focus and qualitative and quantitative in nature. It aims to establish a framework for
the study of proactive approach to safety and for the evolution of airline safety
management system by the development of a `safety mechanism model'. Interviews
were firstly undertaken to gain the empirical knowledge and experiences necessary to
develop a safety mechanism for proactive safety management with the combined
findings from the literature review. The safety mechanism model, as well as the
influential factors, is subsequently developed. Meanwhile, the use of a retrospective
case study and the safety survey aim to provide the reader with an appreciation of the
safety mechanism model in relation to how the safety mechanism has been tested, and
how the organisational factors affect the airline safety management system.
For the purpose of better illustrating the methodologies mentioned in this chapter,
Figure 3-1 demonstratesthe structureof methodologiesadoptedin the thesis.
98
C haptt'r 3 Rt'st ar h llt'tlwilrrlogic's
.
Chapter 3 follows to present the research methods to soli e the I und problem in order
to achieve research objective Three, Four and Five ultimately. As such, the
investigation requires a fully explored method.
This chapter introduces the methodologies used in this thesis, %%

hick are exploratory in
focus and qualitative and quantitative in nature. It aims to establish a framework for
the study of proactive approach to safety and for the evolution of airline safety
management system by the development of a `safety mechanism model'. Interviews
were firstly undertaken to gain the empirical knowledge and experiences necessary to
develop a safety mechanism for proactive safety management vvith the combined
findings from the literature review. The safety mechanism i node!, as well as the
influential factors, is subsequently developed. Meanwhile, the use of a ! eetrospc'ctive
case study and the safety survey aim to provide the reader with an appreciation of the
safety mechanism model in relation to how the safety mechanism has been tested, and
how the organisational factors affect the airline safety management system.
For the purpose of better illustrating the methodologies mentioned in this chapter,
Figure 3-1 demonstrates the structure of methodologies adopted in the thesis.
98
Chapter 3 Research Methodologies
Figure 3-1 The Structure of Used Research Methods
Chapter 2
Literature Review
Chapter 4
('liaptcr 4
1. Define proactive safety mechanism model
2. Influential factors proposed
3. Making hypotheses
a
Chapter 5
Retrospective Case Study
Test of the
model
1
Chapter 5
Safety Survey Study
99
(7r(Ju( r i' Kcc, ',n h i(l&'t/U) Mlo li,
. ,
3.1 Interviews
The method chosen for this part of the study is a scrics Of, ctni ýtrurtuicd qualitative
interviews conducted to include various airline safety people. The use of
semi-structured interviews meant that the topics discussed were grounded on Chapter
2 but were flexible enough to incorporate other issues brought by the participants. In
this way, the interviews were designed to build upon literature review and empirical
findings in the previous part of this study. Following sections will present the details
of how the airlines and interviewees are selected as well as the interview format.
3.1.1 The Selection of Airlines
In order to gain a varied perspective on how safety is managed, five safety managers
in different airlines were interviewed19, based upon the airline's willingness, size,
scope and nationality, and the author's time plan and convenience. The diversity in the
size and nationality of airlines was thought to be critical to ensure that a variety of
organisational and safety cultures, and safety management practices were presented
through the interviews. A detailed description of the companies is presented in Table
3-1.
Table 3-1 The Profile of Selected airlines
:
1
A BCDE
Nationality Western Eastern Eastern Western jEastern
Size Large Medium Medium Medium (Small

International International International International Domestic
Service Regional Regional Regional ' Regional
Domestic Domestic Domestic
Fleet I >100 1100>, <50 1100>, <50 50> i 20>

Note: All of them run passenger, charter, and cargo services.
io Two civil aviation authorities were interviewed as well (one western, one eastern). The ('AA
provided a view from the regulatory angle. Although these opinions were not included in their
entirety in the interview results, they played an important part as a reference vvhen completing the
interviews with airlines, and when summarisini the final re, ult, of this research.
100
Chapter 3 ResearchMethodologies
3.1.2 The Selection of Interviewees
As the set of interviews was intended to provide a broad view of the safety activities
within the airline industries, interviewees were approached on primary criteria. People
were sought who were felt to have sufficient experience and knowledge to be able to
provide an overview of his or her airline, as well as the airline industry as a whole. In
this way, the opportunity sample therefore consisted of people who are in charge of a
safety department or mostly related to safety activities of the organisation. Their titles
are varied according to the different companies, such as Safety Manager, Director of
Safety or Safety, and Security VP.
3.1.3 Interview Format
A number of topic areaswere identified from the previous analysesasbeing critical to

the managementof safety system and these formed the basis of the interview format.
The topics were not only focused on proactive safety managementor organisational
safety culture as more practical discussionswere sought (semi-structured).As reactive

safety practices and proactive safety culture are inextricably linked, the topics that
formed the basis for the discussions ranged from the more tangible and easily
describablesafety managementpractices to the less easily quantified issuesof safety
culture. They included:
+ Safety managementsystem
+ Risk management programme
+ Safety audit
+ Organisationalstructure
+ Culture and safety
+ The role of regulator
+ Improvement
+ Others
101
Chapter3 Research
Methodologies
Interview followed a semi-structured format. The detailed questions can be found in
Appendix J. Each interview lasted from one to one and half hours approximately. The
majority of the interviews were conducted in the interviewee's place of business.

Where this was not possible, the interview was conducted by telephone. Given the
requirement from the participants, the names of airlines are omitted for reasons of
confidentiality.
The findings from these interviews, combined with the results of literature review,
were then used to develop a safety mechanism model of how a proactive safety
approach develops and evolves within the airline industry. The interview findings are
discussed in Chapter 4.
It should also be kept in mind that the interviews were semi-structured in format and
therefore these formed only the basis of the interviews. In addition, although
interviewees were willing to take part in the interviews, there might be some inherent
difficulties in obtaining complete information from the interviewees due to the
policies of individual companies.
3.2 Model Development
3.2.1 Purpose of the Model
As discussed in previous chapter, a proactive approach to safety is called for to

improve airline safety management system. This calls stem largely from the
shortcomings in existing retroactive safety programmes, which focus on the

identification of the accident causes and the human factors which are thought to
constitute effective safety management.Some programmes have been designed to

help to maintain safety proactively as introduced previously but few of them have had
significant success.Therefore, apart from the method of interview which provides the
existing knowledge of is
what thought to contribute to an effective proactive safety (a
safety mechanism), what is further required is a model which builds upon the safety
mechanism by adding an increasedknowledge of the organisational factors which
102
Methodologies
Chapter3 Research
serve to influence a proactive safety mechanism, and which will serve to be the
predictors20of the performanceof airline safety managementsystem.
The model focuseson the hypothesisedorganisational factors which impact upon the
developmentof a safety mechanism.Many such factors have been identified from the
analyses in the chapter of literature review and will be identified in the interview.
Building upon the results of this work, a hypothesisedmodel of the developmentof a
proactive safety mechanismwill be presented.
In short, the purpose of the model is to provide a greater understanding of the

important concept of a proactive safety mechanism and its
complex and potentially
influential factors for the airline industry. In addition to this, this model intendsto:
Identify the layers which contribute to the development of a proactive safety

mechanism
Identify the factorswhich exert an influence on the safety mechanism
Identify the organisational factors which are the predictors of the safety
mechanismand safety performance
In this way, the relationship between airline safety managementsystem and safety
mechanismand safety performance should become clear.
20 In statistical term, `Predictor' meansa possible factor (input) to forecast the consequenceof
something (output). For example, in Multiple Regression procedures, it will estimatea linear
equation of the form:
Y=a+b, *Xi+b2*X2+... +bP*Xp
In this equation,the regressioncoefficients (or B coefficients) representthe contributions of each

independentvariable (Xs) to the prediction of the dependentvariable (Y). Therefore, Xs are the
predictors of Y.
103
Chapter3 Research
Methodologies
To this end, four general hypotheses will be explored and tested. The term hypothesis
as used here should be understood in the broad sense of the word as these are intended
to outline the aims and objectives of this part of study. These statements should be
considered to be propositions, made from known facts, which form the basis for this
investigation, rather than hypotheses in the statistical sense, which will be tested and
either accepted or rejected. The outcomes of the investigations with respect to these
hypotheses will be discussed in order to summarise the results of the investigations
which formed the substances of this thesis.
3.3 How to Test the Model
Two methodshave been used to test the model. One is a retrospectivecasestudy and
the other is the use of a safety survey.
3.3.1 A Retroactive Case Study
The use of an "organisational accident" to demonstrate the failure of socio-technical
systems is a technique employed extensively by Reason in developing and making a

case for the Resident Pathogen Model. Working backwards from the accident site
provides what Reason (1995b) calls a pathogen trail. This case study approach is
based on Reason's concept, but it will focus mainly on the commercial airline industry.
The use of the after-the-fact approach is not gifted with the spirit of "proactive"
concept. However, at this stage, by dissecting the pathogen trail, the bottom-up
metaphor helps to illustrate how the model explains the safety mechanism, which
develops under the various factors identified in next chapter.
Selection Criteria
This case,the crashof Air Ontario Flight 1363,was selectedin accordancewith the
following criteria:
104
Firstly the case must be informative, i. e. sufficient and available information is
required. In the case of accident, formal accident investigation reports can help to
provide a sufficiently lengthy pathogen trail to facilitate the analysis. Usually the
more severe cases are the accidents/incidents, as the better known they are, the more
they are subject to intense scrutiny. The case of Air Ontario is well qualified to serve
as a subject for informative discussion in this section. However, it is worth noting that
this does not mean that only serious or severe accidents/incidents are suitable to
illustrate the model.
By using the case study, it aims to provide a useful illustration of the myriad forces,
which conspire to define a system's safety mechanism. Given the complexity of the
case, it is hoped that the case study will serve to provide the reader with a greater
appreciationof the explanatorypower of the in
model relation to the in
way which the
safety mechanismdeveloped,and how the safety managementsystemhas evolved.
3.3.2 Safety Survey
What is required next is an in-depth examination of the structureunderlying the model.

As such, the following methodology will use quantitative studies aimed at looking
more closely at the relationships between the forces hypothesisedto be at work under
the domains of the model.
The study was conducted using a three-stage process of, questionnaire development,
questionnaire piloting and questionnaire distribution. In other words, the survey was
developed based on the safety mechanism model. The pilot questionnaire allowed the
instrument to be tested and modified to suit the target audience for the final
distribution.
3.3.2.1 Stage 1: Questionnaire development
1. The concept of a safety survey
The term "survey" meansto make a detailed investigation of the behaviour, opinions,
105
etc, of a group of people (Collins, 1995). There are three types of survey, based on the
degree of their structure or formality (Hague, 1993). These are structured,
semi-structuredand unstructuredsurveys.
In structuredsurveys,the wording and sequenceof most of the questionsare fixed and
are identical for eachrespondent,

with eachquestion having predefined answers.This
ensures that any differences between responses are attributable to individual

differencesand not to variations in the survey.
The semi-structured survey has a mixture of questions with predefined answers, as
well as those where the respondents are free to answer anything that he/she wishes.
The advantages of this type of survey are the greater flexibility that they offer, and the
opportunity of finding out more in-depth reasons for certain answers.
With unstructuredsurveysthere are no pre-specifiedquestions.Usually a checklist of
questions is used to assist the respondentsin describing their experiences,opinions

and attitudes.
A perception survey, or called a culture or climate survey, is used to measure/assess
the respondents' concept or behaviour. Bailey and Petersen (1989) suggest that a
perception survey is a better measure of safety performance and a much better
predictor of safety result, as it can identify the strengths and weaknesses of elements
of a safety system. In other words, a safety survey is essentially used to review the
extent of satisfaction with aviation operations, and to diagnose any problems that may
be apparent or suspected. By assessing safety attitudes, the real safety level of an
organisation can be determined.
Interviews and questionnairesare usually used in safety surveysto determinewhether
a particular facility or operation presents the risk of hazards. The former includes
telephoneor face-to-face interviews, while the latter are self-reported questionnaires,
which may contain "positive and negative measurement"or "attitude scaling" types of
questions,also called Safety Climate Questionnaire(SCQ).
106
Formation and use of the safety survey - Safety Climate Questionnaire (SCO)
The safety survey has gradually increasedin use since the 1980s. Some researchers
(see examplesin the next paragraph)claim that measuring safety climate can indicate
the changes in organisational safety behaviour and would therefore be useful for
evaluating safety programmes. They also argue that any effort to improve safety
should be perceivedas such by employees,and that the only way to measurethis is by
using a safety climate questionnaire(SCQ). SCQ is a structuredsurvey,in which most
of the questionshave predefined answersand there is little latitude for a respondentto
stray beyond them. However, it does have the advantagesof a structured survey
mentionedin previous section.
For example, Zohar (1980), who was the first to develop a safety climate survey, used
it to establish the high agreement in employees' perceptions regarding the safety
climate in their company. The level of this climate is correlated with safety
programme effectiveness, as judged by safety inspectors. Zohar found eight safety
climate dimensions in the resulting responses, they are (1) Importance of safety
training, (2) Management attitudes toward safety, (3) Effects of safe conduct on
promotion, (4) Level of risk at work place, (5) Effects of work place on safety, (6)
Status of safety officer, (7) Effects of safe conduct on social status, and (8) Status of
J
safety committee. Brown and Holmes (1986) found three dimensions by having
Zohar's model validated on American sample. These three retaining factors were: (1)
Employee perception of how concerned management is with their well-being, (2)
Employee perception of how active management is in responding to this concern, (3)
Employee physical risk perception. Dedobbeleer et al. (1990) studied the relationships
between safety climate and organisational factors prevalent in most safety

programmes. They also attempted to validate Brown and Holmes's three factors, but
only found two factors, one of which measured Management's commitment to safety
in terms of management's safety attitudes and practices; the other factor labelled
Workers'involvement in safety..
In 1990s,increasedefforts were made in the discussion of the measurementof safety
climate and its subsequentapplications. Safety climate scales have been developed
107
primarily on the basis of attitude items (e.g. Niskanen, 1994), or based exclusively
upon safety-related perceptions, with both attitudinal and perception items (e.g.
Williamson et al., 1997).
Meanwhile, several questionnaires have been developed in an attempt to determine
the key factors that comprise safety climate. Flin et al. (2000) and Guldenmund (2000)
identified twenty-seven such studies. By 2000, over thirty studies using safety climate
questionnaires have been published (Guldenmund, 2000). It is worth noting that there
is no explicit distinction or discussion between safety culture and safety climate
within these studies, and safety climate is generally taken to comprise a summary of
employee perceptions of a range of safety issues. As such, whether safety culture and
safety climate are the same or they have not been sufficiently defined to identify the
deficiencies will be further investigated in the model.
2. Design of the questionnaire
Budworth (1996) refers to measuring safety climate as taking the "safety temperature"
of an organisation. As such, potential uses for a safety climate questionnaire include:
measuring employee perceptions of management commitment to safety, detecting
areas for safety that require improvement, identifying trends in an organisation's
safety performance and establishing benchmarks for safety climate, emerging from
research on organisations.
According to the characteristicsof the safety mechanismmodel, the concept of the
safety climate survey will be adaptedfor questionnairedevelopmentin this research

as a reference point. In order to obtain an individual's self-assessmentof the
conditions hypothesised in the model, a questionnaire was developed to reflect
concerns at each layer identified in safety mechanism. The main objectives of this
safety mechanism survey are to provide an assessmentof the concernsaffecting the
safety mechanism; to verify its influential factors, and to validate the hypothesised
factor structurein the safetymechanism.
In accordance with the hypothesised organisational factors, this questionnaire is
108
Chapter3 Research
Methodologies
designed to consist of two parts - internal factors and external factors. These items
deal with the concerns of individuals both in the internal and external environment.
They were intended to examine the degree to which the individual felt part of their
organisations or considered organisational concerns when making choices at work.
A statement was constructed for each item (potential concern) so that participants
could be asked to rate the extent to which these aspects of their working environment
were considered when making choices at work. The response format consisted of a six
point scale which ranged from "Strongly disagree" (1) to "Strongly agree" (6), plus
"N/A" (7).
Therefore, these two sections of the questionnaire sought to verify the salience of
various concerns which were identified as being relevant to the development of a
safety mechanism by assessing the frequency with which each was considered in
making day-to-day choices at work.
3. Method of interpreting the questionnaire
In order to determine the underlying dimensions of the organisational safety
mechanism,a principal component of factor analysis, followed by a varimax rotation

was performed on the questionnaire to interpreting the results.
Factor Analysis
Many statistical methods are used to study the relationship between independentand
dependentvariables. Factor analysis is different. It is used to discover the patterns of
relationship between many dependent variables, with the goal of discovering

something about the nature of the independentvariables that affect them, even though
those independent variables are not measured directly. The inferred independent
variables are calledfactors.
In other words, factor analysis is a form of multivariate analysis which is basedon the
assumption that human behaviour is rarely attributable simply to one cause or
109
Chapter3 Research
Methodologies
influence, and that underlying factors, components or elements can be identified and
used to explain complex human behaviour. As such, the purpose of this technique is to
identify theseunderlying factors in terms of their common underlying dimensions.
Principal Components Factors Analysis
As factor analysis is a technique which is able to reduce a system of many scaled
measurements (or variables) down to a small number of factors, it is accordingly a
method of data reduction. By analysing a set of attitude data using this technique, it
may be possible to gain a more coherent comprehension of the human behaviour

being examined. In factor analysis, the factors are not directly observed; rather, they
are defined by a group of variables or items that are components of the abstract
factors.
As mentioned above, factor analysis is often applied to many attitudinal items in a
survey,so it is commonly used on Likert scaleswith at least five categories.A typical

factor analysissuggestsanswersto four major questions(Darlington, 2002)21:
1. How many different factors are neededto explain the pattern of relationships
amongthesevariables?
2. What is the nature of those factors?
3. How well do the hypothesisedfactors explain the observeddata?
4. How much purely random or unique variancedoeseachobservedvariable

include?
21 Darlington, R. B., FactorAnalysis.

http://comp9.psych.cornell.edu/Darlington/factor.htm, 2002.
110
Stepsin conducting a factor analysis
There are four basic stepsin factor analysis:

1. Data collection and generation of the correlation matrix
2. Extraction of initial factor solution
3. Rotation and interpretation

4. Construction of scalesor factor scoresto use in further analyses
This questionnairestudy is going to follow these main stepsto conduct the analysis.
The results are detailed in Chapter4.
4. Target sample for study
The target sampleselectedfor the presentstudy is airline safety managersworld-wide,

including Africa, Asia and Pacific, Middle East, Latin America and the Caribbean,
Europe, and North America. The rationale behind this decision included several
factors to do with the nature of the required data.
Firstly, a cross-airline approachwas taken in order to ensurediversity in the sample;a
general population survey provided a means of this diversity without having to

approach multiple airlines. Secondly, the information which was sought from the
survey was general in nature. The present study dealt with the factors which
influenced an individual's actions, which may affect safety on a daily basis,
necessitatinga relatively in-depth knowledge of the industry. Therefore, the general
population, i. e. airline safety managers world-wide, was deemed to be the most

readily available target audience which could provide the diversity and knowledge
required for the study. The method and means of acquiring this sample will be
discussed.
111
There are around 2000 commercial airlines in the world, including scheduled, charter,
cargo, and helicopter airlines (ATI, 2001)22. However, 80 percent of the revenue of the
total airline industry was generated by only 20 percent of these in the year 2000. As
such, the initial idea was to find out the most profitable top 400 (2000x20 percent)
airlines in the world. Steps taken were as follows:
1. Airline Business (September, 2001), listed the top 200 passenger airlines,
ranked in terms of RPK (Revenue Passenger Kilometre) in 2000.
2. Air Transport World (July, 2001), showed the most powerful airlines by
region.
Table 3-2 Samples and Regions of Questionnaire Administration
Region Africa Asia Latin Europe Middle North Total

& Pacific America & East America
Source
Caribbean
Airline
Business
Sept20011 13 48 22 65 13 39 200
I op 200 in terms
ul ItI'K)
Air
Transport
World 7 39 1 42 3 33 125
July 2001
(World traffic
Statistic 2000)
Total 20 87 23 107 16 72 325
22 Air Transport Intelligence (ATI), Online Database.

http: //www. rati. coni
112
Table 3-2 shows the results of these findings. By putting the results of these two
databases, these are the top 325 airlines in terms of RPKs and traffic in 2000 (a full
list in alphabetical order is shown in Appendix K), and the breakdown result for each
region. As there was no available database to search for the remaining 75 airlines to
make the sample total 400, it was decided to use these 325 airlines as the distribution
sample.
Addressesand contact details for the safety managers of the targeted airlines were
obtained from the ATI database.In cases where the names of safety people were
unavailable,the name of the recipient of the survey packagewas replacedby a generic
title of "Safety Manager".
3.3.2.2 Stage 2: Questionnaire piloting
A pilot study was conducted in order to identify any specific problems with the
questionnaire in terms of understanding the questionnaire items, understanding of the
questionnaire instructions, or other problems with the data obtained (such as
possibility of restriction of range or response bias).
Due to the restraints of geography and time, plus the fact that the factor analysis
requires a large sample to conduct the analysis, it was decided that the pilot survey
would not entail the standardprocedure of carrying out a mailshot to membersof the
target population. Instead the nearestand most convenient personswere chosen as
respondentsfor the pilot test. The questionnairewas completed and criticised by eight
individuals who possessedskills and experiencesin aviation, by personal hand-in or
by email. These respondents worked at a management level in different airlines,
which is one of the limitations of the targetedsample.
113
3.3.2.3 Stage 3: Distribution procedure
The questionnaire, printed in the style of an A4 booklet, was sent by post in a sealed
envelope to the representative (Safety Manager) at each participating company. Each
survey package was contained in an envelope which bore the name of the University
and a label which identified the package as the Safety Research Study. Included
within the survey package was a questionnaire, along with a covering letter from the
author and a freepost envelope. The covering letter was printed on the School of
Engineering letterhead, and was hand addressed and signed by the researcher. It
explained the purpose of this research and ensured the confidentiality and anonymity
of all responses. The pre-addressed freepost envelope enabled the direct return of the
completed information to the author. Both the questionnaire and covering letter are
available at Appendix L.
3.4 Summary
With respect to fulfilling the main objectives of this research, an interview study will
be conducted followed the findings of literature review in order to develop the safety
mechanism model. The case study and questionnaire study are followed to plan and
implement with the aim of testing the hypotheses of the safety mechanism model.
With respect to the hypothesis, the results of the factor analytic study will be
presented, which will serve to group the myriad variables identified within the model
into a number of factors. As such, the hypothesised factor structure in the safety
mechanism can be identified as well as a framework for understanding the factors
which underlie the development of a safety mechanism across regions. The following
chapter will present the result of these studies.
114
CHAPTER 4
Interview Findings and

Model Development
"Many of the ills of organisations stem from imposing an inappropriate

structure on a particular culture, or from expecting a particular culture
to thrive in an appropriate climate. Vines don't grow where the sunshine
doesn't fall in the right proportions with the rain. "
C. B Handy, 1985
4.1 Interview Findings
Aligning with the interview format described in Chapter 3, the findings of the
interviews are discussedas follows. For the purpose of this discussion,these findings
have been collapsed across airlines and organised according to the topics discussed
during the interview.
4.1.1 Safety Management System
The purpose of this section of the interviews was to obtain an idea of how safety is
managed within different airlines and how the safety management system is
conducted.All of the airlines interviewed use the term "Safety Management System"
(SMS) to describetheir systemof safety management.
Although the structure of a SMS might not be the same in every airline, all
interviewees agreed that SMS should be an integrated system, including risk
management tools, associated support, training and communication systems, which
115
Chapter4 InterviewFindingsandModelDevelopment
involves all the departments in the company. According to the interviewees, the
components of a SMS should cover the following:
Clear policy and objectives
Top management commitment
Identification of safety risks
Safetyperformanceevaluation:performanceindicators of safety to result in
enhancedoperational and maintenanceperformance

A well establishedsafety culture
Feedbackto ensurecontinuousimprovement
One airline employee interviewed portrayed the airline's concept of a safety
managementsystem,which is similar to the diagram presentedby Britannia Airways

(see Figure 4-1). Members from Policy Committee and Quality Assurance(different
names are used in different airlines) involved with advising on flight safety. The
managersof these groups freely discuss and communicate inter-departmentalsafety
concerns. The flight safety committee directly report to the president (CEO). The
direct accessto the top president was also emphasisedby other airline interviewees
(seenext section4.1.2).
Meanwhile, both interviewees from westernairlines emphasisedthat a SMS should be

built as a proactive system, which can anticipate potential safety problems, ensuring
they are addressedbefore an incident or accidentoccurs.
116
Chapter 4 Interview Findings and Model Development
Figure 4-1 Airline Safety Management System
CEO
................................
..................................
.................................
..................................
tiplit: $:äfe... öti.....
Planning Finance ~ngr er. jg:::: FjighE:: :::.qi.::. lj g Personnel

äfey... Crew.... .
Quality Assurance Group*
Engineering Flight safety audit

Audit Cockpit crew
Cabin crew
Line Checks
Policy Committee Group
Engineering Training Chief pilot Cabin service Operation
Source:Adapted from Sharples,1996
* Note: Within Quality AssuranceGroup, training audits should be included and applied
to cockpit and cabin crew on a regular basis.
117
4.1.2 Organisational Structure
Some attention needs to be paid to the issue of the safety manager's position within
the organisational structure. With traditional organisational safety approaches, the
safety officer, who is independent of the operational departments, reports on the
company's safety performance to the chief executive in order to make the final
decisions about safety improvement and investment. All the safety managers in the
interviewed airlines have direct access to the top managers and relative independence
compared with other departments in their airlines.
4.1.3 Risk Management and Risk Programmes
The purposeof this section is to explore one of the values of a SMS - risk assessment
and evaluation, in the airline industry.
All the five airlines involved in the interviews were using some form of risk
assessmentprocedures and tools. Some were highly formalised (large and medium
size), some were less so (small size). The managementof the risk assessmentprocess
is often termed "risk management"23.One interviewee put the aim of risk
managementwell: "Our aim is to find a technique that helps us to manage risk,

without having conflicts with our financial goals and fitting in the decision-making
processes.More importantly we want to find a technique that enablesus to spot a
hazardbefore it had manifesteditself."
One interviewee called the process"hazard management"and provided the following

definition of hazard management: "Hazard Management is kind of the risk based
management approach basically. The techniques used in hazard management are

identifying hazardsby risk-control measuresto prevent the occurrenceof hazards."
23 The concept of risk managementhasbeen discussedin Chapter2 from the academicpoint of view,
which shows that a robust SMS can be achievedby practising safety directives and initiatives.
Theseprocessesconstitute quality and risk management.
118
The value of risk management or hazard management is the processes/ procedures to
spot the hazards and then reduce/ eliminate them. The success of this system largely
depends on the quality of safety information gathering process that fit in the system.
As such, in some airlines in the industry, the quality manager replaces the safety
manager in charge of the safety function. Yet this view is not totally agreed by
interviewees. One interviewee noted: "Safety has always been considered implicit in
the quality management; accordingly, many companies regard their quality policy as
their safety policy too. It is really a misunderstanding. " For example, ISO 9000 will
provide a process that assures consistent application, however, standards may be good
or bad!
The following are the detailed results of discussions with interviewees regarding risk
management. Under the safety management system, a common technique is to use
varying techniques to identify risk and then to apply controls to manage the hazard.
This process is assisted by determining a level of severity based on the product of the
seriousness of an event against the frequency of such an event happening. Hazard can
then be managed to the level of "As low as reasonably practical" (ALARP). Figure
4-2 outlines the common risk assessment framework and concepts24in the interviewed
airlines.
24 Pleaserefer to Chapter2 for the description of various risk programmesand analyses.
119
Figure 4-2 The Common Framework of Risk Assessment

in the Interviewed Airlines
Airline safety management system

" Policies and principles, manuals,accountabilities,training, auditing, communication
Accidents/Incidents investigation:
" Investigatestructuralfailure
" Identify human erroneousbehaviour
(humanfactors)
1. Hazards identification ý2
" Incident reporting system

" Identify process / procedural risks
Risk Management
(Hazard Management)
2. Risk analysis
" Evaluate data,trend analysis
Feedback 3. Risk prevention and control

" Auditing,assessment,
risk programmes
Safety performance
of safety services
120
1. Hazard identification:
ºý Identify all undesirable events or circumstances associated with each

hazardwhich could result in harm (from accidents/incidentsinvestigation,
and incident reporting systems)

+ Identify and categorise potential sources of harm, whether process or
proceduralrisks
Apart from identifying and categorising source of harm, the major aim is also to feed
the database with data and parameters. Not only can the database be used for risk
analysis, but also safety cases can be developed in the following stages. It is worth
noting that in terms of flight safety data, two western airlines use BASIS to input,
analyse and manage flight crew-related errors for their safety information database,
from which safety cases are developed. Two eastern airlines develop their own safety
databases similar to BASIS, due to the fact that their safety case technique is still at
the research stage, and the software interface is unable to meet their requirements,
although they acknowledged the value of BASIS. The other just has a simple safety
database and no plans to develop safety cases so far.
In terms of maintenance aspects, two airlines use MEDA for the input data; one is
under consideration. One western airline has developed a database similar to MEDA.
The small easternairline usesa simple input database.
2. Risk analysis:
Systematically evaluateall potentially harmful eventsand circumstancesto

identify the degreeof risk exposure
The aim of risk analysis is to evaluate the input data and undertake a trend analysis
(even human reliability analysis). Within the interviewed airlines, one of them uses
Risk Analysis Matrix and FORAS. Two use FORAS and one has developeda similar
121
technique but with a different name. It was found that larger airlines tend to use
multiple risk analysesto analyserisks due to the scopeand complexity of their fleet.
3. Risk prevention and control
+ Develop/implement appropriate measures to prevent occurrence of
accidents
+ Evaluate possible harmful effects of undesirable events/circumstances
actually occurring and define those measuresnecessaryto limit/contain

them and re-establisha safeoperatingsituation
The aim of risk control is to minimise or mitigate the risk exposure by using defences
(process/technology). Safety policy should mandate equipment fit. There are a few
techniques employed by the interview airlines, and some equipment is installed on a
voluntary basis, such as EGPWS.
Two points are noticeable here. One is that larger airlines are willing to invest in
expensiveequipment for safety, and are more willing to spendtime in auditing25(e.g.

an internal survey). The other is that the interview airlines tend to focus on the risk
control tools for flight safety (for instance, CRM and FOQA have been paid a great
deal of attention and have been developedfor a long time); however, western airlines
show a greaterawarenessof maintenanceaudits than do easternairlines.
To sum up, the following risk programmes have been adopted by the interview
airlines, and are also regarded as the basic requirement for current airline safety
managementoperations.
Air Safety Reports/Incident report system

BASIS (or other safety information database),and confidential data
exchange(only for BASIS users)
GPWS
25 Another view on this is, Westernairlines comply with a regulatory requirement,while others may
have no regulatory requirementsimposed for that. However, without the leverage will hesitate to
commit to investment.
122
Traffic Collision Avoidance System (TCAS, not for the smaller airlines)
FOQA, or Quick Access Recorders (QAR)
FORAS or Risk Matrix
CRM
Human Factors (HF) and Error Reduction (particularly in maintenance)
4.1.4 Safety Audit
A safety audit is an important componentof the safety managementsystem.It is also
part of the As
safety assessmentprocess and risk management. accidents are either
unforeseeable or unforeseen, the objective of safety auditing activities is to avoid the
latter type of accident or incident. As such, some attention needs to be paid to the
issue of safety monitoring and audit.
The intervieweesthought that it was possible to eliminate risks as a matter of routine

by objectively examining all aspectsof a department's activities that impinge on air
safety.This can be achievedby carrying out a safety audit and taking remedial action
as soon as shortcomings (potential hazards)are identified.
Safety auditing includes both an internal and external review, and inspection26.The
purpose of the routine self-inspection is to confirm that minimum safety requirements

are met for the purpose of the company's safety goals. The result of the safety audit
can also be used as indicators of safety performance27. Some safety managers
mentioned specific auditing programmes (e.g. the large western airline has several
audits respectively for flight operation, ground operation and engineering. There is
also a cross-department audit. The complexity of the programmes is higher than for
the medium and small airlines). Besides, with the growth of alliance and network
26 The external examination and inspection from the regulators will be discussedmore in the section
about the regulator's role.
27 There are other safety performanceindicators such as accident/incidentrates, aircraft proximity

hazardreports and equipment failures to show the level of risk, and the shortcomingsin need of
improvement.
123
sharing, an increased risk of cross-cultural issues also impact operation. Most of the
safety manager mentioned that the new joint relationship aims to increase the
competitive advantages and market share by cooperation and standardisation, which,
nonetheless, inevitably results in significant changes and difficulties in airline
operation, in particular with safety services. Differences in company culture and
safety system result in problems. KAL and Delta are good examples (Please refer to
section 2.6.2.3, footnote 15 for the case of Delta's safety review on KAL).
One interviewee especially mentioned the value of auditing the cultural aspects of
safety28.However, its is
current state of progress to promote the non-blame and
reporting culture (as is the casewith the other four interview airlines) and evaluatethe
safety culture by distributing a safety culture survey (only one interview airline).
There is no practical programmeto supportcultural auditing so far.
4.1.5 Culture versus Safety
The relationship between culture and safety management was one of the major
focusesof the interviews, and Chapter2 has showedits importanceto proactive safety
management. The culture approach was recognised by the interviewees, but as

mentioned in the previous section, the culture approachis making slow progress,due
to the difficulties of measuringso called "culture".
As discussed in Chapter 2, layers of culture can be understood as overlapping
elements. Therefore, a safety culture exists as a subset of organisational culture. Since

the interviewees for the research were selected for their ability to manage safety,
organisational culture was not mentioned explicitly by all interviewees, compared
with safety culture. Below are the resulting discussions of organisational and safety
culture respectively.
28 Glennon (1982) reveals that organisations with poor safety culture scores had higher accident
than those organisations with better safety culture scores. However, there is no further
rates
academic evidence to support this statement.
124
4.1.5.1 Organisational culture as a source of reliability
Although organisational culture attracted less attention by the interviewees, the

interview findings reveal that organisationalculture is likely to be seenas a sourceof
reliability to sustainthe airlines and their safety culture. Reliability is conventionally

embodied in structure and training, but seems to be up against some limits. Here
however, it means the support that organisations can commit and the resourcesthat
they can allocate. That is to say,if the organisational culture places safety as a higher
priority than production, the resource allocation within the airline will favour the
concernsof safety over those of cost.
It was also found that the concept of organisational culture, to the interviewees, was
very abstract. One of the interviewees used the term "organisational climate" to
express a concept similar to that of organisational culture. Other interviewees did not
distinguish between organisational culture and organisational climate. Generally
'
speaking, interviewees regarded organisational culture/climate as the culture which
reflects the belief of employees towards mission, activities, etc that have worked well
in the past, have been assumed to translate into behaviours and norms.
4.1.5.2 Safety culture/climate
Aspects of safety culture were more remarked upon by the interview airlines than
were those of organisational culture. The premise put forth was the observationsof
interviewees, which is that a safety culture consists of a combination of safety
practices (from SMS) and the interaction of the organisation with various aspectsof
its environment.
The phrase "safety culture" is used very commonly in the airline industry, so no
interviewee chose to use "safety climate". However, when asked the difference
between'safetyculture and safety'climate, four interviewees
reckonedthat there was a
slight'difference betweenthesetwo:
125
Chapter 4 lntervie t Rulings ant/ Model Development
Moreover, all interviewees believed that how a firm responds to safety depends
largely on its safety culture. To build a good safety culture is thus the goal of the
interview airlines because an organisation with an appropriate safety culture would be
more likely to implement an appropriate safety management system, and safety
management programmes would not be effectively used if they were not supported by
an appropriate safety culture. It therefore shows the relationship between safety
culture and safety management system (See Figure 4-3).
Figure 4-3 The Relationship between Safety Culture and SMS
SMS
(b) (a)
Safety culture
(a) Interviewees show how a good safety culture can implement a robust SMS
and safety initiatives (e.g. a no-blame culture can encourage line workers to
report errors, which can help to identify hazards)

(b) Premise: Safety culture consists of safety practices and initiatives in the SMS.
4.1.6 The Role of Regulator and the Relationship with Airlines
The final area of interest is the role of regulator and its relationship with the airlines,
because the civil aviation authority is closely involved in the development of the
airlines. The impact of regulatory environment on airline safety has been discussed in
Chapter 2. Below is a summary of the interviews in terms of the interviewees'
thoughts on regulators, which also contain some opinions from the interview on
regulators:
126
Viewfrom Interviewees-
There are many functions provided by aviation regulators. Especially in recent
years, the increasingly intense competition, privatisation, strategic alliances

and new type of operation like low cost operators are taking place with the
developments in the world's airline industry, which reflects a continuing
growth in demand for air transport. This complexity places more stress on the
regulatorsthan ever before.
For airlines, regulators, such as Civil Aviation Authorities, act as monitors and
inspectors in terms of the safety function. The purpose of the routine
inspection by the external regulators is to confirm that minimum safety
requirements are met for approval purposes, both for airlines' and passengers'
own good. Given the commercial competition between airlines, monitoring
and inspecting airline safety is becoming more and more important, in
particular with other issues resulting from the alliance and network sharing,
such as the exchange of cockpit and cabin crew, the requirement for safety
standards, the use of risk programmes and so on. The standardisation and
improvement of global alliances' joint safety performance are therefore in
need of regulators' close attention and inspection.
Viewform regulators -
In addition to this inspection function, regulators also act as assisters.Their
guidance and principles are needed to help to maintain civil aviation

operational safety risks at the level of ALARP (mentionedby the interviewees
as well). By gathering and sharing safety information, the regulatory
environment aims to provide' airlines with the industry's latest principles,
information and programmes.For instance,the concept of human factors has
beenrecognisedas important in the drive to improve aviation safety.
Human Factors encompassespsychology, physiology and engineering. In the

broad array of civil aviation, Human Factors include design and operation of
aircraft, maintenanceof aircraft, provision of air traffic services,etc. As such,

given CRM, accident investigation, design of flight instruments, human
127
limitations, maintenanceerrors, etc being traditionally attracteda great deal of
attention, regulators must pay and have paid much more attention to human
factors than before although remedy of HF is getting more about influencing
organisational culture and value as to the individual. Some areas are in need of
more exploration, such as organisational issues, safety management concepts,

integrated design and learning from incident data.
So far in this industry, with the help of various methods, techniques and tools,
Human Factors advice has been developed to help to apply human factors in
many areas by the trained specialists. For example, the western CAA
interviewed has four HF specialists under its regulation group. The main areas
of their work are:
1. Supplying human factors input to specific tasks/projectswithin the

technical divisions of regulation group;
2. Raising awarenessof human factors and initiating human factors

training acrossthe division appropriateto divisional needs;
3. Fostering relationships with and/or participating in relevant groups

outside CAA, such as JAA, and ICAO;
4. Keeping abreast of developments within human factors and applying this

knowledge as appropriate.
These functions not only feature in the jobs of regulators, but also imply the
interaction between airlines and regulators and the assistanceoffered by the
regulators.
4.1.7 Improvement and Other Comments
The following is a summary of the interviewees' observations regarding safety

improvementsin the air transport industry and in their airlines so far:
128
Safer aircraft designs
Developing and improving national and international regulations

Enhanced staff training
Integratedand intelligent systemsto support first line crews
Human factors issues
One airline interviewee particularly mentioned total safety management,which was a
contribution from the study of human factors and the development of a no-blame
culture.
In addition, one interviewee stressedthat the importance of safety control lay in a
multi-channel feedback system, and the ability to respond effectively to events or

hazards.
One interviewee statedthat "In the current airline industry, the misunderstandingstill
exists quite common in some airlines, such as, `More (extra) safety is costly', `quality
assurancesystemis there to ensureall work is done in accordancewith procedureand
process', etc. These complacent thoughts will only lead to nowhere in safety
improvement,particularly when there is a conflict betweencost and safety."
Another interviewee said that "Risk managementmust be effective becausethrough

this process,risk can be identified, measured,reviewed and controlled. Nevertheless,
it is more important that all employeestruly understandwhat's the value of safety in
all operationsfor the purpose to createa sustainable "

operating environment.
The results of interview will be usedto strengthenthe underpinnedknowledge of the
safety mechanismand the developmentof organisationalfactors.
129
4.2 The Safety Mechanism Model
As previously discussed in Chapter 3, what is required at this point after conducting
the interview is to develop the safety mechanism model, building upon the existing
knowledge of what is thought to contribute to an effective proactive safety by
increasing the knowledge of the factors which affect a proactive safety mechanism,
and the SMS performance, to improve the shortcomings of existing retroactive safety
programmes. The followings will illustrate the development of the model.
4.2.1 Overview and Hypotheses of the Model
The framework of this model is designed at Figure 4-4. The safety mechanism is
organised as a series of concentric circles which are intended to represent the
relationship of each layer, which combineswith organisationalfactors to form a safety

mechanism model. It has both bottom-up and top-down metaphorscharacterisingthe
safety mechanism. The former is that of throwing pebblesinto a pond. If one stone is
thrown in, ripples emanate from the point of entry and continue unchecked. Thus,
carrying the metaphor over to the issue of a safety mechanism,with the stone thrown
into the pond, each level of the system is relatively affected, from each individual's
understanding of the concept, which translates to knowledge, and subsequentlyto

behaviour,which eventually shapesthe organisationalclimate.
Taking this one step further, the organisational climate exerts an influence on
organisationalculture, which in turn affects safety climate and safety culture. Safety
culture influences safety philosophy, and safety philosophy shapes managerial

decision-making,which results in good or bad consequencesfor safety performance.
130
('1l If rr 4 /nh'; !w Fm (It ii. c u, i(I Ii// I )cvc1opnlrn!
Figure 4-4 The Structure of Proactive Safety Mechanism Model
Organisational
Factors
Organisational climate
Organisational Culture>
Safety climate
Safety culture
Safety philosophy>
Decision-making
U
Safety Performance
(of SMS)
131
With having each layer of the composite defined (see Hypothesis 1), this model will
focus on the top-down metaphor of the safety mechanism, which illustrates the way in
which climate/culture presents both opportunities and threats for safety management,
because the climate/culture of an airline is often held as being of critical importance to'
corporate strategic decision making. Namely, safety is ultimately a state of mind.,

(Sayce, 2001). It is supported by organisational structure and regulations, but
fundamentally it is the matter of concept and behaviours, which influence the outcome
of safety (see Hypothesis 2).
The safety mechanism model is not a model of proactive safety management or
organisational climate/culture. It is a model of the evolution of safety performance in

the context of proactive safety management. The implication of using this model toR
assessproactive airline safety is to identify the organisational factors which will affect
the safety mechanism. Meanwhile, external factors are there to present the forces
outside the airline which reflect the environmental factors (see Hypothesis 2).
The difference between retroactive and proactive safety management is in the
treatment of the contributing causes of accidents and their underlying factors to'
prevent accidents from recurring. The traditional approach to studying aviation safety
has involved analysing accident data. Research has focused on identifying factors that
link to, for example, pilot-error accidents through systematic accident investigations.
However, the accident rate has not been significantly improved for decades, which
points to the need for a new approach in safety management.
Proactive safety management moves away from pilot-error as the focus of
understanding accident causation, to the organisational approach, which assigns

responsibility for an accident to all systemswithin an organisation.The investigation
of organisational factors, which acts as the mediator and predictor of the safety,
mechanism and safety performance can help to establish a proactive approach to
safety and improve safety performance(seeHypothesis 3).
The systematicapproachis designedto ensurecontinuous improvement by analysing
132
whether the values of an airline's climate/culture are consistent with good safety
practices. For this purpose, management need to establish a support system so that all
employees can use operational definitions to analyse the resulting data and to identify
barriers to continuous improvement.
Most important of all, chapter 2 stated that the current proactive concept and
approaches aim to uncover the latent organisational conditions, and avoid the
company surrounding errors or inadequaterecognition, which degrade safety in the
workplace. This research therefore aims to establish the organisational factors
affecting the safety mechanismin order to establish the proactive approachto safety.
Certainly it requires an understanding of the traditional aspectsof human factors-
communication, stress, and ergonomics, as well as accident investigation, incident

investigation data. etc, which have been identified in Chapter 2. An airline's safety
health and performance need the co-ordination of both reactive and proactive safety
management(seeHypothesis 4).
Hypotheses
In keeping within the aim of this part of the research,to develop the safety mechanism
within the proactive management,four generalhypotheseswill be explored and tested.

The term hypothesis, as used here, should be understood in the broad senseof the
word as these hypotheses are intended to outline the aims and objectives of the
remainder of this thesis. These statements should be considered as propositions,

developed from known facts, which formed the basis for this investigation. This is
quite different from the hypotheses in the statistical sense that implies specific
inferencesand will be testedand either acceptedor rejected.
133
Hypothesis 1:
The safety mechanism is the composite of organisational climate/culture,
safety climate/culture, safety philosophy, and decision-making. The definition
of each layer will be redefined (except decision making) to suit the airline
industry.
Qr' This hypothesiswill be testedby a literature review.
Hypothesis 2:
The safety mechanism model will be demonstrated as a multidimensional
construct, which will be influenced by a complex factor-structure. Meanwhile,
this factor structure will also reflect the current environmental factors in the
airline industry.
ov" This hypothesiswill be testedby the use of casestudy and questionnaire.
Hypothesis 3:
The organisational factors mediate the relationship between the safety

mechanism and safety performance, and act as predictors of safety
performance.
This hypothesis aims to be tested by the literature and questionnaire,which

includes a self-rated safetyperformancemeasure.
Hypothesis 4:
is
The safety mechanism model seen as being critical to the development of
proactive safety management and evolution of safety performance; however,
134
the airline safety health and performance need the coordination of proactive
and reactive safety management.
r,r This hypothesis will be tested by the combined findings, including the
literature review, casestudy and questionnaire.
4.2.2 Layers and Their Definitions
For the purposesof the model, six systemlayers have been specified. The six layers of
the safety mechanism are: (1) organisational climate, (2) organisational culture, (3)
safety climate, (4) safety culture, (5) safety philosophy and (6) decision-making.
These layers will be defined in the following sections to suit the application to the
airline industry.
4.2.2.11(1) Organisational climate & (2) Organisational culture
The characteristics of organisational culture and their relationship with safety have
been discussedin Chapter2. However, when conducting interviews with airline safety
managers,the term "climate" was found to be used by one interviewee, although the
term "culture" tended to be more commonly used. According to the Collins
Dictionary (1995), "climate" is defined as: a) typical weather conditions of area;b) a
prevailing trend. This would seem to have a different meaning as compared with
culture. Yet sometimes the two are used interchangeably when they are applied to
organisational research. Identification of organisational climate and organisational
culture is neededhowever becausethese two concepts are so global and so abstract
that they can run the risk of becoming virtually meaningless.
What is organisational climate and the distinction between the two
In the 1970s and 1980s, organisational culture and climate attracted a. great deal of
attention because they provided views for managers to overlook their business. Much
135
research was undertaken under the heading of organisational climate in the 1970s.
Gradually, during the 1980s, the term "culture" replaced the term "climate" in this
type of research (Guldenmund, 2000). Organisational culture is seen as a kind of
reliability to sustain a business, as proved in the interview. Nowadays, this concept is
referred to by the term "organisational culture", whereas the term "organisational
climate" has come to mean more and more the overt manifestation of culture within
the organisation.
Nonetheless, debates on distinguishing organisational climate from organisational
culture always exist. For example, De Cock et al. (1986) argue that organisations are
characterised by a coherence of numerous processes. Organisational climate then is

the perception of this coherence by all the members. On the other hand, organisational
culture is the underlying meaning given to this coherence, which forms a pattern of
significance and values (quoted from Guldenmund, 2000).
Conversely, Schein (1992) conceives climate as preceding culture, i. e. climate is
culture in the making. So climate is replaced by culture and culture then conveys a
broader, and more profound and comprehensive meaning. Furnham and Gunter
(1993) regard organisational climate as being an index of organisational health, but
not a causative factor of it. What organisational climate measuresmay accessare

somedimensionsof organisationalculture within a limited range.
Culture is part of organisational climate
In order to have a clear construction of the safety mechanism,this researchintends to

follow the themes of De Cock et al. (1986) and Schneider (1975). In particular, the
latter brought the distinction between descriptive attributes (perception of
organisational practices) and affective attributes (reaction to same practices and
procedures) for organisational climate and organisational culture respectively, given

the difficulties of distinguishing organisational climate from culture. Therefore,
organisationalculture is seenaspart of organisationalclimate.
136
This concept is also supported by Handy (1985). He provided an analogy to explain
why an appropriate organisational climate is important if a management initiative
(efforts to manage safety or some other aspects of performance) is to thrive within the
organisation:
"... many of the ills of organisations stem from imposing an inappropriate
structure on a particular culture, or from expecting a particular culture to thrive

in an appropriate climate. Vines don t grow where the sunshine doesn't fall in
the right proportions with the rain - nor has anyone yet found a more effective
technology for tending the vines than the human hand. "
Using the analogy of the garden, organisational climate is just like the environment,
which needs to be full of sunshine and rain to cultivate the flowers and plants.
Organisational culture is like the ways of coping with the influence of the
environment. For instance, nationally-owned airlines are literally more

politically-orientated than privately-owned airlines in terms of organisationalclimate,
i. e. nationally-owned airlines value their nations' benefits higher than the airlines
themselves.Thus, this type of airline's organisational culture tends to be bureaucratic
and inflexible29.
The detailed definitions and differences between organisational climate and
organisationalculture are listed in Table 4-1.
29 Refer to Table 2-4 (page 81) for Bureaucraticorganisationalculture
4 137
Table 4-1 Definitions of Organisational Climate and Culture
Organisational climate Organisational culture
Describing aspects of the airline's A complex phenomenon of social

state, (political, economic, grouping, serving as the prime medium
brand-value,.. ), tend to be affected for all members of an airline to interpret
by external factors. their collective identity, beliefs and
behaviours.
Perceptions of organisational Reactions to those same practices and

practices, a more superficial concept procedures, created by all organisation
than culture. members (internal factors), not owned
by any group.
An index of airline's health, but not a Used to be taught to the new member as
causative factor of it. the framework for cognitions and
behaviours (reaction) to the problems.
Multi- dimensional Multi- dimensional

(a number of different climates exist (one of the important and direct impact
within an airline ). is safety).
138
4.2.2.2 (3) Safety climate & (4) Safety culture
What is safety climate?
Contemporaneous with the derivation of safety culture from organisational culture is
the associated term "safety climate". The concept of safety climate emerged from
the research on organisational culture and climate. Schneider (1975) argues that a
number of different climates exist within an organisation. Researchers began to
measure one specific type of organisational climate - safety climate. Since then, a
few researchers have defined safety climate (see Table 4-2). As we can see, most
researchers aim at the same concept but differ on what this concept might encompass.
In other words, safety climate tends to be thought of as regarding an employee's
perception of safety, but its operation of the concept varies according to different
companies.
Table 4-2 Definitions of Safety Climate
Researcher Definition
Zohar (1980) A summary of perceptions that employees share about their

work environments
. .......................
............ .......... ........ .........
Glennon (1982) A special kind of organisational climate making employees'
perceptions of the characteristics of their organisation and have
a direct impact upon their behaviours to reduce or eliminate
danger
................................................................................................................................................
...........................................................................
Cooper and Philips Safety climate is concerned with the shared perceptions and
(1994) beliefs that workers hold regarding safety in their work place
... ........ .........................................................

..
Cabrera et al (1997) Shared perceptions of organisational members about their
working environment and about their organisational safety
policies
Williamson et al A summary concept describing the safety ethic

(1997) in an organisation which is reflected in employees' belief
about safety
119
The influence of safety climate
Recently in safety literature, the emphasishas shifted from the individual human who
might be responsible for incidents/accidents (single human error) towards a systemic
or organisational approach. Defined in the previous section, organisational climate
derives from aggregate employee perceptions and consequently, it is
multi-dimensional and can potentially influence safety-related behaviours. Safety

climate is then thought of as the mediating factor betweenorganisationalclimate and
safetyperformanceby someresearchers.
Neal et al. (2000) examinethe impact of organisationalclimate on safety climate, and

the impact of safety climate on safety knowledge, motivation and performance of
individuals in organisations. They found that safety climate operated as a mediating
variable between organisational climate and safety performance, as measured by self-
reports of compliance with safety regulations and procedures, as well as participating

in safety activities, which were also mediated by employees' safety knowledge and
motivation (see Figure 4-5).
Figure 4-5 The Impact of Safety Climate
Safety Safety
Knowledge Compliance
Organisational Safety
Climate 1 -1 Climate
Safety Safety
Motivation Participation
Source:Neal et at., 2000
140
The relationship between safety climate and safety culture
While perceptions are more associatedwith climate, attitudes are consideredto be a
part of safety culture30. The safety culture of an organisation is the product of

individual and group values, attitudes, competenciesand patterns of behaviour. These
characteristics determine the commitment, the proficiency and effectiveness of an

organisation and in particular, its safety management.The safety culture within any
organisation is an indicator of the state of respect for safety consciousness,the
willingness and determination to comply with the company's policies and procedures,
and compliance with regulatory requirements. To individual, safety culture is the
accountability that an individual has to himself or herself; the accountability goes up

to the line supervisionand finally to the employer.
Certain attributes are critical to a strong culture. Understanding performance
requirements,respect for training, respect for peers and supervisorsand professional

pride are examplesof positive attributes that contribute to a healthy culture.
As such, it is found that the model provided by Neal et al. (2000) contains the
31although Neal et al. did not specify it as such. In order
componentsof safety culture
to have a clear understanding of safety climate and safety culture within the airline
industry, this study redefines these two concepts (see Table 4-3) and follows these
definitions when developing the safety mechanismmodel. It is worth noting that one
of the main differences between safety climate and safety culture in the airline
2
industry is that the former formed according to mandatory regulations. As it, is
is
governedby a minimum requirement, safety climate is supposedto pervadethe whole

airline and tends to be regulation-orientated.
3o The definition of safety culture is listed in Table 2-5.

3t According to the Collins Dictionary (1995):
Attitude: the way a person thinks and behaves,opinion, frame of mind
Behaviour: manner of behaving
Belief: trust of confidence, faith, feeling
Motivation: desire, incentive, drive
32 Pleaserefer to Chapter2, Section 2.1.3 for the regulatory environment.
'141
Chapter 4 niter 'icit Finýliýr, L.ý nd 11udc'l /)et clopment
.
Table 4-3 The Definitions of Safety Climate and Safety Culture
Safety Culture
derived from the aggregate employee a series of attitudes, behaviours, and

perceptions of safety social and technical practices
governed by the minimum requirement established to minimise the exposure

from regulators (regulation-oriented) of employees, managers, passengers,
or other aviation authorities and third parties to hazardous conditions
pervading the whole airline " frequently identified as being fundamental

to an airline's ability to manage safety-
potentially influencing safety-related related aspects of operations
behaviours
particularly technical departments are
involved (flight operations, engineering,
maintenance)
A strong safety climate does not guarantee a strong safety culture, but a strong safety
culture must mean a strong safety climate being achieved since safety culture is
fostered by safety climate. For example, CAL suffered two serious accidents in the
four years from 1994-1998. The whole external and internal environment, including
regulation, equipment, safety programmes, training, etc. have been changed in order
to improve its safety climate. However, the accident in 2002 proved that CAL's safety
culture is still in need of improvement, although the general environment and the
perceptions of employees towards safety have been substantially changed.
Safety culture is defined as a series of attitudes, behaviours, and social and technical
practices, which are established to minimise the exposure of employees, managers,
passengers and third parties to hazardous conditions. It is frequently identified as

being fundamental to an airline's ability to manage the safety-related aspects of its
operations (particularly in technical departments, such as flight operations,
maintenance, etc.)
142
Organisations or authorities have provided a number of instructions regarding how to
achieve a "good"33 or "strong" safety culture. ICAO (1994), for example, provided the
following indications of a good safety culture:
+ Senior managementplacing a strong emphasison safety;
ºý Staff having an understandingof hazardswithin the workplace;

Senior management'swillingness to acceptcriticism and an opennessto
opposingviews;
Senior managementfostering a climate to encouragefeedback;
Emphasisingthe important of communicating relevant safety information;
+ Promoting realistic and workable safety rules; and
+ Ensuring that staffs are well-educatedand trained so that they understand
the consequencesof unsafe acts.
Some safety programmes try to develop a Safety Culture Index as an indicator of
safety performance, for example the BASI-INDICATE Safety Programme,discussed

in Chapter2. When there is a measurementsystemin place, the operational deficiency
and the room for improvement are easier to spot. As such, now culture presentsboth
an opportunity and a threat to the aviation industry.
4.2.2.3 (5) Safety philosophy and (6) Decision-making & action
The culture of an organisation is often held to be of critical importance to corporate
strategic decision-making (Johnson and Scholes, 1988). Nevertheless, there should
exist a philosophy before management make decisions and employees take action.
The term "philosophy" signifies the airline management's overall view of how they
are going to "shape" the company and conduct the business. For employees,
philosophy means the priorities and plans that they follow.
33 The meaning of "good" or "bad" safety culture is sometimesvery subjective becausethere appears
to be no worldwide benchmarkingtool or measurementto evaluatesafety culture yet. As such,this
study choosesto use "strong" or "weak" culture, insteadof "good" or "bad".
143
Degani and Wiener (1994) argue that a company's philosophy is largely influenced by
the individual philosophies of the top decision makers and company culture. Although
most airline managers (in this study, the safety managers interviewed) cannot clearly
state their philosophy, such a philosophy of operation does exist within airlines. They
can be in combination with economic factors, political factors, major organisational
change, etc. to generate policies, and can be inferred from policies and procedures to
exercise training, punitive actions, etc. Figure 4-6 shows the three `P' (philosophy,
policies, procedures) framework.
Figure 4-6 The Three `P' Framework in the Airline Industry
(action)
Source:Degani and Wiener, 1994
Safety philosophy is the foundation of safety culture. In other words, safety
defines the organisation's goals. Safety climate may exist in part as

philosophy a
result of the regulation's rules, requirements, and the company's procedures and
to
standards meet them; however, if a philosophy is not extracted at CEO level, safety
culture cannot exist. It constitutes the logic, rationale, plans and priorities towards
safety. It is safety philosophy that drives safety culture in practice and carries out a
continuous self-assessment. Without such a philosophy, an airline cannot ensure that

its safety standards and performances are improving. For example, quoted from
Wiener et al. (1991), Delta Air Lines used to produce a statement of flight deck
automation philosophy, which acknowledged that "the purpose of automation is to
aid the pilot in doing his or her job" and- that the 'pilot must be proficient in
144
operating their airplanes in all levels of automation... must be knowledgeable in the
selection of the appropriate degree of automation. " Furthermore:
"Automation should be used at the level most appropriate to enhance the
priorities of Safety, Passenger Comfort, Public Relations, Schedule, and Economy,
as stated in the Flight Operation Policy Manual.
In order to achieve the above priorities, all Delta Air Lines training programmes,
training devices, procedures, checklists, manuals,... and the day-to-day operations
of Delta aircraft shall be in accordance with this statement of philosophy.
There are two issues about the automation here. The first arises from the ability of the
pilot to properly use the automation in all needs. Different technology requires
different operating methodology. Accidents have occurred where the transitional
issues were not fully understood. The second is that any operational definition to
automation is understood and SOPS amended to account for this. This requires
supports at the training level. By this statement, Delta clearly shows its safety
philosophy on flight deck automation, given the emergent operational problem
generated in the philosophy of operations. With the priority and referents made in the
statement, pilots as well as other employees are provided with a better understanding
of the safety achievement.
The definitions of safety philosophy and decision-making, which are applied to the
development of the safety mechanism, are described in Table 4-4.
Table 4-4 The Definitions of Safety Philosophy and Decision Making
Safety - Logic + plan+ priority towards safety

Philosophy Inferred from policies and procedures
exercise training, punitive actions
Decision- Action
making
145
4.2.3 Organisational Factors
4.2.3.1 Individuals influenced by the safety mechanism model
Before listing the factors influential in the safety mechanism model, it is necessary to
distinguish the different roles of individuals in an airline. These roles serve to define
the individuals influenced by the safety mechanism. The impacts of these specific
individuals on the safety mechanism are likely to be complex and at each level, a
number of contradictory goals are also likely to compete for resources. As such, three
categories of people have been identified as being important to understanding the
factors which contribute to the development of a safety mechanism.
1. Line worker
Individuals at this level perform the essential work of an airline. They are the
resource used by management to achieve its objectives. Frequently they are

described as working at the "sharp end", and include pilots, cabin crew, ramp
personnel, ground staff, etc.
2. Middle management
Individuals at this level are responsible for implementing policy through the
managementof operational resources(e.g. line personnel and equipment). They

do not set the strategic directions for the company. In the airline industry, they
include management pilots, supervisors of ground staff, maintenance line
managers, etc.
3. Senior Management
At this level, individuals are responsible for setting the strategic direction of the
organisation by establishing policy. They do not directly manage line workers,

who are required to report to middle management.Senior managementincludes
headsof department,the vice president and board members.
146
4.2.3.2 Concerns and needs for different groups
This section aims to identify the needsand concernswhich need to be addressed,and
which consequentlyinfluence the acts of individuals in eachgroup, and thus influence

the safety mechanism.
Line workers
Concernsand needsof line workers are thought to include:
1. Individual safety and health

Employees are motivated to participate in safety initiatives safely becauseno one
would wish to be ill or injured at work. This should have a positive influence on
the safety mechanism.
2. Role of line worker towards senseof safety and security

The main role of line worker is in delivering the standard of operation through
adherenceto proceduresor SOPs.As such, line personnelare in the best position

to spot hazards in the workplace. Vigilance of individual members is critical to
safety in an organisation.Besides,many safety initiatives dependon a "no blame"
culture for their success.A perceived threat to job security for speaking up on
safety issueswould erodethe safety mechanism.
3. Safety concepts, skills, knowledge and attitudes

Donald and Cantor (1993) point to that individuals' attitudes and behaviour are
related to the level of safety in operations.Thus, individual and collective attitudes

might influence the developmentof safety mechanismpositively or negatively.
4. Decision-making style
Employees require the training to make, contribution to safety. For example,
incident reporting systemsare used to identify the potential significance of events,
and these not only require the knowledge, but also the co-operation: of line
workers. The processof decision-making will exert an important influence on the
result.
147
5. Communication channel
In order to fully understand the corporate policy, management style, awareness of
safety, etc., line workers need to communicate to others; to their colleagues and to
the management, in the organisation.
Middle management
Concernsand needsof middle managementare thought to include:
1. Management ability
Middle managers must be aware of all aspects potential hazards in order to
manage safety effectively. Also, these managers are expected to encourage line
workers to participate in safety initiatives and express the safety concerns.
2. Role of middle management on decision-making style and process

Senior management usually use performance indicators to judge the performance
of middle managers so that top managers can determine how to allocate the
resources on production and safety. When the financial success of a department is

decided by such process, middle mangers' responsibility is to decide and provide
safety needs while completing the job function under budget limitation. In other
words, middle management is the level at which the organisation can be changed.
If his/her line workers repeat deficiencies, it needs management support to fix it.
Financial pressures can be perceived to limit management freedom to change and
improve.
3. Communication bridge
Middle managersare like the bridge between line workers and top management.
Thus they should act as the primary channeldownwardsto the line and upwards to
senior in
management order to deliver the correct information.
4. Manager's leadership
Being emphasised for long by researchers, management's leadership is important
to the organisation. Budworth (1996) is one of them. He identified the critical
148
essence to the effectiveness of safety professional is leadership and

communication skills. As such, managers' leadership is important to the
developmentof a safety mechanism.
Senior management
The concernsof senior management,basedupon the findings of these interviews and

factors noted in previous chapters,are thought to be:
1. Senior management's commitment

Grimaldi and Simonds (1984) are some of the researchers to stress 'that the
attitudes and behaviour of management have a profound effect on safety practice.

In other words, the management's commitment to safety operations may influence
safety mechanism and consequently safety performance would be affected in an
airline.
2. Maintain the stability in the aftermath of accidents/incidents

Discussedin Chapter 2, risk in the aviation industry is most commonly associated
with threatsto life and limb. Air transport accidentscould risk airline business'and
at worst, might result in an airline's demise. The impact of accidents thus has a
significant impact on an airline's long-term survival if it is not well managed.

Therefore, it is the responsibility for senior management to maintain the stability
of the airline.
3. Leadership and communication capability

Similar to the middle management,senior managementneeds the interpersonal
skills and leadership to communicate with their subordinates.Meanwhile, they
must be able to listen to the 'safety concerns of middle management and line
workers and act upon the input of their subordinates.
149
4. The way to monitor safety performance
Discussed in previous chapters, the use of accidents as indicators of safety
provides an insufficient indication of the actual safety health of the system. As

to
such, senior managers are expected explore more meaningful indicators of
safety and concentrateon organisationalprocessand organisationalgoals not only
outcomes.
5. Financial health of the organisation

As mentioned previously, the goals of safety and production are competing for
resources in airlines and sometimes conflicts are caused as a result, especially in
the era of cost cutting. It is senior management's biggest task to maintain the
airline's financial health because a highly competitive commercial environment
might cause lack of profit, which will reduce the incentives to invest in safety as a
result.
6. Decision-making style - to meet safety and production goals and shareholder

demands
Generatingmaximum profitability has always been airlines' primary goal, but risk
is very expensive to eliminate completely. Thus, airlines' managers make
decisions(allocate limited resources)on the basis of cost-effectivecomprehension,
which results in a trade-off situation between two strategic goals, i. e. safety and
investment costs. Senior managementfocus on what safety investment for how
long may have a negative or positive impact on the airline's safety mechanismby
showing managementsupport.
7. Relationship with the regulator

How to directly develop a relationship of trust with the regulatory authority and
compliance with regulatory practices within the industry is not the biggest task of
senior management. However, the behaviours of senior management are so

indicative to the employees and the public, which will consequently affect the
safety mechanism.According to one CAA regulator, many of senior management
150
underestimatethe extent to which the regulators "trust" the senior managementof

the airlines to deliver a safe system.
8. Relationship with the public
Similar to the above relationship with the regulators. Meanwhile, some

interviewees identified that to establish a good relationship with the public has a
positive impact on airline safety.
Table 4-5 lists the results of the concerns and influences of line workers, middle
managementand senior management.For simplicity of further analysis, Table 4-5

also summarisesthese concerns according to their similarities, and this forms the
constructof the organisationalfactors of the safety mechanism.
151
Chapter 4 /nterrietir Findings any! Model Development
Table 4-5 Various Groups' Concerns & Implied Factors
Group Concerns Implied Factors
Individual safe and health - Perceived safety

Sense of safety and security - Perceived safety
Line Safety concept, skill, knowledge Perceived safety
-
worker and attitudes
Decision-making style - Decision-making style
Communication channel - Communication
Management ability - Management control

,
Middle " Decision-making style and process' - Decision-making style
management Communication bridge - Communication
Manager's leadership - Management control
Senior management's commitment - Management control

Maintain the stability in the - Post-incident/accident
aftermath of incidents/accidents
" Leadership and communication - Communication
capability
" The way to monitor safety - Operation &
Senior
performance maintenance
management Financial health of the organisation Commercial
- pressures
Decision making style- to meet - Decision making style
organisational goals and + The investment
shareholder demands community
Relationship with the regulator - Industry regulation
Relationship with the public - Public relationship
4.2.3.3 Summary of organisational factors
Given that organisational factors are the macro forces that affect safety in an aviation
(Westrum, 1996), these influential factors are divided into internal factors
organisation
factors. The former are those forces from within the company, while the
and external
152
latter express the forces coming from outside the company. These hypothesised
organisational factors are listed in Table 4-6. In addition to the factors summarised in
Table 4-5, are additional factors found in the interviews and the literature, which
include safety information, organisational structure, documentation, country influence
and region influence.
Table 4-6 Breakdown of the Factors
External Factors Internal Factors

1. Industry regulations 1. Perceived safety
2. Public relationships 2. Operation and maintenance
3. The investment community 3. Risk management (control)
(shareholeders) 4. Management control (quality control)
4. Country influences 5. Commercial pressures

5. Regional influences 6. Safety information (info technology)
7. Organisational structure
8. Personnel communication and
relationships
9. Post-incident/accident (impactof
incident/accident)
10. Decision-making style and process

11. Corporate safety policy
12. Documentation
In terms of external factors, country and regional influences are hypothesised to have
influence on the safety mechanism in accordance with Hofstede's (1994) layers of
culture, discussed in Chapter 2, which should take account of national culture,
regional culture, gender culture, etc. Since organisational cultures are the subset of
national cultures, the latter will provide the context in which organisational culture
will develop. Moreover, people are usually part of a number of groups, and they are
potentially influenced by the culture of these groups. Researchers have studied and
confirmed the influence of national and regional cultures on cockpit crew
performance (Westrum, 1996; Ho, 1996; Helmrcich, 1999) and also on maintenance
153
workers (Al-Harabi, 2001). Their analyseshave led to the cultural assumptionarising

from the other groups of people, which has simultaneously affected the safety
mechanism.
Internal factors are summarised from the previous literature review, the interview
findings and the concerns of different groups. Among these factors, "Commercial
Pressures" is the one not acknowledged (or less so) by the airline interviewees, but
one which obviously has a significant influence on safety. Reason (1990) describes
how in a company, production and safety goals compete for a finite amount of
investment in terms of capital, personnel, time, equipment, etc., and Westrum (1996)
notes that safety is often one of the targets for cost-cutting34. As such, the following
paragraphs will try to demonstrate the impact of commercial pressures on the safety
mechanism from an economic point of view.
The economic consequencesof safety translate into accident costs and safety
investment costs. The costs and benefits of safety cannot be measured only in
economic terms; however, the concept of safety costs can be illustrated in the
environment as follows:
According to Pasman (2000), safety costs are the sum of safety investment,
maintenance,insuranceand residual costs,which can be expressedby this equation:
Safety Costs = Safety investments + Maintenance cost

+ Insurance cost + Residual risk costs
34 Actually to "target" safety meansit may manifest itself by deferring equipment fit, defer/minimise
maintenance, training, etc. That does not mean safety is literally affected but does require the
approach to managing safety to be revised.
154
This equationcan be simplified as follows:
Total Safety Costs (TSC 1) = (Safety investments+ Maintenancecost) +

(Insurancecost + Residual risk costs)
= Risks prevention costs (RPC 1) + Accident costs (AC 1)
Prevention costs are the costs investedby airlines in order to improve flight safety and
prevent accidents from occurring. As illustrated in Figure 4-7, the more prevention
costs are invested, the larger is the risk reduction achieved. For example, a
well-trained crew will have more awarenessof abnormal situations which will affect
safety.Accident costs will then be reducedwith the increaseof the risk reduction and
prevention costs.
Figure 4-7 The Concept of Safety Cost
Cost (p) II [deal situation
Total safety cost (TSC 1)
Risk prevention cost (RPC 1)

ALARP ýP1)
Accident cost (AC 1)
ALARP (Q1) Risk reduction level (Q)
E: the optimal safety cost

ALARP: As low as reasonablypracticable
Source:Adapted from Pasman,2000; Hsu, 1999

1
155
Ideally safety can be obtained by maximising risk reduction, which means that the
right side of the diagram is a closer fit. Nevertheless, this comes with a higher safety
cost. From the point of view of an airline's management, they may see the ideal point
as being to the left of point E on the total safety cost line, in spite of the higher
accident cost. An increased effort just to stay even or to attain modest reductions will
not appear unless they are extremely farsighted. Within this area, it means that
is
management willing to risk more accidents so that they can pay less for prevention
and place the resources elsewhere.
To this end, Figure 4-8 portrays the hypothesised safety mechanism model: both
internal and external factors exert influence on safety mechanism,which will manifest
itself on the airline safety services,i. e. the performanceof safety managementsystem.
156
OL
W
c
ö N
-511 iß uý
a)
2
rn
O O
c o
-
O
(u m
O
c
-EC
Q
O D v 0 -
V ö C Ný7ý iß aý cý
aoi U - C
ýC
_ V Co
U
O - O O Ü O .a N
W. C +^+ Q
J p N o v
aýi aýi
E M o o
z a cu c E °0-
ca c a) E
6. o f 0)
LLJ
(0) (ß (L) C -
O
C:
0 ' y
(ß OJ -
a O C) ' Do
«- T CL)
cl) 0 (13 i C
131
= Z E O
Cl) 2 O O .-aN Q
r U ( > -
i , -
O O O
Q. .
ü) iü 0 oO
10
O r N
N M 4 41 (9 hý N 6
r r- r r
w
w
U
v
CD II a
0
0
a
aº ni
I
V
I s
O Eý
a 4 ý.
w,. r 00
4
cv
vL
ý_
.C
7
'n
N
E
L-L3 co
O °
O o
is c v
o
ü o °' a> CL) n
L. a c
J - o
Z a) CL) 0 OD
> C
v
CC O ß_ ,v - O
7 a1
LLI
:3 ,- co 0)
W . N M V Lf)
157
Blank Page
CHAPTER 5
Results of Model Test
"The most important thing in life is to have the great aim and
to possessthe aptitude and perseveranceto attain it. "
Johann Wolfgangvon Goethe
5.0 Introduction
In accordancewith the researchmethodologiesdescribedin Chapter3, a retrospective
casestudy and a survey of opinions with questionnaireare adaptedto test the model,
which will subsequentlybe presentedin the next sections.
5.1 Qualitative Method -A Retrospective Case Study
5.1.1 The Event
At 11.55 EST (Eastern StandardTime) 35on Friday, March 10 1989, The Canadian
carrier - Air Ontario Flight 1363, a Fokker F28 1000, departedThunder Bay about one
hour behind schedule.The aircraft landed at Dryden at 11.39 CST (Central Standard
35.
Time) It was being refuelled with one engine running, becauseof an unserviceable
35 EST = GMT minus 5 hours

CST= GMT minus 6 hours,
GMT: Greenwich Mean Time _.,
159
ChapterS Resultsof Model Test
Auxiliary Power Unit (APU). Although a layer of 1/8-1/4 inch of snow had
accumulated on the wings, no de-icing was done because de-icing with either engine
running was prohibited by both Fokker andAir Ontario.
Since no external power unit was available at Dryden, the engines could not be
in
restarted caseof engine shutdown on the ground. At 12.09 CST the aircraft started
its take-off roll using the slush-coveredRunway 29. The Fokker settled back after the
first rotation and lifted off for the secondtime at the 5700ft point of the 6000ft runway.
No altitude was gained and the aircraft rushed in a nose-high attitude, striking trees.
Less than one kilometre from the end of runway, Flight 1363 became a mound of
smoulderingmetal and the deathtrap of 21 passengerand three crew members.
After a 20-month investigation of the probable cause of the accident, the inquiry's
report concluded that "Captain Morwood, as the pilot-in-command, must bear
responsibility for the decision to land and take off in Dryden on the day in question.:
However, it is equally clear that the air transportation systemfailed him by allowing
him to be placed in a situation where he did not have all the necessary tools that
should have supported him in making the proper decision. " To the benefit of the
Canadian Aviation System, the accident became the subject of the most pervasive and
intense inquiry in the history of aviation.
In the introduction to the inquiry's lengthy report, Mr. JusticeMoshansky outlined the
systemsperspective adopted by the commission. The Inquiry set out to identify the
elements of the aviation system and examine each in turn. It is shown that this
accidentwas the result of a failure in air transportationsystem.
The failure to which Moshansky refers are those events and conditions which led toý
the Captain of Flight 1363 finding himself in Dryden, behind schedule in poor
weather conditions, without the possibility of de-icing the aircraft and without
stranding the passengerson board in Dryden at the start of a holiday weekend. It is
these conditions which defined the operational environment encountered by the

individual, and to a large extent predeterminestheir responseto that environment. In
essence, theseconditions combine to form the "safety mechanism" in Air Ontario.
160
Chapter 5 Results of Model Test
5.1.2 The Pathogen Trails by Appling the Safety Mechanism Model
The following are the pathogen trails dissected by applying the safety mechanism
model with a bottom-up metaphor:
Result (Failure)
The crash happened because of a loss of lift caused by a build-up of ice on the
wing and the aircraft crashed shortly after take-off from Dryden.
Decision- Making
Layer
Decision- making
Individuals
1. Captain took off without de-icing in snowy condition.
Line 2. Maintenance left APU un-serviced36 when the aircraft left

Winnipeg that morning.
personnel
The dispatch team sent the disabled aircraft to an airport lacking

Middle
ground start facilities in bad weather; furthermore, APU was left
management inoperative.
Senior Managers intended to maximise the utilisation of aircraft and

profit so the aircraft had to fly anyway.
management
3e APtJ is MLI. (Minimum Equipment List) item, Captain has to determine the effects.
161
('hapter 5 Rcmilts u/ Model Test
Safety Philosophy
Layer
Safety Philosophy
Individuals
1. Line personnel were highly motivated to keep the aircraft flying

Line
and took necessary steps to do this.
personnel
2. Ground handlers were reticent.
1. Managers appeared to regard the on time performance and
Middle
aircraft dispatch rate as prime performance criteria.
management 2. Staff put the aircraft into service before an adequate supply of
spare parts had been obtained, and did what was necessary to
keep the aircraft flying.
1. Managers saw the flight was behind schedule in poor weather
Senior
conditions
management itlhout the

Managers required the aircraft had to llv, Cvcii %N
possibility of de-icing, because passengers could not be
stranded in Dryden at the start of a holiday weekend
162
Safety Culture
Layer
Safety Culture
Individuals
1. Cockpit crew behaviour: did not walk around to have the aircraft
inspected, so the development of ice was left unchecked.
2. Cabin crew attitude: did not communicate well.

(instruction was given to flight attendants, discouraged them from
informing the air crew of the ice building up on the wings)
Line
personnel 3. Maintenance: deferred maintenance had become normal practice
for maintenance and flight crew working on the F-28.
4. Ground handling personnel: it was okay to refuel while the engine

was running. (the aircraft left Winnipeg that morning with an
unserviceable APU. Dryden Airport possessedno ground-start
equipment, so at least one engine had to kept running during the
stop over)
1. Management failed to deliver and communicate safety information

Middle to the line personnel.
management
2. Management ineffectively observed and coached safe behaviour.
(Flight crew and maintenance staff used different information to
determine what constituted "essential equipment").
Senior Management lacked communication with key persons, and safety

management managers had no direct access to CEO.
163
Safety Climate
Layer
Individua S Safety Climate
I. Maintenance: there was no harm in running the risks of

maintenance deferment.
2. Crew: a number of mandatory occurrences were never

Line reported to Transport Canada. Crew were obliged to keep the
personnel aircraft flying.
(The involvement of line personnel in any safety initiatives was

probably minimal because there was a significant degree of labour
unrest in the months preceding the accident. )
I. Had no approved MEL for F-28, which served to outline which

systems were of a degree of importance that the aircraft could
not be operated without them.
Middle
management 2. Had no operation manual for F-28, so pilots referred to various
other manuals for information, including a Fokker manual
which had not been updated for four years.
3. Failed to provide training in jet operations and performance to

dispatchers.
I. Failed to develop a safety organisation (committee).
2. Appointed a management pilot as CP who had a significant

Senior number of duties in the turbulent period while. (The check pilot
management (CP) and first flight safety officer (FSO) resigned after one
month in the position due to lack of management support, the
position of FSO remained vacant until the accident)
3. Forbad de-icing with the engine running. (and aircraft

manufacturer of F-28)
164
Organisational Culture
Layer
Organisational Culture
Individuals
The airline was new (merger of two small airlines with different
Line operating cultures). It was experiencing considerable
personnel organisational change, including a lengthy pilot strike,
introduction of two aircraft types, reduction of the workforce,
etc.
1. Either accepted or were unaware of the improper safety

practices throughout the company.
Middle
management 2. Over taxed, inexperienced with jet operations.
3. Lacked the support of senior management.
1. Need to deal with different company culture, labour unrest,

replace aging fleet with newer equipment, etc., and were faced
with a merger.
Senior
management 2. Lack of the commmmcation of an organisational mission
including safety, organising work to achieve the mission and
striving for the continuous improvement in terms of safety.
165
Chapter 5 R< s,ilt. scof Model Test
Organisational Climate
Layer
Organisational Climate
ndividuals
Given the company environment being turbulent (merger, pilot

Line strike, etc.) and the recession in the 1980s where aviation jobs
personnel were difficult to find, were likely to do what was necessary to
ensure their own job security.
1. As above.
2. The developed working environment was "a new airline which

Middle is responsible for the emphasis placed upon operational goals
management by senior management".
1. To maintain viability in the competitive airline industry,

commercial imperatives were seemed to be felt by senior
management to a greater degree than the requirement to
Senior increase safety, partly in the absence of regulatory pressure
management from Transport Canada.
2. The developed working environment was "a new airline which

is responsible for the emphasis placed upon operational goals
by senior management".
5.1.3 Conclusions of the Organisational Factors Which Influence the

Safety Mechanism
After analysing the safety mechanism of Air Ontario, the underfyin (influential)
factors of the safety mechanism can be illustrated with a backward approach as
follows (see Figure 5-1):
166
ChapterS Resultsof Model Test
Figure 5-1 Findings and Conclusions of the Safety Mechanism Model
Applied to Air Ontario Accident

Note: 1. Org. = Organisation,Mgt. = Management,Phil. = Philosophy, a/c =Aircraft
2. Pleaserefer to section 5.1.3.1 & 5.1.3.2 for the description of the organisationalfactors.
Line Worker Middle Mqt. Senior Mat. Ora. FACTORS
Theinvestment
Org. " job security " future of company " operation goals """"""""""""";:tº E 3: community
"ý Commercial
I 11: pressures
Climate " future " operation goals " viabilit Y+ survival -
" country regulatory E 1: Country
influence
"ý."" """""""ý
Org.
I " turbulent
pressures
" lack communication

".. E 2: Industry
regulation
I 10: Personnel &
" org. changes relationship
communication
Culture " turbulent environment with middle mgt.
environment (overtaxed, ) " merger undergoing I 9: Moýe't
...
"""senior,
" lack of support from
mgt.
18: Documentation
Safety " minimal safety " failure to issue MEL " poor regulatory ".......... """º E 2: ndust
ön
Climate initiatives " failure to provide environment
" bent to keep training and " short-term focus in 17: management
a/c flying manual turbulent environment 16: operation
and
maintenance
Safety t
" deferral of " safety information not " lack communication
Culture maintenance communicatedwel with key persons
I 5: Sf Tation
" lack of training, " little observationan " safety manager has
in safety coaching in safety no direct access to
14: Organisational
behaviour CEO
knowledge structure
13: Perceived
Z safety
Safety " highly motivated--w-on time performance-+" flight was behind /

12: Corporate
Phil. to take necessary was prime criteria schedule safety policy
steps to keep -'" take necessary steps passengercould not be
a/c flying to keep a/c flying stranded on board
1
Decision " Captain took --- l- Sent the disable -º" a/c had got to fly I1:
Decision
making
process
-making off without de-icing a/c to an airport
" Maintenanceleft without ground
APU unserviced starting facilities
Result
1
Dryden Accident
167
5.1.3.1 Conclusions of internal factors
1 1. Decision Making Process

This factor featured the action and decision-making processof the individuals
in Air Ontario. It clearly showed that the decisions of line workers were
influenced by senior management.
1 2. Corporate Safety Policy

Safetypolicy was motivated by on-time performanceand aircraft utilisation.
1 3. Perceived Safety
The F-28 aeroplanewas new to Air Ontario and it was the first jet aircraft to
enter service with the airline. As such, inexperience could be attributed to a

failure on the part of the maintenancecrews, dispatchersand pilots to realise the
potential implications of an unserviceableAPU where no ground start facilities

were available.
14. Organisational Structure

The structure of Air Ontario was not clear or healthy enough to fulfil its safety
responsibilities.
1 5. Safety Information
Safety information was not communicatedwell throughout the company.
16. Operation and Maintenance

Deficient scheduling (over-commitment of F-28). There was no standardised
manual which caused ambiguous operative, dispatch, and maintenance

procedures. Deferred maintenance resulted from inadequate sparespurchasing
and inexperienceon the part of maintenance(lack of skills and knowledge).
168
1 7. Risk Management
The reporting of hazards was insufficiently practiced and the regulatory
environmentwas not completewithin the company.
1 8. Documentation
Maintenance personnel regularly defer repairs while awaiting spare parts, and
pilots would delay recording technical problems in the aircraft log until the end
of day if they thought the report would serve to ground the aircraft pending
repairs. The practice of passing notes from one crew to the next so the last crew
of the day could record all of the technical problems was commonplace.
1 9. Management Control
The merger of two airlines caused some managementchanges and pressures
becauseof the different cultures of the two companies.The fact that managers
could not have full support from the top manager resulted in the management
control problem.
1 10. Personnel Relationships and Communication
Low morale caused inefficient communication between line personnel and
management.
I 11. Commercial Pressures

To ensure that the organisation continued to exist, the primary concern facing
senior management was the viability and survival of the airline. Increasing
competition and commercial imperatives increased commercial pressures and
the dangerof decreasedsafety investment.
169
5.1.3.2 Conclusions of external factors
E 1. Country Influences
Deregulation failed to produce an effective regulatory environment.

Governmentfailed to respondto warnings from the industry and the regulator.
E 2. Industry Regulations
Transport Canada failed to provide clear minimum operating standards for
operators.It also failed to be aware of the operational failings.
E 3. The Investment Community

The merger had caused the investment community to focus on operating
revenue.
The case study does not reveal the hypothesisedinternal organisationalfactors - post
accident/incident (because of no previous accidents), and external organisational

factors - public relationships and regional influence (because of no supportive
information). It shows that any single occurrencewill be unlikely to demonstrateall
the factors of the model, and as such the model should not be rigidly applied to
Nevertheless,
particular cases. it is hoped the readerswill have a better understanding
of how the safety mechanism is useful in terms of identifying the number and
complexity of interactions involved in the formation of a safety mechanismas set out

in the hypothesisof this model.
170
Chapter 5 Resultsof Model Test
5.2 Quantitative Study - Safety Questionnaire
Followed the case study, safety questionnaire is the next method used to validate the
model. As described in Section 3.3.2, this questionnaire is designed to consist of two
parts - internal factors and external factors in accordance with the hypothesised
organisational factors. As such, fifty-seven items were included in the first part of the
questionnaire, dealing with the concerns of individuals in the internal environment.

These items were intended to examine the degree to which the individual felt part of
their organisations or considered organisational concerns when making choices at
work. Included in this section were the questions related to the internal factors: "The
concept of perceived safety", "Concerns in the aftermath of incident and accident",
"Personnel relationships and communication within the company", "The recognition
of organisational structure", "The development of information system technology",

"Internal pressures from the commercial activities", "The ability of management
control", "Risk management programmes", "Operation and maintenance related
activities", ý"Decision-making style and process", "Corporate safety policy", and

"Concerns of written documentation".
In this section, questions related to the "Perceived safety", "Personnel relationships
and communication", and "Operations and maintenance" are referred to in the SCQ
developed by Glendon and Litherland, (2001), and Glendon et al. (1994), whose
research investigated the structure of factors within a safety climate and the
relationship between safety climate and safety performance. Six factors, including
"Communication and support", "Adequacy of procedures", "Work pressure",

"Personal protective equipment", "Relationships" and "Safety rules", were identified
by these two research groups led by Glendon. Due to the similarity of the intended
questions, four items loaded on the factor "Communication and support", three items
loaded on "Relationships", two loaded on "Work pressures" and two loaded on
"Adequacy of procedures", one loaded on "Safety rules", and these were selected for
use in the internal environment part of the safety mechanism questionnaire.
In the second section regarding the external factors, eighteen items were included, to
171
examine whether the external environment can exert an influence on airline safety,
and in particular, on safety culture. Included in this categorywere the external factors:
"The influence of industry regulation", "Public relationship", "Investment community
(such as investors, etc)", "Country influence" and "Regional influence", "Terrorism",
and "Political issues".
5.2.1 Finding from Pilot Study
Section 3.3.2.2 has outlined the method and participants of pilot study. The result of
pilot questionnaire was encouraging in a number of respects. Although these
participants generally accepted the questionnaire, some issues were raised which
needed to be addressed in the questionnaire itself. These problems are reviewed below
and actions taken to rectify there are outlined.
1. Re-editing some questions (issues)in the survey:
For example, over 80 percent of people suggest that the question about the
satisfaction of "organisational learning" was too conceptualto be included in the

survey, terrorism and political issuesshould not be included, etc.
2. Enable respondents to select a category of "don't know":
In the pilot questionnaire, there is no category for "don't know" because the
respondents (the management) were expected to know the answers. However, in
an attempt to enhance the accuracy of the questionnaire, steps were taken to
expand the scale of response. In addition, "don't know" responses may be

indicative of safety system failure.
3. Increase one self-rated question to assessthe safety performance of

the organisation:
This will be analysedseparatelyand comparedwith the results of questionnaire.

The question: How would you rate the safety performance of your company with
172
respectto the rest of airline industry? This variable was measuredon a six point
Likert scale ranging from "Below Average for Industry" through to "Average for
Industry" and "Above Average for Industry".
In summary, the substantive changes made to the questionnaire between the pilot
study and the final study were:
The exclusion of three issues (one safety concept, terrorism, world-wide
political issues).
The addition of a response scale and self-rated question item.
The addition of an expandeddefinition of the purposeof the questionnaire.
For the questionnairedistributed to Taiwan and Mainland China, a Chinese

translation copy was addedfor the convenienceof the respondents.
No changeswere made to other parts (other items) of the questionnaire.
5.2.2 Results of the Safety Survey
This sectionpresentsthe results of questionnairestudy.
A total of 325 questionnaireswere sent out (see Appendix K for the distribution list).
Nine were returned and marked "cannot deliver" or "addresseehas gone away".
Therefore 316 were assumedto have reached the intended recipient. Of these, two
people returned questionnaireswhich were only partially completed and they were
deleted from the analysis. 104 questionnaires were returned completed. This
constitutes a total responserate of 33.54 percent (106 out of 316) and a completed
responserate of 32.91 percent37(104 out of 316).
37 The responserate for a market researchsurvey is 25 percent in average.
173
5.2.2.1 Respondent demographics
Figure 5-2 shows the distribution of questionnaires and the sample obtained around
the world. As can be seen, 72 questionnaires were sent to North America, of which 22
were returned. In Europe, 107 were sent out and 36 were returned. In the Middle East,
16 were distributed and 2 were returned. 20 questionnaires were sent to Africa and 3
were returned. In Latin America and the Caribbean,23 were sent out and 5 were returned.
87 were distributed to the Asia and Pacific region and 36 were returned. Appendix M
lists the comments and suggestions of the participants.
Figure 5-2 Distribution of Final Sample by Region
IVUIC. l'! ti- lave ul till la. l ILu, LAIi- L{I LII! i La. a. lwu uuu uiL ý. ul lva/Lul , .a Lulvýý.,
AF- Africa, ME- Middle East, AP- Asia and Pacific
174
The distribution of respondents from around the world was not equivalent to the target
sample. The proportion of returned questionnaires from Africa and the Middle East
were much lower than the intended sample rates; both of them were lower than 3
3. In from North America
percent addition, the proportion of completed questionnaires
and Europe were only one or two percent different from the percentage of actual
sample rate. The proportion of returned questionnaires from Asia and the Pacific
region was higher then expected. Apart from the personal preference, the reason
affecting the willingness to response the questionnaire may have a lot to do with the
company `culture' and policy, some of which clearly state that response to external
questionnaire is prohibited. The proportions of returned questionnaire are as indicated
in Table 5-1.
Table 5-1 Distribution Percentage of Final Sample by Region
Region Intended sample Actual sample
(% of total 316) (% of total 104)
Africa 6.32 2.97
Asia and Pacific 27.5 35.6
Latin America and the 7.27 4.95
Caribbean
Europe 33.86 35.6
Middle East 5.06 1.98
North America 22.7 21.7
38 The potential impact of an uneven sample on the survey result will be discussed in next chapter
when interpreting the results of survey.
175
5.2.2.2 Survey Results
Table 5-2 shows the factor structure extracting from the survey responses, slightly
different from the hypothesised model (Figure 4-8 in pagel57), which, however,
confirm the existence of organisational factors within the safety mechanism, and
provide a greater understanding of the forces at work within the internal and external
working environment of the airline industry. The following chapter will further
interpret the results by applying the statistical analyses and discuss their implication.
The hypothesised model in Figure 4-8 and the result of this safety survey are
compared in Chapter 6, section 6.6 (page 211).
Table 5-2 The Factor Structure of a Proactive Safety \lechanism

from Survey Result
Internal factors External factors
1. Employee safety attitude& behaviour 1. Influences of region and country
2. Employee safety concept 2. Public and the media influence
3. Level of operational safety in 3. Impact of regulatory environment
operation and maintenance 4. Involvement of investment
4. Corporate safety policy community
5. Personnel- quality of working life

6. Employment of risk programmes
7. Impact of accident/incidents
8. Financial concern
9. Procedures and documentation
10. Commercial cost pressures
11 Organ isational structure and
management commitment
12. Communication system
13. Necessity of safety reports
176
CHAPTER 6
Analyses and Discussions
"There is unlikely to be a single universal set of indicators for all

types of hazardous operations, one way of communicating how safety
health can be assessed is by listing the organisational factors that are
currently measured. "
-Prof. J. Reason,1995
6.0 Introduction
This chapter aims to analyse the survey results, explore and discuss the
inter-relationships between variables (question items) in the questionnaire.Since the
questions are classified into internal environment and external environment, the
analysis of the organisational factor structure of the safety mechanismwill be divided
into two categories,i. e. internal and external factors. The results for internal factors
and external factors are presentedfirst to analyse their relationship with the items in
the questionnaire.Then these factors are discussedrespectively in section 6.5 (page
195) as well as comparedto the hypothesisedmodel in section 6.6 (page211) in order
to further investigate their implication and how these factors can make their
contribution to the industry.
:;4,.
177 1 "'
Chapter 6 Analysesand Discussions
6.1 Organisational Factors in Internal Environment
In keeping with the aim of this chapter,Principal ComponentAnalysis, followed by a

varimax rotation, was performed on the 56-item (internal environment) questionnaire
data from 104 respondentsin order to examinethe organisationalfactor structure.
The data were deemed to be suitable for the analysis (data reduction procedures), as
indicated by the Kaiser-Meyer-Olkin Measure of Sampling Adequacy value of 0.74
(Hair et al., 1995). The Bartlett Test of Sphericity was significant [x2 = 4527.621, P<
0.05], indicating that correlations exist among some of the response categories (see
Table 6-1).
Table 6-1 KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of 740

.
Sampling Adequacy.
Bartlett's Test of Sphericity Approx. Chi-Square 4527.621

df 1326
Significance 000
.
This method of factor extraction was preferred since it usesthe maximum amount of
variance available in the data, and the varimax rotation was selectedsince orthogonal
factors would simplify both factor interpretation and later analyses by providing
uncorrelatedfactors.
The first analysis yielded a sixteen-factor solution, which accountedfor of 76 percent
of the variance. However, this interpretation was rendered problematic because of

four complex items, each of them loaded on two factors. As a result, theseitems were
removed from further analysis (Question 14,36,37 and 44). A subsequentanalysis of

the remaining variables yielded a new fifteen-factor solution. Severalcomponentshad
high factor loading (>.70) and some variables had low loading numbers, such as
variable 20,32, etc. These items were treated as suspectand may later be considered
for removal from the analysis (Hair et al., 1995). Appendix N shows the reproduced
RotatedFactor Loading Matrix.
178
Factor loading (or loading number) means the correlation coefficients between the
variables (questionnaire items) and factors. It shows the extent to which the
questionnaireitems are correlatedto the factor. The subsequentsection therefore aims
to illustrate the underlying nature of each factor, which is characterised by the
groupedquestionnaire items.
6.1.1 Loading Factors from Principal Components Analysis (PCA)
The Principal ComponentsAnalysis produced some interesting results. The primary
groupings of concern are shown together with the loading factors in the following
The
paragraphs. items loading primarily onto each factor were examined to seeif the
factors made theoretical sense and each factor is labelled in terms of its common
underlying dimension nature. This result not only shows the current situations, but
also demonstrateswhat representativesthought to be important.
1) Factor 1: Employee safety attitude and behaviour
All of the items which loaded onto this factor were concerned about the safety
behaviour and attitude demonstratedby airline employees. Therefore, factor 1 is
labelled as "Employee safety attitude and behaviour". The questionnaire items
which correlated/loadedagainst"Employee safety attitude and behaviour" are listed

in Table 6-2.
Table 6-2 The Underlying Nature of Factor I "Employee safety attitude and behaviour"
v Loading Item
03 815 Employeesuse correct safety proceduresto carrying out the job.
.
04 774 Employees ensure the highest levels of safety when carrying out the job.
.
07 640 Employeesvoluntarilycarry out the tasks or activitiesthat help to improvesafety.
.
02 608 Employeesknow how to perform their job in a safe manner.
.
1 462 There is an organisationalawarenesstowards safety in the airline.
.
Note- V: Questionnaire item number, Loading: Correlation to the factor
2) Factor 2: Employee safety concept

This factor dealt with the safety concepts of airline employees, and the degree to
179
which that safety perception was shared. Thus, factor 2 is labelled as "Employee
safety concept". The questionnaire items which correlated/loaded against

"Employee safety concept" are listed in Table 6-3.
Table 6-3 The Underlying Nature of Factor 2 "Employee safety concept"
v Loading Item
09 834 Safety rules can be followed without conflicting with work practices.
.
05 830 Employeesbelieveflight safety is an importantissue.
.
06 609 Employees feel that it is important to maintain safety at all times.
.
08 537 Employees are encouraged to submit ideas to improve safety in the airline.
.
35 434 Safety managementaims are sufficientlysupportedwithin the airline.
.
Note:V- Questionnaireitem number,Loading:Correlationto the factor
3) Factor 3: Level of operational safety in operation and maintenance

This factor "Level of operational safety in operation and maintenance" was
concerned with the level of operational safety in association with front line, i. e.
operation and maintenance issues, especially regarding training. Therefore, it is
named "Level of operational safety in operation and maintenance". The
questionnaire items which correlated/loaded against "Level of operational safety in
operation and maintenance" are listed in Table 6-4.
Table 6-4 The Underlying Nature of Factor 3 "Level of operational safety in operation &
maintenance"
V Loading Item
46 940 There are adequate opportunities to express views about operational problems.
.
43 930 Training is carried out by the individualswith relevantoperationalexperience.
.
48 885 There is an effective mechanism by which the safety manager or the safety
.
committee can report to the CEO and can make recommendations for a change or
action
42 420 Potential errors, consequences and recovery point are identified in training
.
4) Factor 4: Corporate safety policy

This factor groups variables concerning about the safety goal and policy issues
within airlines. Therefore, it is labelled as "Corporate safety policy". The
questionnaire items which correlated/loaded against "Corporate safety policy" are

listed in Table 6-5.
180
Table 6-5 The Underlying Nature of Factor 4 "Corporate safety Policy"

v Loading Item
53 821 The airline has a clearly stated set of goals and objectives.
.
52 763 The policy statementsdefine the airline's fundamentalapproachtowards safety.

.
55 568 An effectivedocumentationmanagementsystem ensures the availabilityof
.
procedures.
51 517 Safety statements and policies of an airline define the management's intention in
.
safety matters and company's commitment to safety.
Note: V- Questionnaire item number, Loading: Correlation to the factor
5) Factor 5: Personnel - quality of working life

This factor dealt with the personnel working quality and relationships within the
airlines. The low variable loading onto this factor of item 20 (0.345< 0.4) would be
deleted in the further analysis. Therefore, it is labelled as "Personnel - quality of
working life". The questionnaireitems which correlated/loadedagainst"Personnel -

quality of working life" are listed in Table 6-6.
Table 6-6 The Underlying Nature of Factor 5 "Personnel - quality of working life"
v Loading Item
17 825 Personnel are confident about their future within the airline.
.
18 821 Morale is good.
.
45 619 Frustrations that arise from factors outside staff control can be accommodated.
.
without adversely affecting work.
19 599 Good working relationshipsexist in the airline.
.
20 345 Employees'jobs are well-defined.
.
Note-V: Questionnaireitem number,Loading:Correlationto the factor
6) Factor 6: Employment of risk programme

This factor consistedof three items and representedthe use of risk programme and
the degree to which the risk programmes have influence on airlines' safety.
Therefore, factor 6 is labelled as "Employment of risk programme". The
questionnaire items which correlated/loaded against "Employment of risk

programme" are listed in Table 6-7.
181
Table 6-7 The Underlvinq Nature of Factor 6 "Employment of risk programme"

v Loading Item
26 718 Data collection,analysisand presentationhave an influenceon safety

.
performance.
38 685 An effectiveongoing hazard identificationprogrammehas an influenceon
. organisational safety culture.
24 515 An adequate system exists for transmitting critical information regarding safety
.
within the airline.
Note- V: Questionnaire item number. Loading: Correlation to the factor
7) Factor 7: Impact of accidentlincidents

This factor was concerned with how to deal with the aftermath of
accidents/incidentsand how airline safety performance is influenced by accidents

and incidents. Thus, it is named "Impact of accident/ incidents". The questionnaire
items which correlated/loadedagainst "Impact of accident/ incidents" are listed in
Table 6-8.
Table 6-8 The Underlying Nature of Factor 7 "Impact of accident/ incidents"
v Loading Item
849 After an accident has occurred, appropriate actions are usually taken to reduce the
11 . chance of reoccurrence.
12 777 After an incident has occurred, appropriate actions are usually taken to reduce the
.
chance of reoccurrence.
13 738 There is a documentedbusinesscontinuityplan in the event of accidents.
.
10 700 There is an appropriateEmergencyResponsePlan.
.
Note- V: Questionnaireitem number,Loading:Correlationto the factor
8) Factor 8: Financial Concern

This factor only consistedof two items and was somewhatdifficult to define. Both
of them seemed to identify airlines' financial concerns,but it was of interest that

they were not groupedwith Factor 10, which presentedthe commercial pressuresof
airlines.
In the case of financial goals, this representsthe degreeto which a conflict exists
between safety and financial goals. In the caseof shareholders'welfare, it represents
whether the welfare of shareholdersand organisationalsafety culture are correlated,
182
i. e. the company's profitability and safety are correlated. According to both of their
similar essence, this factor was resolved by the level of "Financial concern". The
questionnaireitems which correlated/loadedagainst "Financial concern" are listed

in Table 6-9. Meanwhile, as the relatively high proportion of variancewas explained
by this factor (93 percent,refer to Appendix 0), it was decided to keep this factor in
for further analysis.
Table 6-9 The Underlying Nature of Factor 8 "Financial concern"

v Loading Item
30 876 There is no conflict between safety and financial goals.
.
31 854 Shareholder's welfare and airline's organisational safety culture are correlated.
.
9) Factor 9: Procedures and documentation

This factor derived from the analysis which focused on the importance of written
documentation and procedures regarding safety within the company. Therefore,
factor 9 is labelled as "Procedures and documentation". The questionnaire items
which correlated/loadedagainst"Procedures and documentation" are listed in Table

6-10.
Table 6-10 The Underlying Nature of Factor 9 "Procedures and documentation"
v Loading Item
54 669 The roles and responsibilities for the personnel in the safety management
.
s stem are early defined and documented.
56 576 Written work proceduresmatch the way tasks are done in practice
.
49 536 In the event of CEO making an unfavourableresponseto a safety
.
recommendation,there is a procedurewhereby the matter is monitored by the
safety manager or the safety committee until it is resolved.
10) Factor 10: Commercial cost pressures

This factor dealt with the degreeto which commercial pressuresinfluenced airline
safety. Interestingly, item 27 and item 28 renderedminus loading numbers,because

the statements of these two items used negative expression to emphasise the
problems. As predicted, they were grouped together within the factor concerning
commercial pressures. Hence, even though the numbers concerning internal
183
consistency" were affected by their low proportion, this factor "Commercial cost
pressures" remained valid. The questionnaire items which correlated/loaded against
"Commercial costpressures" are listed in Table 6-11.
Table 6.11 The Underlying Nature of Factor 10 "Commercial cost pressures"

v Loading Item
27 -. 818 Management is concerned for cost more than safety.
28 675 Safety budget is the first item to be reducedwhen commercialpressures
-.
emerge.
34 534 The values of management are identified as being safety orientated.
.
29 517 Safety rules are adheredto even under cost pressures.
.
Note: V- Questionnaire item number, Loading: Correlation to the factor
11) Factor 11: Organisational structure and management commitment

This factor consisted of four items, which mainly dealt with the degree to which
airline safety is affected by organisationalstructure and managementcommitment.

Therefore, it is labelled "Organisational structure and managementcommitment".
The questionnaire items which correlated/loadedagainst Organisational structure
and managementcommitmentare listed in Table 6-12.
Table 6-12 The Underlying Nature of Factor 11 "Organisational structure& management

r_nmmitmAnt"
v Loading Item
21 695 The size of the airline has an influence on organisational safety culture.
.
22 665 The airline's history has an influence on organisational safety culture.
.
33 437 Senior management commitment plays an important role in determining the
.
safety performance.
23 403 Airline ownership has an influenceon organisationalsafety culture.
.
12) Factor 12: Communication system

This factor derived from the analysis focused on the importance of a safety
communication system. It was of interest that the internal consistency of this factor
(refer to footnote 39), consisting of items 25,16,15 and 50, increasedfrom 54 percent
to 66 percentif item 50 was excluded.As such,in the further analysis,item 50 is
39 Pleaserefer to section 6.1.2 for the calculation of internal consistency.
184
Chapter 6 Analyses and Discussions
is excluded from the factor. As such, this factor is labelled as "Communication
system". The questionnaire items which correlated/loaded against "Communication
system" are listed in Table 6-13.
Table 6-13 The Underlying Nature of Factor 12 "Communication system"

v Loading Item
25 605 An adequatesystem exists for exchangingcritical informationregarding
.
safety problemswith other airlines.
50 -. 502 Personnel's decision-making is affected by the organisational safety culture.
15 441There is good communication between different groups in the airline.
.
16 328Changes in working procedures and their effect on safety are effectively
.
communicated to employees.
13) Factor 13: Necessity of safety reports

This factor consisted of only two items and representedthe degree to which the
safety reports were concerned.The limited number of items loading onto this factor
indicated that further study in this area might be warranted. Thus, this factor is
named "Necessity of safety reports". The questionnaire items which

correlated/loadedagainst "Necessityof safety reports" are listed in Table 6-14.
Table 6-14 The Underlying Nature of Factor 13 "Necessity of safety reports"
v Loading Item
39 817 Confidential reports should be properly de-identified in order to foster
.
organisationalsafety culture.
40 528 There should be a procedureestablishedfor acknowledgingsafety-related
.
reports.
14) Factor 14: Requires further definition
This factor derived from the analysis focused upon to which degree the risk
programmes have the influence on organisational safety culture, i. e. how safety

programmesaffect safety. It is of interest to note that this item was not groupedwith
factor 6 "Employment of risk programme" or factor 13 "Necessity of safety report".
More work is required in this area. Given the low number of variables loading onto
this factor, and the difficulty ässociatedwith its interpretation;-this factor was' not
included for further analyses.
185
Table 6-15 The Underlvina Nature of Factor 14

v Loading Item
41 701 Risk audit, risk assessment,and risk evaluationhave an influenceon
.
organisationalsafety culture.
Note-V. Questionnaireitem number,Loading:Correlationto the factor
15) Factor 15: Requires further definition
This factor was somewhatdifficult to define as it containedonly two different items.

In the case of variable 47, it represented the decision-making process regarding
safety.For the caseof variable 32, it representedthe degreeto which the role of the
safety committee was played in determining the safety performance,providing the
low number of variables loading onto this factor. This factor requires further
definition and investigation to provide a precise interpretation. As such, this factor
was deletedfrom further analysis.
Table 6-16 The Underlying Nature of Factor 15

v Loading Item
47 820 Final decisions about safety investmentare made by the Chief ExecutiveOfficer
.
(CEO).
32 392 Safety committeehas an influenceon organisationalsafety culture.
.
Note-V. Questionnaireitem number,Loading:Correlationto the factor
6.1.2 Statistical Analyses for Internal Factors
+ Reliability and consistency

To judge the internal reliability of factors, Cronbach'sAlpha statisticswere calculated
for the factors (See Appendix 0). For example, factor 1, with all items loaded onto it
demonstrated an internal consistency of 85. Similarly the other factors all
.
demonstratedacceptablelevels of internal consistency,except for factors 10,12,13
and 15 (factor 2= 846, factor 3= 89, factor 4= 83, factor 5= 82, factor 6= 74, factor
. . . . .
7= 759, factor 8= 93, factor 9= 72, factor 10= 04, factor 11= 61, factor 12= 66,
. . . . . .
factor 13=. 54, factor 15=. 33).
As mentioned in the previous sections,the low reliability generatedby factor 10 was

becauseof the expressionof the variables of 27 and 28, while the reliability of factor
186
12 rose to 0.66 when variable 50 was excluded.
In addition, factor 15, which remained undefined, demonstrated predictably low

internal consistency. Meanwhile, a reliability score could not be calculated for factor
14 as it only contained one variable. These three factors were excluded from further
analyses, leaving thirteen factors: Employee safety attitude & behaviour, Employee
safety concept, Level of operational safety in operation and maintenance, Corporate
safety policy, Personnel- quality of working life, Employment of risk programme,

Impact of accident/incidents, Financial concern, Procedures and documentation,
Commercial cost pressures, Organisational structure & management commitment,
Communication system, Necessity of safety reports.
+ Mean, Standard deviation & Scale scores

For the purpose of subsequentanalyses,mean and factor scores(Appendix P) /scales
scores(Appendix Q) were calculated for each of the scalesidentified in the Principal

Component Analysis in order to further investigate what respondents"thought to be
important and what they felt satisfied with, and in which regions.
Factor scoresare coefficients of caseson the factors, while scalescoresare the sum of
the responsesfor all items loading onto the factor, which was calculated and divided
by the number of items loading onto that factor, which is named the scale score. In
order to indicate the highest and the lowest scoresrated by the representatives,mean
of scale scoreswere applied within this section. Mean and standarddeviation of scale
scorescan be found in Appendix P.
+ Oneway Analysis of Variance (ANOVA) across regions

In order to see whether these underlying factors differed acrossthe regions, oneway
Analyses of Variance (ANOVA) were conducted for each factor across regions.
Respondentswere askedto reveal which regions they came from when completing the
survey. There were six regions in the world; namely Africa, Asia and Pacific, Latin
America and the Caribbean, Europe, Middle East and North America, as shown in
Figure 5-2 (page 174) and Table 5-1 (page 175). The complete tables of ANOVA
187
plus Mean and Standard Deviation for all 13 internal factors can be found in
Appendix Q.
The observation from these various analyses will be discussed and compared to
the empirical experiences in section 6.5.1 (page 196), when interpreting the
internal factors respectively.
6.2 Organisational Factors in External Environment
As mentioned previously, this questionnaire consisted of two parts: internal
environment and external environment. The following sections are the analyses of
external factors exerting influences on the structureof safety mechanism.
The purpose of this section is to explore the inter-relationships between external
variables on the questionnaire, and to develop factors which could be used in
subsequent analysis. Similar to section 6.1, PCA followed by a varimax rotation was
performed on the 16-item questionnaire data from 104 respondents in order to
examine the underlying structure.
The data were deemed to be suitable for the analysis (data reduction procedures), as
indicated by the Kaiser-Meyer-Olkin Measure of Sampling Adequacy value of 0.724
(Hair et al., 1995). The Bartlett Test of Sphericity was significant [x2 = 633.276, P<
0.05], indicating that correlations exist among some of the response categories (see
Table 6-17).
Table 6-17 KMO and Bartlett's Test: External factors

Kaiser-Meyer-Olkin Measure of 724
.
Sampling Adequacy.
Bartlett's Test of Sphericity Approx. Chi-Square 633.276

df 120
Sig. 000
.
188
The first analysis yielded four factors with eigenvalues greater than one, which
together accounted for 68.51 percent of the explained variance. A varimax rotation
to
was performed enhancefactor interpretability. Appendix N displays factor loadings
from the varimax rotation.
The Principal ComponentsAnalysis produced some interesting results. Each factor
will be discussedin turn. Before that, the individual items loading onto eachfactor are
outlined in the tables (from Table 6-18 to Table 6-21), which accompanythe result of
analyses.
6.2.1 Loading Factors from Principal Components Analysis (PCA)
1) Factor El: Influences of Region and Country
Factor El "Influences of region and country" dealt with the degree to which the
impact of region and country was perceived to have on organisationalsafety culture.
Included within this factor were regional economic, regional geography,regional
culture, regional religion and country economics.Interestingly, regional and country

influences were originally designedas two separatedfactors. This finding revealed
that they were seen as one concern (factor) from managers' point of view, and
regional economic influence attracted more concerns than country economic
influence. The questionnaire items which correlated/loadedagainst "Influences of
region and country" are listed in Table 6-18.
Table 6-18 The Underivina Nature of Factor El "Influence of region and country"
v Loading Item
70 879 Regionaleconomicinfluence and organisationalsafety are correlated.

.
69 857 Regional geographical influence and organisational safety culture are correlated
.
72 804 Regional cultural influence and organisational safety are correlated.
.
68 715 The country economic influence and organisational safety culture are correlated
.
71 685 Regional religion influence and organisational safety culture are correlated.
.
2) Factor E2: Public and the Media Influence

Factor E2 "Public and the media influence" focused on the company's relationship
189
with the public and the media, and to what extent organisational safety culture was
influenced. The items loaded onto this factor and were all indicative of the media
and public influences; customers' and consumers'perspectives were included. The
questionnaire items, which correlated/loaded against "Public and the media

influence" are listed in Table 6-19.
Table 6-19 The Underlying Nature of Factor E2 "Public and the media influence"
v Loading Item
61 881 The customer's reaction has influences on organisational safety culture.
.
63 777 The perceived corporate image has influences on organisational safety culture.
.
62 757 Consumer habits have influences on organisational safety culture.
.
60 546 The relationship with the media has influences on organisational safety culture.
.
3) Factor E 3: Impact of the Regulatory Environment
This factor dealt with the degree to which organisational safety culture was
influenced by industry regulation/regulators.Included within this factor were the
to,
adherence and influence of, safety authorities. As such, this factor is labelled as
"Impact of the regulatory environment" which shows that safety authority has been
regarded as one of the most influential and important factors of the safety
mechanism.The questionnaireitems, which correlated/loadedagainst"Impact of the
regulatory environment", are listed in Table 6-20.
Table 6-20 The Underlying Nature of Factor E3 "Impact of the regulatory environment"
v Loading Item
58 809 Safety informationfrom aviationsafety authoritiesis highly valued.
.
756 The regulations from the aviation safety authority have an influence on
59
. organisational safety culture.
750 The recommendations and suggestions from the industry safety committee are
57
. adhered to all the time.
4) Factor E 4: Involvement of Investment Community

Factor E4 "Involvement of investmentcommunity" was concernedwith the degreeto
safety culture was influenced by the investment community. It
which organisational
is unclear why the last item was included in this factor. Yet since the number of
190
loading is so low, this item is deleted in the further analysis. The questionnaire items,
which correlated/loadedagainst "Involvement of investment community", are listed

in Table 6-21.
Table 6-21 The Underlying Nature of Factor E4 "Involvement of Investment community"

v Loading Item
65 797 The investors' perspective has influences on organisational safety culture.
.
66 788 The stock market's reaction has influences on organisational safety culture.
.
64 555 The organisation's investments in other businesses are correlated with the
.
organisational safety.
67 219 The country culture and organisationalculture are correlated.
.
6.2.2 Statistical Analyses for External Factors
+ Reliability and consistency

Cronbach'sAlpha statistics were calculatedto assessthe internal consistencyof these
factors. All of the factors were shown to be reliable (Factor E1= .8575; Factor
E2= 7854; Factor E3= 686; Factor E4= 636), as shown in Appendix 0. As such,
. . .
thesefour external factors are: Influences of region and country, Public and the media
influence, Impact of regulatory environment, and Involvement of investment
community.
+ Mean, Standard deviation & Scale scores

For the purposes of subsequentanalyses, factor scores (Appendix P) /scale scores
(Appendix Q) were calculated for each factor identified in the Principal Component
Analyses (PCA) in order to further investigate what respondents thought to be
important and what they felt satisfied with, and in which regions. The same method
was used as in the internal factor structure. In order to arrive at the best interpretation,
both scale scores and factors scoreswere calculated. However, mean of scale scores
were applied within this section becausethey indicated the highest and the lowest
scoresrated by the Mean
representatives. and standard deviation of scale scorescan
be found in Appendix Q.
191
+ Oneway Analysis of Variance (ANOVA) across regions

In order to see whether these underlying dimensions differed across the regions,
oneway Analyses of Variance (ANOVA) were conducted for each questionnaire scale
across six regions: Africa, Asia and Pacific, Central/South America, Europe, Middle
East, North America. Appendix Q lists the table ofANOVA for all 4 external factors.
The observations from these various analyses will be discussed and compared to
the empirical experiences in section 6.5.2 (page 206), when interpreting the
external factors respectively.
6.3 Relationship between Organisational Factors and Safety

Performance
Looking back to the hypothesisedmodel illustrated in Figure 4-8 (page 157), both
internal and external factors (input) exert influence on safety mechanism,which will
manifest on the safety performance (output). Therefore, this analysis aimed to explore
the relationship between the organisational factors (both internal and external)
identified in this chapter and safety performance through the survey. As it is not
possible to identify the safety performance from the anonymous survey, self-rated
is
safety performance therefore adopted. It was thought that an examination of the
factors identified in the study were the biggest predictors of self-rated (perceived)
i.
safety performance, e. what organisationalfactors contributed to the explanation of
the perceivedsafety performance,and to what degree,would provide someinteresting
insights into the development of a proactive safety mechanism. Hence, a stepwise
multiple regressionprocedure was used. This part of the researchwas intended only
as a guide to indicate which factors were the best predictors of perceived safety
performance.
By applying the standardmultiple regressionanalysis,the dependentvariable was the
192
Chapter 6Analyses and Discussions
self-rated safety performance, and the independent variables were the organisational
factors - the influential dimensions of the safety mechanism; the latter were assessed
in order to forecast the former. Appendix R shows that the result of the multiple
regression and variance accounted for by the factor scores in the regression equation
was significant (F4,99 = 28.822, P<0.05). It indicates that there is a relationship

between self-rated safety performance and organisational factors. And without
the existence of external factors, internal factor 2 "Employee safety concept", factor
12 "Communication system", factor 7 "Impact of accidents/incidents" and factor 3
"Level of operational safety in operation and maintenance" are the best predictors of
self-rated safety performance among these organisational factors. If expressed by a

linear equation, they have the form, according to Appendix R:
Y1= 486+ 403 * (factor 2) +. 244* (factor 3) +. 113* (factor 7) +. 133* (factor 12)
. .
Where
Yl is self-rated safetyperformance
Factor 2 is Employee safety concept
Factor 3 is Level of organisational safety in operation and maintenance
Factor 7 is Impact of accidents/incidents
Factor 12 is Communication system
It is worth noting that in the equation, the regression coefficients (or B coefficients)
representthe independentcontributions of each independentvariable (factors) to the

prediction of the dependentvariable (self-rated performance), i. e. for example, factor
2 "Employee safety concept" is'correlated with the Yl (self-rated safety performance)
after controlling for all other factors (factor3, factor 7 and factor 12). The direction of
the correlation coefficients in the equation also indicates that higher factor scoresare
associatedwith better rated perceived safety performance.
193
6.4 Relationship between Internal and External Factors
One of the four hypothesesof this thesis is that the safety mechanismis a multiple and
complex construct. This has been demonstratedthrough the case study included in
Chapter 5 and will be further explored here in order to shed some additional light on
this hypothesis.
As such, the following analysis was carried out to explore the relationship between
the internal and external factors that were identified to have influence on a proactive
safety mechanismin the previous sections,given that in Table 6-1, the Bartlett Test of
Sphericity, indicates that correlations exist between some factors. For this purpose,
correlations were calculated for each internal factor against each external factor.
Appendix S shows the correlation and significant relationship between internal and
external factors, including:
Correlation between El "Influences of region and country" and internal

factors:
As shown in Appendix S, external factor El "Influences of region and country"

demonstratedquite low correlations with all thirteen of the internal factors. The
highest correlation (r=.216) was observedbetween this factor and internal factor
11 "Organisational structure and managementcommitment". A moderate and
significant correlation (r=. 198) was also observedbetweenthis factor and internal
factor 10 "Commercial costpressures".
Correlation between E2 "Public and the media influence" and internal

factors:
Factor E2 "Public and the media influence" is found to be most correlated with
internal factor 1 "Employee safety attitude and behaviour", factor 2 "Employee
safety concept", factor 9 "Procedures and documentation", and factor 11

"Organisational structure and managementcommitment", shown in Appendix S.
The highest correlations were observed between this factor and internal factors
194
relating to "Organisational structure and management commitment" (r=. 367) and

"Employee safety concepts" (r=. 272). Further significant correlations were
observed between this factor and internal. factors relating to "Procedures and
documentation" (r=. 244), and "Employee safety attitude and behaviours" (r=. 216).
ºý Correlation between E3 "Impact of regulatory environment" and internal

factors:
This factor E3 "Impact of regulatory environment" was significantly correlated
with all internal factors. Among the internal factors, the highest correlation was
observed between this factor and internal factor 3 "Level of operational safety in
operation and maintenance" (r=. 455).
Correlation between E4 "Involvement of investment community" and

internal factors:
Factor E4 "Involvement of investmentcommunity" demonstrateslow correlations
with all thirteen of the internal factors. However, this factor was found to be
highly correlated with "Commercial cost pressures" (r=.226), and "Financial
concern" (r=. 198). Meanwhile, significant correlations were observed between
this factor and those factors relating to "Corporate säfety policy" (r=.235), and
"Impact ofAccident/incidents" (r=.219).
The implication of the results will be interpreted together with the discussion of
each factor respectively in the following section.
6.5 Interpretation of Individual Factors
With the confirmation of the existence of organisational factors in the safety
mechanism, a greater understandingof the forces (input) is provided at work within

the internal and external working environment of the airline industry as well as the
195
relationship to safety performance (output). Table 6-22 lists the summaries of
statistical analysesfrom the preceding The

analyses. following sectionswill carry out
the in-depth discussion of the result findings.
6.5.1 Discussion of Internal Factors
+ Factor 1: Employee safety attitudes and behaviour
The first factor derived within the internal environment is "Employee safety attitude
and behaviour". Thus, the importance of employee competence regarding safety is
emphasised.Among the skills and abilities (loading items) included within these
factors are: employees' ability to comply with safety procedures,safety knowledge,
safety participation, and safety motivation, which echoes the existing literature on
safety culture explored in Chapters2 and 4.
Furthermore, it is interesting to note that this factor did not predict perceived safety
i.
performance, e. individuals who rated the "Employee safety attitude and behaviour"
highly on the questionnaire did not rate the Perceived safety performance more
highly.
Additionally, the table of means indicated that means of factor 1 "Employee safety
attitude and behaviour" scores were in the range of 4.2- 4.9. Safety managers in
Africa reported the perceived factor 1 "Employee safety attitude and behaviour" more
frequently than other regions within the airline industry, which also demonstrated an
awareness of such concerns has more effect in this region compared to others.
A significant difference was observedon this factor acrossdifferent regions (F5,98

2.997, P<0.05). When the effects are significant, the meansmust then be examined in
order to determine the nature of the effects. There are procedures called "post hoc
tests" to assist to perform this task. One of the post hoc multiple comparisonsused
was Least-Significant Difference (LSD). It indicated that this was attributable to a
significance between respondents in North America compared to Europe as well as

Asia and Pacific (p<.05), i. e. the mean difference was significant at the .05 level,
which meant that there was variability of the opinions (rating scoresof questionsin
196
Chapter 6 Anah'ses and Discussions
W
S 4 4 4
S." ED
W \ c
\ \ \ \ \ \ \ \ \ ::
7 7 T Oc
.0
6
0 O 0
W U U U U
1)
.yo
CA U
w w w w
0) 0) 0) 0
44 4 z
ice. -
0 Z.
w w w w w wC wC wý+ wý+ p Q w w
b
6) b
iy yq+p
U
P. L14 13. ol W
rrh: rr---II ..--JJ

W:4ü
U
,e w w V
b
w
U
v
rý
Q,
i."
m cccl
U
E Q
rf V
ma
N
6r
c 0)
CC o) C
O ° O ü
(0
4 7 ü
U
O c
t ?i C
ý: a (0 C
m U) i O
(1)
41
E
C E C
U E E
O
2 O
a O m a O U ca
O E CO f17 O D
C (ß O ca
Ia (4 a1 fE U = C
.t
Q ö Ql
O O O
0)
D a) M ° _0 E w
CL E
L) ö_ °ý v
c ° - ä
= <n c aa)
C6 0
U
Q °
O
° > 0 4) 0 7 a.
0 A " -- O ý6 -1 c 75
z ;M- 13 a U a)
° = U
G.' U) Ul
U s C
U
U
0
Od O -o C w
y UU) (E y ,
OU QC m U 2
O CD O E
o c E ö U) °
°
ö
m
ö ö
o N ° U C) ü.
° v a c U)
ö o Q1 E E n a v¢
W Q m a co cn ° o
Ü 0 U = i
E E , EW ÖU Ü z E 0
w W E _ LL d°
> _J U L
0 N m
N [n v Ln co r` co cyý W W W W
J
197
the survey) towards employee safety attitudes and behaviour within these three
regions. Generally speaking, the higher the rating scores, the more satisfaction there
was towards employee safety attitudes and behaviour, and the more importance and
attentions of safety culture were placed in the company. Further referring to the table
of mean (Appendix P), the mean scores of this factor in six regions showed that the
lowest ratings on this factor originated from the Middle East while the highest ratings
came from Africa. This seems to contradict industry indicates: Middle East airlines
have better airline publicity and aircraft performance than Africa airlines which tend
to have poor publicity, poor accident record and lack of regulatory framework. The
contradiction will be further explained when discussing the bias of the questionnaire
responses.
A significant correlation was also observed between factor 1 "Employee safety
attitude and behaviour" and external factor E2 "Public and media influence" and
external factor E3 "Impact of regulatory environment", in particular E3. As to E2,
since public and media influence involve company reputation management, it is

therefore more related to employee safety knowledge and safety motivation, which
constitute this factor. Thus, a higher level of understanding and activity on the part of
the regulatory environment and the public and media are associated with employee
safety attitudes and behaviour which favours the development of a proactive safety
mechanism.
+ Factor 2: Employee safety concept
The safety concept factor included employees' beliefs, shared perceptions and
organisational The
atmosphere. presenceof this factor is consistentwith the literature
which describedthe contribution of a safety climate to the organisation,in Chapter4.
No significant differences were observed for this factor (F5,98= 2.123, P>0.05). It
showedthe degreeof perceived employees'safety conceptsrated by respondentsat all

regions were similar to each other. Furthermore, an examination of the mean scores
for each group on this factor revealedthat higher scorestendedto be rated (except for
Latin America' and the Caribbean, and the Middle East), which showed the highest
198
Chapter6 AnalysesandDiscussions
satisfaction of the perceived safety concept, and the importance of a safety climate
was felt by the respondents. Across the regions, scale scores reveal the highest ratings
,
(higher satisfaction) were observed in Africa, while the lowest were in the Middle
East (lower satisfaction).
Factor scoresfrom the multiple regressionanalysis (Appendix R) show this factor to

be a significant predictor of perceived safety performance. It meant that those who
rated the employee safety concept highly on the questionnaire rated the perceived
safety performance more highly. Meanwhile, the direction of the correlation
coefficients indicates that higher factor scores are associated with better rated
perceived safety performance.
A significant correlation was observed between factor 2 "Employee safety concept"
and external factor E2 "Public and media influence", and external factor E3 "Impact
of regulatory environment", shown in Appendix S. In other words, employee safety
concept is likely to be more influential on a proactive safety mechanism when the

media and public and the regulatory environment becomemore active and involved.
ºý Factor 3: Level of operational safety in operation and maintenance
Operation and maintenanceissueswere highlighted since this factor is concernedwith

the degree to which individuals are empowered to deal with operations and
maintenanceissues,especially regarding training. In Chapters2 and 4, the importance
of theseitems was identified asbeing critical to the continual improvement of safety.
The oneway ANOVA for this factor approached, but failed to reach statistical
significance (F5,98 = .647, P>0.05). Therefore,-it could be concluded that managers

working in all regions displayed a similar degree of concern for operation and
maintenanceaspects.
Factor scores from the multiple regression analysis show that this factor is a
significant predictor of perceived safety performance (Appendix R) within the airline

industry, i. e. individuals who rated this factor highly on the questionnairerated the
199
Perceivedsafety performance more highly. It also indicates that higher factor scores
are associated with better rated perceived safety performance.
A significant correlation was observed between factor 3 "Level of operational safety
in operation and maintenance" and external factor E3 "Impact of regulatory
environment", shown in Appendix S. This finding reveals that the influence of
regulators and regulations for safety is associated with the level of operational safety
in operation and maintenance to a significant degree. With African region reveals the
highest mean score of this factor, it seems to conflict with the conclusion in previous
paragraph and industry findings since African regulatory regimes are largely regarded
as ineffective. Again, similar to factor 1 "Employee safety attitude and behaviour", the
bias will be explained in section 6.7 and 6.8.
+ Factor 4: Corporate safety policy
The safety policy within airlines was representedby a factor of its own which
includes items relating to clear statedpolicy towards goals, objectives and approaches
to safety.A significant difference (F5,98= 2.363, P<0.05) was observedon this factor
between groups of respondentsrepresentingdifferent regions, which described the
degree to which the airline focused on and practiced its safety policy. In a similar
manner to factor 1"Employee safety attitude and behaviour", the post hoc multiple
comparison - Least-Significant Difference (LSD) indicated that region of North
America differed from the regions of Europe and Asia and Pacific (P<.05), indicating
the variability which exists within theseregions. Meanwhile an examination of means
revealed that Africa perceived this aspect to be more satisfactory in their working
environment, and regarded this factor to be more important than those who were in'
other regions. The lowest ratings on this factor originated from the Middle East (refer
to previous factors for the African issue).
A significant correlation was observed between factor 4 and external factor E4
"Involvement of investment community", and external factor E3 "Impact of regulatory
environment", presented in Appendix S. It indicates that the investment community is

likely to be more associated with the organisations which value safety policy and/or
200
post event impact in order to develop a proactive safety mechanism.
+ Factor 5: Personnel - quality of working life
Factor 5 "Personnel - quality of working life" included items relating to personnel
attitudes, morale, working relationships, and frustration accommodation. No
significant differences were observed between the responses from various regions on
this factor (F5,98 =. 111, P>0.05). Thus, there is no need to conduct the post hoc
multiple comparisons because groups of respondents representing all regions reported
similar attention to their companies' personnel working life quality.
Given the similar degree across regions, two points were observed from the
Mean/Standard Deviation Table for factor 5 "Personnel - quality of working life".

One was that the scores rated by all representatives were relatively low compared with
other factors, which showed less satisfaction with the quality of employee working
life, because no significant difference was observed on this factor across different
regions40 and the observation of lower ratings is shown in the table of mean. The
other was that the highest rating on this factor originates from Asia and Pacific, while
the lowest one is from the Middle East. As such, Asia Pacific representatives had a
better acknowledgement of this factor, given the low tendency of mean scores of other
regimes.
A significant correlation was observed between factor 5 "Personnel - quality of
working life" and external factor E3 "Impact of regulatory environment", shown in

Appendix S. As such, regulators are likely to take greater stepstowards encouraging
proactive safety in the organisations in which employees' working quality support
such action.
ao Notwithstanding the lack of regional variations, there is likely to be variations between the
different workforce group, e.g. pilots, engineers, cabin crew, etc.
201
+ Factor 6: Employment of risk programmes
This factor was representedby three variables relating to the importance of data
analysis systems, safety information and hazard identification systems to safety

culture, identified in Chapter 2. Respondents
were thought to have a similar degree of
feeling towards the employment and benefit of risk programmes because no
significant differences were observed (F5,98 = 693, P>0.05) between regions on this
.
factor across different regions. As such, representativesfrom various regions have a
similar degreeof view regarding airline risk programmes.
A significant correlation was observed between factor 6 "Employment of risk
programmes" and external factor E3 "Impact of regulatory environment. It indicated
that regulators are likely to encourage proactive safety in the organisations in which
the employment of risk programme helps to support such action.
+ Factor 7: Impact of accidents/incidents
This factor includes the existenceof emergencyplans, actions taken in the aftermath
of accidentsand incidents, and businesscontinuity plans. In Chapter 2, the importance
of theseitems was identified as being critical to safety improvement.
The oneway ANOVA across regions on factor 7 "Impact of accident/incidents",

dealing with the management'sview of the influence of incidents/accidents,did not
demonstrateany statistical significance (FS,98= 598, P>0.05). Therefore respondents
.
acrossdifferent regions did not vary appreciably in their views of the actions that the
companytook to deal with the aftermathof incidents/accidents.
However, Mean/StandardDeviation Table of factor 7 showed standarddeviations of

five regions were over one, showing that there was variability within all regions
except Latin America and the Caribbean.This interesting observationmight be caused

by that not all the airlines were used to have accidents, which had the responses
deviatedwithin regions.
202
Multiple Regression analysis shows that this factor is a significant predictor of
perceived safety performance (Appendix R) within the airline industry. It indicates

that there exists a positive relationship between "Impact of accident/incident" and
Perceived safety performance, i. e. individuals who had high rating scores of the
former was likely to rate the latter with high scores.
A significant correlation - was also observed between factor 7 "Impact of
accident/incident" and external factor E3 "Impact of regulatory environment" and

external factor E4 "Involvement of investment community", demonstrated in Appendix
S.
ºý Factor 8: Financial concern
Factor 8 "Financial concern" is the one split from commercial pressures. It

-is
representedby two variables relating to conflict between safety and financial goals
and shareholders'welfare. ANOVA of factor 8 "Financial concern" did not reveal any
significant difference across regions (F5,98= .675, P>0.05). Given this fact plus the
mean scores were lower, it implies that representativesrated this factor -Financial
concern, to a similar degree, which indicates that financial concerns did influence
safety goals,but not by much, comparedto other factors. .
Five standard deviations within regions were observed to be higher than one
(Appendix Q), showing the potential differences of concerns towards factor 8 within
individual regions. Secondly, the managers in North America demonstrated the
highest concern in this regard compared with managers in other regions.
+ Factor 9: Procedures and documentation
Factor 9 "Procedures and documentation" included items relating to documented
responsibilities, written work procedures and written monitor procedures.

Respondents across the regions had a similar degree of opinions towards procedures
and documentation in their airlines since no significant difference was observed on
this factor across the different regions (F5,98 = 384, P>0.05). As such, respondents
.
203
across different regions displayed a similar degree of concern regarding factor 9

"Procedures and documentation" with their airlines. An examination of the mean:
scores for each group on this factor (Appendix OJ revealed that managers in all
regions had their companies' documentation process rated with a similar frequency of
4.3. The highest rating on this factor originates from Africa, while the lowest one is
from Latin America and the Caribbean. The contrary to the empirical experiences will
be explained in later sections.
+ Factor 10: Commercial cost pressures
This factor derived from items relating to managementconcerns,safety budget, value
of management,and trade-off between cost and safety, demonstratedin Chapter 4.

Similar to factor 8 "Financial concerns", oneway ANOVA for regions approached,
but failed to reach statistical significance (FS, 98 = 415, P>0.05). Therefore,
.
respondents across different regions did not vary appreciably in their views of
commercial cost pressureson safetyperformance.
An examination of the table of means (Appendix Q) revealed the highest rating on

this factor originated from Asia and Pacific, while the lowest one was from the Middle
East. The lower mean scores suggest that respondentsacross the regions had low
appreciation of the influence of this factor upon safety. It reflected the fact that
commercial cost pressureswere not significantly perceived by safety mangersacross
regions, especiallyMiddle East.
"Commercial cost pressures" was significantly correlated to El "Influence of region
and country", E3 "Impact of regulatory environment" and E4 "Involvement of

investment community" (Appendix S). This shows that commercial cost pressures
exert a significant impact on promoting proactive safety with the influence of region
and country, regulatory environment and investment community. The smaller
correlations indicated the decreasing impact actions.
204
+ Factor 11: Organisational structure and management commitment
This factor represented by five. items relating to the size and history of airline,
leadership, senior management commitment and airline ownership. Although no
significant difference (F5,98 = .981,. P>0.05) was observed on this factor across the
different regions, the table of mean (Appendix Q) shows the high rating tendency of
respondents, which means that the importance of organisational structure and
management commitment was highly appreciated by representatives across regions to
a similar degree, as factor 11 "Organisational structure and management
commitment" indicated the importance of understanding and involvement of
organisational structure and management commitment in the company. The highest
rating on this factor originated from Africa, while the lowest one was from the Middle
East and Central/ South America.
Factor 12: Communication system
This factor included the exchanging safety information system, personnel's decision-
making, group communication, and employee communication. No significant

difference was observed on this factor across different regions (F5,98=, 1.313, P>0.05).
By this result, respondents across different regions did not vary appreciably in their
views of the communication system. Moreover representatives tended to have a low
rating across regions on this factor, showing that similar degree of view of airline
representatives was less satisfaction towards the individual airlines and

Communication system is seen as less effective.
In addition, factor 12 "Communication system" is a significant predictor of perceived
safety performance within the airline industry (Appendix R). This relationship not
only indicates the positive correlation between these two factors, but also the
contribution of the communication system to perceived safety performance. Across

the regions, the highest ratings were observed in Africa, while the lowest were in
Latin America and the Caribbean. A significant correlation was also observed between
factor 12 "Communication system" and external factor E3 "Impact of regulatory
environment", indicating the positive relationship between these two factors.
205
+ Factor 13: Necessity of safety reports
Factor 13 "Necessity of safety reports" included two items relating to confidential
report identification, and procedures for safety reports, described in Chapter 2. The
oneway ANOVA across regions on this factor approached, but failed to reach
statistical significance (F5,98= 565, P>0.05), i. e. respondents across the regions were
.
deemed to have a similar degree of concern, which tended to be high on this factor.
Thus, respondents across different regions did not vary dramatically in view of the
safety reports, and the representative across regions all placed importance on the
necessity of safety reports. This can be explained, for example, although various
European JAR requirements for mandatory operating on maintenance and operational
issues, the necessity and importance of safety reports are never been neglected and
still been highly concerned.
The highest rating on this factor originated from North America, while the lowest one
was from Asia and Pacific. A significant correlation was observedbetween factor 13
"Necessity of safety reports" and external factor E3 "Impact of regulatory
environment", shown in Appendix S. Again, it showed that the necessity of safety

reports and the impact of regulatory environmentwere positively correlated.
Within the interpretation of internal factors, it showed some findings are
contrary to the industry experiences. The contradiction will be further
interpreted in section 6.7 (page 214).
6.5.2 Discussion of External Factors
+ Factor El: Influences of region and country
Factor El "Influence of region and country" included items relating to economic,

geographic, cultural and religious influence of region and country. No significant
difference was observed on this factor across the different regions (F5,98 = .669,
P>0.05), so respondentswere seen to regard the impact of region and country to a
similar degree and they had a common feeling of less impact of this factor, shown by
206
the lower mean scores.
Factor El "Influence of region and country" demonstratedquite low correlationswith
all thirteen of the internal factors as shown in Appendix S. As such, one can conclude
that the influence of region and country does exert an influence on the safety
mechanism but it was seen to play a relatively minor role in the development of a
proactive mechanism. However, a number of moderate correlations were observed
which warrant somefurther discussionat this point.
The highest correlation was observed between this factor and internal factor 11
"Organisational structure and management commitment". This finding indicates that
a higher level of influence of region and country is associated with organisational
structure and management commitment towards developing proactive safety

mechanism to a significant degree. It also suggests that regions and countries are
likely to exert an influence on encouraging proactive safety in the organisations in
which structure and commitment support such action.
A moderate and significant correlation was also observedbetween El "Influence of
region and country" and internal factor 10 "Commercial cost pressures", which
demonstratesthe impact of commercial pressures.It shows that the degreeto which
region and country exert an influence to establish a proactive safety mechanism is

associatedto a significant degreewith commercial cost pressures.
+ Factor E2: Relationship with the public and the media
Factor E2 "Relationship with the public and the media" was representedby four
variables relating to customer relations, perceived corporate image, customer habits

and media relationships. Representativesacrossthe regions were thought to have no
different degree of concerns regarding this factor becauseno significant difference
was observedon this factor acrossdifferent regions (F5,98= .909, P>0.05). Therefore,
respondentsacrossdifferent regions did not vary appreciably in their views of how the
public and the media influence organisational safety. Compared to El "Influence of
region and country", the mean score of E2 "Relationship with the public and the
207
in various regions was much higher, indicating that the public and media were
media"
deemed as an importantly influential factor.
A significant correlation was observed between E2 "Relationship with the public and
the media" and internal factor 11 "Organisational structure and management

commitment", factor 2 "Employee safety concept", factor 9 "Procedures and
documentation" and factor 1 "Employee safety attitude and behaviour", shown in
Appendix S. The highest correlations were observed between this factor and internal
factors relating to "Organisational structure and management commitment" and

"Employee safety concept". As such, a higher level of understanding and activity on
the part of the public and media is associated with an organisational structure and
management commitment and employee safety concept which favours the
development of a proactive safety mechanism.
Further significant correlations were observed between E2 "Relationship with the
public and the media" and internal factors relating to "Procedures and
documentation", and "Employee safety behaviours". Accordingly, a better
appreciation of procedures and documentation with respect to safety and

well-behaved employeesafety behavioursis likely to involve more positive relation to
influence media and public or receive its influence, which in turn affect a proactive
safety mechanism.
+ Factor E3: Impact of regulatory environment
This factor was representeda factor of its own, which included items relating to: value
of safety authority information, the influence of regulations and the adherenceof

industry safety committee suggestions.The oneway ANOVA for region on Factor E3
"Impact of regulatory environment", dealing with safety authority's influence on
organisational safety demonstrated statistical significance (F5,98 = 2.435, P<0.05).
Post hoc multiple comparisons - Least-Significant Difference (LSD) indicated that

this was attributable to a significance between respondents in North America
to Africa as well as Asia and Pacific (p<.05), i. e. the mean difference is

compared
significant at the 05 level.
.
208
The examination of table of means revealed the highest rating on this factor originated
from Africa, while the lowest one was from North America. However, the empirical
experiences show North America is heavily regulated, perhaps too much, while Africa
less so with safety information promulgated by regulators. The Africa result may
indicate a dependence upon regulatory impact due to inadequacies in the company
management structure. Yet there is still a need to further investigate the impact of bias
of questionnaire in subsequent sections.
Additionally, a significant correlation was observed between E3 "Impact of regulatory
environment", and all internal factors. (Appendix S), showing the close relationship
between airlines and regulatory environment. This finding is of importance, and it
indicates that a higher level of understanding and activity on the part of the regulatory
environment is associated with all internal factors which favour the development of a
proactive safety mechanism. It demonstrates that the role of the regulator and
regulation were deemed to be important to the internal organisational factors for
promoting proactive safety.
Among the internal factors, the highest correlation was observedbetween this factor
and internal factor 3 "Level of operational safety in operation and This

maintenance"..
finding reveals that the influence of regulators and regulations for safety is associated
with the level of operational safety in operation and maintenance to a significant

degree4'.Regulators are likely to take greater steps towards encouraging proactive
safety in the organisationsin which operation and maintenancesupport such action.
+ Factor E 4: The influence of investment community
Factor E4 "The influence of investment community" included items relating to

investors' prospective, stock market's reaction, and business investment. No
significant difference was observedon this factor acrossdifferent regions (F5,98

= .910,
P>0.05). Thus, respondentsacross different regions did not vary appreciably in their
view of how investment influenced organisationalsafety.
41 It is interesting to note that despitethe regulatory environment and the findings here it is
non-compliancewith regulatory requirementsthat after leads to incidents/accidents.
209
Also, table of mean/standard deviation showed relatively lower mean scores,
comparedto other factors. Most of the standarddeviations were greater than one or
to
close one, indicating the variance within the individual regions. The highest rating
on this factor originated from Latin America and the Caribbean,while the lowest one
was from Africa.
According to Appendix S, factor E4 shows low correlations with all thirteen of the
internal factors. A significant correlation was observedbetween E4 "The influence of
investmentcommunity" and "Commercial cost pressures", and "Financial concern".
As such, the degree to which the investment community's influence on establishing
is
proactive safety mechanism associatedwith organisations' financial issues to a
significant degree.
Meanwhile, significant correlations were observed between E4 "The influence of

investmentcommunity" and those factors relating to "Corporate safety policy", and
"Impact of Accident/incidents". It indicatesthat the investment community is likely to
be more associatedwith the organisationswhich value safety policy and/or post event
impact in order to develop a proactive safety mechanism. It reflects a tendency of
investors to continue investing if an airline does not have an accident or bad press
imperative of its safety policy.
In summary,the evidence from these findings support the secondhypothesis of this

thesis which statesthat the safety mechanismwill be demonstratedas a complex and
multidimensional construct, which will be influenced by a set of factor-structure.
Hypothesis three, stating that the organisational factors mediate the relationship
betweenthe safety mechanismand safety performance,and act as predictors of safety
performance,is also supportedby the findings.
210
6.6 Comparison with the Hypothesised Factor Structure
The results of survey reveal a slightly different factor structure in comparisonto that
of the hypothesisedmodel (see Table 6-23). In terms of internal underlying factors,

the factors derived from the internal environment confirm the general structure of the
safety mechanism model with few changes,when compared with the hypothesised
internal factor structure.Thesechangesare listed below:
Factor 1- "Perceived safety" in the hypothesisedmodel was separatedinto two

internal factors: "Employee safety attitude and behaviour" and "Employee safety
concept" in the survey results and the latter is the predictor of perceived safety
performance. This finding is of interest because, as mentioned in the previous
section, safety behaviour is related to safety culture, while safety concept is more
related to safety climate. It therefore indirectly proves the construct of the safety
mechanism and would be able to provide the airlines with exact influential factors
to promote proactive safety.
Factor 8- "Personnel communication and relationship" was split into two factors
in the survey results- "Personnel-quality of working life" and "Communication
system". It shows that to employees, good working relationship in the company
manifest itself on the quality of work life, including morale, well defined job
function, and accommodation of frustration. These are what personnel deem to be
as important in the working environment and have influence on proactive safety.

Meanwhile, the appearance of "Communication system" indicates the importance
of the organisational communication system, especially when the issue is related
to proactive safety and examined as one of the predictors of perceived safety
performance.
+ Factor 6- "Safety information" was reconstructed to create a new factor-

"Necessity of safety report". This result reveals that within airline safety
information systems,safety reports attract higher attention and are thought of as
one of the influential factors on a safety mechanism, and as one of the predictors
211
of perceived safety performance.
*ý Items in Factor 4- "Management control (or quality control)" were not grouped
together; instead most of them were included in "Organisational structure and
management commitment". As such, "Management control" was dismissed and

replaced by "Organisational structure and management commitment" to cover the
managerial issues, which echoes to Chapter 2 when describing the important of

these managerial issues on airline safety.
*3' Factor 10- "Decision-making style and process" was out of the factor structure
becausethe results did not reveal the characteristicof this factor. Similarly, a new
factor emergedand grouped the financial issue items together. This new factor
was therefore named "Financial concern". Again, it showed the degreeto which
the survey respondentsregardedtheseconcernsimportant to the developmentof a
safety mechanism.
In terms of external factors, in total the results of the survey identified four factors in
the external environment. Comparedwith the original ones proposed (five in total),
country influence and regional influence were combined into one factor. This finding,
togetherwith the fact that the scoresrating of this factor was lower than that of others
pointed to the fact that influences of region and country were not exerting aspowerful
an influence on proactive safety mechanismsas was initially proposed.As to the other
three factors, the survey results revealed the same factor structure. However, the
factor names were changed slightly in order to describe their characteristicsmore
accurately.
212
Uhrar, I (, I,... ý .,, r.l /)I cussiuit. c
Table 6-23 Comparison of the Factor Structure
.1
-A. _----- i Survey result
1. Perceived safety 1. Employee safety attitude& behaviour
2. Operation and maintenance 2. Employee safety concept
3. Risk management (control) 3. Level of operational safety in

4. Management control (quality control) operation and maintenance
5. Commercial pressures 4. Corporate safety policy
6. Safety information (info technology) 5. Personnel- quality of working life
7. Organisational structure 6. Employment of risk programmes
c
Z 8. Personnel communication and 7. Impact of accident/incidents
oý
ZZI relationships 8. Financial concern
9. Post-incident/ accident (impact of 9. Procedures and documentation
incident/accident) 10. Commercial cost pressures
10. Decision making style and process 11. Organisational structure and
11. Corporate safety policy management commitment
12. Documentation 12. Communication system
13. Necessity of safety reports
1. Industry regulations 1. Influences of region and country

2. Public relationships 2. Public and the media influence
3. The investment community 3. Impact of regulatory environment
oý
w 4. Country influences 4. Involvement of investment
5. Regional influences community
213
6.7 Discussion of the Bias in the Result
The conclusions outlined in the previous section support the -perspectives which
underline the safety mechanism in Chapter 4. This section aims to identify the ways in
which the conclusions regarding organisational factors expand the understanding of

the development of a proactive safety mechanism. As such, some consideration of the
interpretation of the data and some discussion of where the results fit within the
existing literature are in order.
There are two points worth noting as the bias the results when applying results from
this study.
1. The uneven representation of respondents across different regions:
The respondentsfrom different regions include twenty-two from North America,

three from Africa, thirty-six from Europe,thirty-six from Asia and Pacific, five from
Latin America and the Caribbeanand two from the Middle East (Figure 5-2, page
174).
.
Two points need to be addressed here. Firstly, according to Figure 5-2, the
proportion of returned questionnaires from Africa and the Middle East were much
lower than the intended sample rate. Secondly, the low number of representatives
from Africa and the Middle East in the current sample revealed that higher scores
were rated by African representatives, while lower scores were rated by
representatives from the Middle East on most of the factors, which may limit the
power of the analyses, especially when discussing these three factors with
significant differences across regions. As such, there is a need to review the results
of factors 1 "Employee safety attitude and behaviour", factor 4 "Corporate safety
policy" and factor E3 "Impact of regulatory environment", which will be discussed
in next section - Implication for the proactive safety mechanism model.
2. Two undefined factors: internal factor 14 and 15:
These two factors remained undefined and were not subjectedto further analyses.
214
Both factors contained items which, although only moderately related to other items
within the questionnaire, still point to elements which were thought to be important
to the development of a proactive safety mechanism. The fact that they were not
included in the other thirteen internal factors would demonstrate that they account
-
for some degree of variance which was not accounted for by other factors. Therefore,
some thought is required before they should be dismissed as unimportant, and. it
may be advisable to expand upon these factors in future research.
Factor 14 contained one item which dealt with the degreeto which the individuals
felt the influence of risk audit, risk assessmentand risk evaluation on organisational
safety culture (Table 6-15, page 186). The fact that this item was not included in
factor 6, which representsthe "Employment of a risk programme", would indicate
that the individual's assessmentof the level of risk posedby the organisationis not
indicative of the correlation between risk managementand organisational safety.
This is curious when interpretedin the context of the finding describedin Chapter2,
which indicated that the system approach to risk managementis known as system
safety, implying that the process of risk management,which is used throughout
industry and commerce, involves identifying work activities and hazards and
estimating, evaluating and controlling the associatedrisk, is the just tool used to
achieveSMS (GAIN, 2000).
Factor 15 associates two items which indicate the Chief Executive Officer's (CEO)
decisions about safety investment, and the influence of the safety committee on
organisational safety culture (Table 6-16, page 186). There is no similarity between
these two items, making this. factor difficult to define, although both items were
deemed to be variables influential on organisational safety culture. The former item
focuses on the decision making style in the organisation, which the interview
suggested may not be symbolic enough to promoting a proactive safety mechanism,
and needs more work. The latter considers the influence of the safety committee on
.
safety, which is supposed to be included in internal factor 11 "Organisational
structure and management commitment". Chapter 2 highlighted Edward's (1999)
and Overall's (1999) 'comments regarding the importance of an effective

1, ",; "4
215
organisational structure for airline safety culture. This result may imply that a safety
committee has a review role to play in determining what are the safety issues;
however, the setting and planning of the safety committee depends on the
characteristics and operations of different airlines, which in turn influence
organisational safety and the proactive safety mechanism.
6.8 Implications of the Survey Results
Described in Chapter 4, the top-down method was selected for the safety mechanism
it
model since provided a useful description of the complex interactions which take
place between the myriad factors critical to the development of proactive safety. As
discussed previously, this analogy represents a way of thinking and the actions of
organisation. The aim of this model is to provide a greater understanding of such
organisational factors as influence an airline's proactive safety mechanism, and the
evolution of safety management system and safety performance (services). The
current result is useful in this regard.
Looking back on the results:
1. Predictors of perceived safety performance
It is worth noting that the significant predictors of perceived organisational safety
performance were: "Employee safety concept", "Communication system", "Impact
of accident/ incidents" and "Level of operational safety in operation and

maintenance". Although a causal relationship could not be established from
correlational analyses, these findings do provide some insight into the critical
factors of perceived airline safety performance.
Firstly, this finding demonstratesa relationship does exist between organisational

factors and safety performance. In particular, it was discovered that these four
factors - the importanceof employeesafety concepts,organisationalcommunication,
post-event adjustment,and the requirement for to

management createan atmosphere
216
,;
where personnel's contribution to safety in operation and maintenance, predict and
contribute to airline safety performance to a significant degree. The regression

equation also shows that "Employee safety concept' 'makes the largest contribution
to the prediction of perceived safety performance, while the "impact of
accidents/incidents" makes the least contribution to the prediction. It reflects the

literature in chapter 2: safety concept defines safety policies and principles, which
assign safety accountability, practice safety directives and consequently influence
the outcome of SMS.
Meanwhile, the highest, rating of perceived safety performance was from the
European region. Interestingly, it was found that regardless of the low number of
African representatives, European representatives rated the highest scores for factor
1 "Employee safety attitude and behaviour", factor 2 "Employee safety concept" and
factor 3 "Level of operational safety in operation and maintenance". This implies
that European airlines acknowledge more readily employees' safety concepts,
attitudes, behaviour and performance on operation and maintenance within their
airlines, and are reported to demonstrate these concerns more frequently and more
satisfactorily than other regions within the airline industry. This result is expected
also because European Authorities influence on SMS and related issues. As such, it
may be concluded that airlines in which these issues are thought to be important and
are acknowledged by the management may have the appropriate motivation and
attitudes to participate in safety initiates.
2. Implication of mean scores and standard deviation of factors
As described previously, the scoresrated by the respondentsnot only showed the
current situation of individual airlines, but also revealed what people thought to be
important, the degree to which they felt satisfied with them, and the regions they
came from. Across different regions, the results of ANOVA indicated that no
significant differences on each organisational factor were observed, except for
internal factor 1 "Employee safety attitude and behaviour" and factor 4 "Corporate
safety policy", plus external factor E3 "Impact of regulatory environment". This

finding pointed to the fact that safety managersfrom different regions viewed each
217
organisational factor influencing the proactive safety mechanism to a similar degree,
with the exception of these three factors. This is useful especially for the regulatory
authorities such UKCAA, FAA, Transport Canada, etc. when they promote the
concept of SMS. As such, followings will be divided into three parts to discuss:
+ With significant difference across regions:
In terms of factor 1 "Employee safety attitude and behaviour", a significant

difference was observed for factor scores across regions. Post hoc multiple
comparisons - Least-Significant Difference (LSD) indicated that this was

-
attributable to a significance between respondents in North America compared with
Europe, as well as Asia and Pacific. Moreover, regardless of the responses from
Africa and the Middle East, the table of means shows that safety managers in
Europe reported to voice this concern more frequently than other regions within the
airline industry. As such, one may conclude that European representatives (safety
managers) are more satisfied with their employees' competence on safety attitudes
and behaviour and are more aware of the influence of employee safety attitudes and
behaviour on the proactive safety mechanism.
In terms of factor 4 "Corporate safety policy", post hoc multiple comparisons -
LSD - indicated that this was attributable to a significant difference between
respondents in North America compared with Europe and Asia and Pacific. Taking
Africa and the Middle East out of consideration, the table of means shows that
safety managers in North America observed this aspect more frequently in their
working environment and deemed the influence of corporate safety policy on the
proactive safety mechanism to be more significant than did those from other regions.
The occurrence may result from the heavily regulated environment and regulatory
compliance in the region of North America.
In terms of external factor E3 "Impact of regulatory environment",post hoc multiple
comparisons- LSD - indicated that this was attributable to a significant difference
between respondentsin North America comparedwith Africa as well as Asia and

Pacific. Without taking into account those respondentsfrom Africa and the Middle
218
East, the table of means shows that safety managers in Asia and Pacific regarded
the impact of the regulatory environment, including regulators and regulations, on
proactive -safety mechanisms as being more important than did those from other
regions. Apart from the effect of European influence, such as in Malaysia, HK,
Brunei, and Singapore, the influence of Power Distance in National culture
dimension, identified in Chapter 2 may be the main effect.
+ With no significant difference across regions:
In terms of internal factors that were found to have no significant differences
across regions, the following factors were rated with lower mean scores on average
when compared with the mean scores of other factors: "Personnel - quality of
working life", "Financial concern", "Commercial cost pressures" and

"Communication system". Tested by ANOVA, safety managers from various regions
had similar degree of concerns regarding these factors. This indicated that in
addition to regarding these factors as being influential upon the safety mechanism,
safety managers from various regions had common feelings of their employees
being less satisfied with the quality of their working life, less effective
communication systems, and fewer perceived financial pressures compared with
other internal factors.
Some factors were rated with higher scores including: "Employee safety concept",
"Organisational structure and commitment", and "Necessity of safety report". This
demonstrated that the safety representatives had common feelings of higher
satisfaction of employee safety concepts, the importance of organisational structure
and management commitment, and the necessity of safety reports.
In terms of external factors, the two following factors were rated with lower scores,
and across the different regions respondents viewed these factors to a similar degree:
"Influences of region and country", and "Involvement of investment community".
The findings demonstrated that firstly, these factors are regarded as being the
influential factors within the safety mechanism in the external environment; and
secondly, respondents from various regions had common feelings of relatively less
219
influence of region and country on organisational safety (the analysis of the
relationship of internal and external factors in section 6.4 also found that the
influence of region and country was seen to play a relatively minor role in the
development of proactive safety mechanism), and less influence of the investment
community than other external factors.
With a difference within regions:
Meanwhile, the two factors: "Impact of accidents/incidents" and "Financial
concern" were tabled to have a standard deviation value greater than one within
most of the regions, except Latin America and the Caribbeanon the factor "Impact
of accident", and Africa on the factor "Financial concern". This meant that there
was some variability in responsesto the factors within the individual regions, and
future study may be needed.
3. Factor 2 "Employee safety concepts" vs. Factor 1 "Employee safety attitudes
and behaviour"
As mentioned in section 6.6, thesetwo factors were split off from one factor in the
original proposed model. In accordance with the definitions of safety climate and
safety culture within the airline industry, given in Chapter 4, safety concept is
related to safety climate, while safety attitude and behaviour is more closely related
to safety culture. Further investigation of the factor scoresrevealed that the mean
scoresfor safety concept tend to be higher than for safety attitude and behaviour. As
such, one may conclude that the former is deemed to be more important and
satisfactorythan the latter.
4. Factor E3 "Impact of regulatory environment" is correlated to all internal

factors
Factor E3 emergedas one of the most influential factors of the safety mechanism,
although the result of ANOVA showed that there was a significant difference
between respondentsin North America comparedwith Africa, as well as Asia and
220
Pacific, on this factor. The most noticeable point is that it is correlated to all the
internal factors by the correlation test. This finding revealedthat the airline industry
is a highly regulated businessand with the rising competitive airline industry (e.g.
Alliance, network sharing), the role of regulator will becomemore crucial.
.-
5. Factor 10 "Commercial cost pressures" and Factor 8 "Financial concern":
The appearanceof factor 10 "Commercial cost pressures" proved that commercial
cost pressures do exert an influence on airline safety mechanism, although, as

mentionedin Chapter4, those airline safety managersinterviewed did not admit any
cost pressureon safety. More interestingly, factor 8 "Financial concern" emergedas
a newly formed factor, which was identified to be one of the influential factors on
the safety mechanism in the survey results. It further showed that financial issues
also attractedthe attention of the respondentsto a critical degree,although the mean
scores of it and factor 10 "Commercial cost pressures"were not as high as other
factors.
6. Implication of organisational learning:
One point worth noting here is the survey results reveal the implication of
organisational learning within the model. Choularton (2001) and Sagan (1993)
presentfour constraintson organisationallearning. Theseare:
1. Feedback from the real world is often ambiguous. This allows
pre-conceived and convenient positions to be supported through different

interpretationsof the available information.
2. Post-event adjustment often takes place in a highly charged or political
in
environment which apportion of blame is sought.
3. Fault reporting from those individuals with vested interests in obscuring
the truth prevents objective analysisof the situation.
4. Secrecy, or the failure of internal organisational communication due to
restrictions on information flow, preventslearning from taking place.
221
These four points correspond to the three predictors of perceived safety performance:
"Level of operational safety in operation and maintenance", "Impact of
accidents/incidents" and "Communication systems". Assume that the barriers of

organisational learning (Y1) is the function of these three organisational factors (A,
B, Q. This could transform the relationship into a new equation, which is that
perceived safety performance (Y) is the function of Yl and D (see Figure 6-1). As
such, organisational learning barriers can be concluded to be one of the predictors

(factors to predict and influence) of perceived safety performance. This equation
indicates the contribution that organisational learning barriers can have on self-rated
it
safety performance, although does not prove the degree of this contribution.
Interestingly, this safety survey initially included the concern of organisational
learning, but the results of the pilot study suggested that this concern should be
dismissed, as stated in Chapter 5. Hence, the validation of organisational learning-
related issues in the organisational factors might need further investigation in the
future.
Figure 6-1 Organisational Learning Barriers vs. Perceived Safety Performance
Perceived Safety Performance Predictors from the survey result
/Barrier of organisational learning identified by Choularton and Sagan
A: Level of operationalsafety in operation

Y1: Barriers of
and maintenance/ Feedback+ Fault reporting
organisational learning
B: Impact of accident/incident Yj= f (A, B, C)

/ Post event adjustment
C: Communicationsystem
/Organisationalcommunication
Y. Perceivedsafety
performance
D: Employee safety concept
Y=f (YI, D)
222
6.9 Examination of the Findings for the Primary Hypotheses
Previous sections have identified the factors which influence proactive safety
mechanism and discussed some relationships between these factors - and their
implications. The next will integrate the findings from the study and the rest of the
investigations reported in this thesis.
Four general hypotheses were explored and tested in order to develop the safety
mechanism within a proactive management.As mentioned in Chapter 4, the term

hypothesis as used here should be understoodin the broad senseof the word as these
are intended to outline the aims and objectives of the model. These statementsshould
be consideredto be propositions, made from known facts, which form the basis for
this investigation, rather than hypothesesin the statistical sense,which will be tested
and either acceptedor rejected.
To this end, four hypotheseswere set to be explored and tested in this thesis. The
outcomesof the investigation with respectto thesehypotheseswill be discussedin the

following sectionsin order to summarisethe results of the investigation which formed
the substancesof this thesis.
6.9.1 Hypothesis I
Hypothesis 1:
The safety mechanism is the composite of organisational climate/culture, safety
climate/culture, safety philosophy, and decision-making. The definition of each layer
will be redefined (except decision-making) to suit the airline industry.
The first hypothesis,the redefinition of organisational climate, organisational culture,
safety climate, safety culture and safety philosophy, is critical to the application of a
safety mechanism for the ' airline industry. As the current existing definitions of
organisational climate, organisational culture, safety climate, and safety culture' are
223
the generalisation for an open system, there is a need to distinguish the differences
betweenairlines and other industries in order to develop an airline safety mechanism.
Building upon the knowledge of existing definitions of organisationalclimate/ culture,
and safety climate/ culture statedin Chapter2 and Chapter4, the primary evidence for
this hypothesis was presented in Chapter 4, where organisational climate,
organisational culture, safety climate, safety culture and safety philosophy were
redefined to illustrate and suit the needsof the airline industry. As such, through the
literature and empirical findings, the first hypothesisis supported.
6.9.2 Hypothesis 2
Hypothesis 2:
The safety mechanism model will be demonstrated as a multi-dimensional construct,
which will be influenced by a complex factor-structure. Meanwhile, this factor
structure will also reflect the current environmental factors in the airline industry.
The hypothesis that each layer of this safety mechanism would have a unique
definition (Hypothesis 1) is critical to the justification of a complex and

multi-dimensional construct underlying the safety mechanism model. Having defined
the. components in the model, this model illustrates a top-down metaphor, which
shows that the organisational climate exerts an influence on organisational culture,

which affects safety climate and culture; safety culture influences safety philosophy,
and safety philosophy informs decision-making, which results in good or bad
consequences in safety services. The multi-dimensional relationships are explained
and supported by the literature described in Chapter 4 regarding the development of
the model.
Meanwhile, the assertionthat this constructwould be found to be influenced by a set
of identifiable factors is important. The identification of a limited number of factors

serves to reduce the myriads of issue relating to the development of a safety
mechanism to a reasonablenumber of items, and to acknowledgethere are real effects
224
imposed by theses factors. The results of principal component analyses were
successful in identifying the internal (thirteen) and external (four) factors for the
safety mechanism; for example external factor El "Influence of region and country"
shows the effect of culture and value of the country and region in which the
organisation exists, although no efficient evidence proves whether the influence of
country or region can override any organisational attempts at developing itself. If
expressed by an equation, it has the form:
Y=F(X X2... X13)

Y=F(EI... E4)
Where
Y is the safety mechanism,(the compositeof Y is defined in Hypothesis 1)

Xs are internal organisationalfactors,
Es are external organisationalfactors.
Various ANOVA in this chapter further reveal the complexity of the factor-structure
by identifying the fact that not all respondentsregard these factors to be important to
the same degree,and identifying which factors are placed as high concernsin which
regions.
Moreover, the case study in Chapter 5 provides sufficient evidenceto test the model
and conclude this hypothesis.However, one point to note here is that there are slight
differencesregarding the factor-structurecomparedwith the one originally proposed,
discussedin section 6.6. As such,there is a need to redraw the casestudy by using the
new factor-structureof the safety mechanism(Figure 6-2).

Figure 6-2 The Safety Mechanism Model Applied to the Air Ontario Accident
Line Worker Middle MQt. Senior Mgt. Ora. Factors
Involvement of
Org. " job security " future of company " operation goals """"""..... ... E4: investment comma
`" g: Financial concerns
Climate " future " operation goals " viability + survival
Commercial
fI. count
country regulatory 10: cost
pressures
Influence of
pressures (culture), """ý""
""" ""º El:
country
regloo
E3: Impact of regulato

environment
Org. " org. changes " turbulent " lack communication -º 12: Communication
Culture " turbulent environment with middle mgt.
environment (over taxed, ) " merger undergoing 5: Personnel

...
" lack of support from 11: Ormýi geonnaalthuc
t
senior mgt. commitment
9: Procedures &
Documentation
Safety " minimal safety " failure to issue MEL " poor regulatory .... ...... . ... º E3: Impact of regulato
environment
Climate initiates " failure to provide environment
" bent to keep training and " short-term focus in 6: Employment

of ris
a/c flying manual turbulent environment 3: L, velof operati
f safety onal
" safety information not " failure of regulatory """""" """"º E3: actof gnlt°
Safety " deferral of environment
Culture maintenanceý communicated wel compliance
" lack of training " little observation an " safety manager has
safety knowledge coaching in safety no direct access to
behaviour CEO " safety manager has
-º 11: Organisational
no direct access to & managementstrut
commitment
time performance-+" flight was behind the 2: Employee

Safety " highly motivated-on _. safety concept
Phil. to take necessary was prime criteria schedule
steps to keep -'" take necessarysteps " passengercould not beº 4: CQTPorate
safety policy
a/c flying to keep a/c flying stranded on board
Decision " captain tookI -º" Sent the disabled -º" a/c had got to fly 1: ä
Emplqyee
ade ssafe
-making off without de-icing a/c to an airport

" maintenance left without ground
APU un-serviced starting facilities
Result
I
DrydenAccident l
Note: 1. Org. - Organisation,Mgt. = Management,Phil. = Philosophy,a/c = Aircraft
226
Moreover, the evidence of a complex factor-structure is also supported by some

in
analysesconducted this chapter, where the factor scores between internal factors
and external factors were correlated to a significant degree. In particular, all the
internal factors were found to be significantly correlated to "the impact of the
regulatory environment", which was evidence that the airline industry is a highly
regulated business, and that regulators strongly desire airlines' co-operation. It also
implies that with the different regulatory structuresacrossregions, a more harmonised
regulatory environment is necessary.
The safety mechanism model is not a model of proactive safety management or
organisational climate/culture. It is a model of the evolution of safety serviceswithin

a proactive safety managementsystem. The importance of this model is to assess
airline proactive safety by identifying the organisational factors which will affect
organisational climate and analogously influence the safety mechanism. These
findings generally support the secondhypothesisthat the safety mechanismmodel is a
multi-dimensional construct,which is influenced by a factor-structure:internal factors
exert influence within the airline, while external factors reflect the environmental
influential forces.
6.9.3 Hypothesis 3
Hypothesis 3:
The organisational factors mediate the relationship between the safety mechanism
and safetyperformance, and act aspredictors of safetyperformance.
The third hypothesis, that the organisational factors mediate the relationship between
the safety mechanism and safety performance, and act as predictors of safety
performance, is important in fully understanding the implications of the safety
mechanism. There are two reasons. Firstly, it highlights the relationship between the
safety performance and organisational factors, i. e. the particular factors at a particular

time, which affect airline safety performance to a significant degree. Secondly, the
identified predictors serve to identify the areas recommended for improvement and for
future study of this research topic (see section 6.11).
227
The evidence of influence of safety mechanism on self-rated safety performance is
presentedin Chapter 4 when discussingthe model demonstratedby Neal et al (2000).

The primary evidence with regard to this hypothesis was presentedin this chapter.
The result of the multiple regression analysis indicates that there is a relationship
between safety performance factor and organisationalfactors, i. e. four factors are the
best predictors of self-rated safety performanceamong theseorganisationalfactors. If
expressedby a conceptualequation,it has the form (refer to section 6.3):
Yj=. 486+. 403*X2 +. 244*X3 +. 113*X7 +. 133*X12
Where Yl is the perceivedsafety performance(output or effect),

Xs are organisationalfactors (inputs or causes):
X2: Employee safetybehaviour
X3: Level of operational safety in operation and maintenance
X7: Impact of accidents/incidents
X12:Communication system
Although one may argue that the essenceof organisational factors in a company's
organisation deliberations will almost certainty affect safety performance as it leaves

the way open to error and failure, it is worth noting that the questionnairerespondents
rated the "safety performance" here becausethere exists the difficulties to define/ rate
airline safety performancein reality plus the questionnaireis anonymous. Therefore,
Hypothesis 3 should be the hypothesis regarding the organisational factors as the
predictors of perceived safety performance.The findings of thesefour internal factors

show internal environment will have more influence on the perceived safety
performance than external environment, and the attention should be paid more than
other factors when proactive safety managementsystemis developed.
228
6.9.4 Hypothesis 4
Hypothesis 4:
The safety mechanism model is seen as being critical to the development of proactive
safety management; however, the airline safety health and performance still needs the
coordination of proactive and reactive safety management.
The last hypothesis stated that the safety mechanism model is critical to the
development of proactive safety management and airline safety health; however,
airline safety performance still needs the co-ordination of proactive and reactive
safety management. The logic underlying this hypothesis is that in the light of
empirical study and literature review, current proactive concepts and approaches aim
to uncover the latent organisational conditions and avoid the surrounding errors or
inadequate recognition, which cause safety to be degraded in the workplace. Building
on this knowledge, this research aims to identify the importance of the safety
mechanism, and the organisational factors which affect it in order to establish a
proactive approach to safety.
As mentioned in Chapter 2, and in keeping with the aim of maintaining a
well-functioning safety services, continual adjustments, and sometimes the

introduction of new methods, are required. Nevertheless, when undertaking such
changes,one must be also careful that those aspectsof safety services that function
well are not destroyed.From the result of literature review, retroactive approachesto
safety have been identified as important existing tools to improve airline safety
managementsystem.This fact is also supportedby the results of survey,which reveal
the factor-structure containing these factors "the employment of risk programmes",
"impact of accidents/incidents", and "the necessity of safety reports", which were
identified as retroactive approachesto safety in Chapter 2. Moreover, this factor
-
"impact of accidents/incidents" is found to be one of the predictors of perceived
safety performance. These findings indicate that the safety mechanism is built on a
proactive concept; however, retroactive and proactive vehicles both exert influences
on this mechanism model in order to achieve safety performance and safety health.
As such, it can be concluded that that the airline's safety health and performancedo
229
need the co-ordination of both retroactive and proactive safety managementand the
fourth hypothesisis accepted.
From these results, the safety mechanism model provides an indication as to what
drives the development of a safety mechanism, what the influential factors of the
safety mechanism are, and what organisational factors are the predictors of perceived
safety performance. Figure 6-3 portrays the safety mechanism model resulting from
the survey. By the acceptance of the four hypotheses, these results are the main
contributions that have been made from the thesis.
230
o
o
CL) M
(3) 0
.0m
ot5 'ý
- ö
V ü 79 a
ý uni L-' C- 0
Ü 3g: O Cl)
Q =ý U) ü G) + 0]
. QC iu to -
a 0)
J .ý 0) W
4 ca 0 üa°m
a' Q1 2 70
- ° co
z .ik (ß (13
i- C4
:Doa c f6
N r5- ++ -cm .-
w (o dl {A ü OO QD O
r, C-)
W Q7 M üii 76 E
OO
O - es t+ aE E F- U)
OO CD p E
IÖE E
Q C2 a) Z U7 0- M
_Q a1 O 41
EE
O O -d) C- CL C
WWJ E OWWE iL W0U
aý
XX? X>ý>ý> >X 7C 7C
0
äý
b
0
U
8
co CD
Eb IE w
Y1
CC
O (D Ct
U7 ý
H
ky [ß +i
, M
G)
i.
bA
"rr_
h+ýl
0 0)
CU v
V G;
H C F3 L' Q)
V Q as o >
d aF-
Lj-
a> > o
CD
0 cl)
.
z CDCo o E >
üU
H 0- °
3 ß o
X U
W W
WWW W
231
6.10 Limitations and Validation of the Safety Mechanism

Model
Although the items (questions)of the safety survey were developedfrom an extensive
literature review, and qualitative data was collected by meansof interviews, it is not
possible to say at this time whether the factors analysed from the survey would be
accuratewithout further investigation. As such, the next sectionsfocus mainly on the
limitations that might causethe factors identified in the safety mechanismmodel to
fail to accuratelyreflect the factors which have an influence on the safety mechanism
and perceivedsafety performance,and why.
Firstly, the safety mechanism contains a number of inter-related elements, which are
not usually differentiated as black and white, such as the difference between
organisational climate and organisational culture, etc. Therefore, the distinctions

between these elements may not be easily understood in the first place, and so may
take some time for the reader to appreciate. Secondly, due to the difficulties in finding
a benchmarking measurement of airline safety performance, the safety performance
measured in this research is respondents' perceived safety performance. Thus, this

handicap may deviate its factor score and regression to some degree as a result.
Thirdly, as statedin section 6.7, the relatively low number of respondentsfrom Africa,
the Middle East, and Latin America and the Caribbean,plus their responsetendency,
might cause a bias in the survey results; nevertheless, a larger proportion of
questionnairesreturnedwould add greaterconfidenceto the conclusionsof the thesis.
With the limitations of the safety mechanismmodel in mind, the next section aim to
explore its validation by comparing this model with other four safety programmes
with proactive concept, which were discussedin Chapter 2. The comparisonof these
four programmesand the safety mechanismmodel are listed in Table 6-24.
232
w
O +N
U
N NN
a'ý Y[ j. Up 0.0
^ cN 7i
aUl
OýJ
Ql Ci YQa OJ
U Z'3 '_'r br U
o o
.t
UR ! y'ý
Y O >
YO O (U ý +' '" Üc UO
. i nyO cq
¢
ý1
ro
V
N
6b
Q] G
ro ýl
r+ý
(V r _1yU 7 "
ýI
Lý4 O
U_
Up Rý Ný' O. ro OU j
ip. ýNN [O U ip+ O
3.1 C, y
w 0 Y wp ^Oý
cU F3 GJ Si 0. ) fy i-ý iU L: N
W
RI SJ W
ö y[y 'Ü
ON . (..ý' Ry
Ri
UC: g90. ten QO cn O
aw OýJ'., . Y "tla U fd
OU ro y ý.
y { r'
O r
ro ro
Siý 1-r. N¢ U OA UO
a ¢" tti " "r (U
U)
.a
'J N
ä
G.+ dl "NOOy "'
OOy..
ý O Q1 ffMrr '-' O 0
Wp W70 UC4 IYC4UCD u ý! 0
N O Gsu
c
Cý U r-i N 1'ßl V' Vl ýO .O N -
"--"
. -. . -.
0
ö
WNaý! CD
KZ
a)
O
'S
y! Ü. o
O
O 47
y^
ý.. Nr U'
KZ G
G
J ° °J' ö ro ro 33
a a , x
RO U
.ý
[l1
W
i: ý4 NNO [ti
w
a
v ui
Ö a)
y
O t_
O U
ti
o dU
Cä
W E r u z
^ C c
4. a) (U cd
N a
W 73 b2
ai
o
i vroi
ro o
ý-ä
ö a) m z
9
L) Z-1 li UOY
-103
w
-ö o aý
Yo o Lei 0
ý'!
U ¢ W
ro
OY
Cd
to u'
Q)
N0Y
m
ö ro & äoß
Dw v
iy
w
o
Z R L'i C OA G O 'ý ý-!
NN
Y Cý . yý
CD a' a'
ý'D p ro i: bA
O ! y., U
Op ý, (U Q
Y'pNL .0 w to 67
P. ro ro , ro ry ý
M
aý Q1 Ö M 4] N- of , ý{
N
row
Uw ((ý
o W
ý4 W A. 7r C/ý U! UZ LZ ý-!
"rl O N rl N T'i
ýýr"`
ffh C
C. 40
O
NM Vl ýO W
^ý . --. NM Vl
a)
ý'
UG
Nr
N
C
SN
O
U
OÖ0.
UGQ
Ö L7
y U y 4y 7
C) _-
qý Uc
ýN
f.K
cý o v CD a) Ö
-7U r-
o 0
a) C. OA O) w
UFU
w ý"ý
p F; O" "C O TO
y ýi ý+ ü
cp
Va y cd
2 yOOOv
'ý Qp
R. vii 'vý i'+ lam. OA
uni y
OO O d'
1-y-y yO pq yQ p O ,Oa
ý-'!
N r
CLO OO Op NM
fD 5ý"" 1y-"
[O (O lO ai UNO
n. n
aJ i+
OCZ CLf-iý C,G Z? "ýXXF [L I- WC)p.
Q p
LZ
.rNM 'S Vl ýO Iý CO .rN
ºa . --. N Cpl Vl 11D r- CO Qý .-.
v
; a_ £ olopoq aw ö
7
233
Reason (1995b) points out that effective safety management requires both reactive
and proactive information in order to guide an organisation to that region of the

"safety space" associated with the greatest resilience to operational hazards. In both
cases, though, it is necessary to identify the organisational and situational factors
contributing to unsafe acts. As such, MESH employed Reason's philosophy and his
model for programme development. Referring to Table 6-24, in MESH, front line
personnel assess the local factors, and technical management assess organisational
factors at periodic intervals. The aim is to identify factors in need of improvement and
track the changes over time, with the help of computer software package. However,
there is no literature which describes exactly how these factors were chosen.
As to INDICATE, Reason's model was applied to develop the programme as well, but
with a difference in focal point. The INDICATE programme focuses on the aims of
identifying and resolving deficient aviation safety defences before the occurrence of
mishaps. By dividing the experiment company into two groups and observing their
operation, it provides a means of evaluating the effects of the programme. It is more
suitable to small and medium sized airlines with budgetary limits.
PERS is an error management system which can be used by non-experts in human
factors because it is based on a human-factors approach to solving errors. PERS
provides a way for airline personnel to analyse an error or potential error, to discover
why it occurred, and then to see how they might go about changing systems,
equipment or work patterns to prevent future errors. The solutions presented by PERS
tend to be limited to exploring potentially hazardous situations by analysing the
ongoing problems.
LOSA highlights exemplary, as well as deficient performance,which shows airlines

the areasin which they excel as well as those in need of improvement. A databaseis
being developedthat allows organisationsto comparetheir results with those of other
airlines. The focus of LOSA is mainly on flight crew and flight operations division.
However, the expenseof LOSA is too high to be afforded by airlines with budgetary
limits.
234
Comparing the safety mechanismmodel with other programmes,it is found that aside
from the contribution of organisational factors exploration and identification within
the academic literature, this model also provides the airline industry with a
"benchmark" of organisational factors which are an aggregationof the viewpoints of
various representativesacross all regions. The rating of each question (response)
presents the tendency of each organisational factor, i. e. how important it is and to

what degree it is to be satisfactory across regions. These organisational factors are
also assessedby ANOVA to investigate whether significant differences exist across
regions. It is important to understand the regional difference, especially when
applying this model in the real world. Meanwhile, this model offers a "prototype" of
the contributing organisationalfactors to the perceived safety performance.
It is not the intention of this part of the thesis to suggestthat the safety mechanism
model is the best among these programmes, although the flexibility of this model's
application exceeds that of other programmes. Its benefits include its relatively low
cost, the fact that there is no limitation to specific divisions, it provides a platform for
the integration of organisational factors and change management, etc. (see the
application of the safety mechanismmodel in the following section). The purpose of
the comparison is to show the differences between different programmes, and to
hopefully validate the model by pinpointing its advantages,and the contribution that it
makesto the literature and to the real world. This point should be emphasised.
235
6.11 The Application of the Safety Mechanism Model
The case study of Air Ontario has demonstrated the retrospective application of the
safety mechanism model, which is quite useful in retroactive safety management, e.g.
accident/incident investigation. This portion of the study will add value to the
application of the safety mechanism model in reality. The application of this model is
its implementation in a practical situation, providing a vehicle for identifying
organisational factors within airlines. In other words, there needs to be a change, but
how can change be brought about in the existing organisation? A process of change
management must be adopted in order to effectively embed the organisational factors

that have been identified in the safety mechanism model. As mentioned previously,
the factors identified serve as a benchmark and a prototype on a world-wide basis.
Therefore, the airline can identify its own factor set (by replicating the questionnaire
in its own organisation, or department, i. e. local factor-set), compare the results with
the safety mechanism model, and decide which factors are in need of change.
There are severalwell-known corporatechangemanagementtools. Six Sigma42is one
of them. Six Sigma provides a means to identify the major areas (project) for
improvement, form the team, and apply the Define (D) Measure (M) Analyse (A)
Improve (I) and Control (C) methodology43to completethe project. By data collection
and analysis, root causesof problems can be found, and direction for improvements
are able to be identified.
.
42 Six Sigma -A vision of quality which equateswith only 3.4 defectsper million opportunities for
eachproduct or service transaction.This concepthasbeenusedin corporatemanagementto
improve deficient processes.In other words, Six Sigma is a highly disciplined processthat helps to
focus on developing and delivering near-perfectproducts and services.The essencebehind Six
Sigma is that defectsare unknown and needto be identified. If you can measurehow many
"defects" you have in a process,you can systematicallyfigure out how to eliminate them and get
as close to "zero defects" as possible. GeneralElectric (GE) is a role model famous as a Six Sigma
organisation.
43 DMAIC - (Define, Measure,Analyse, Improve and Control) is a processfor continual

improvementpurposes.It is systemic and fact based.This closed-loopprocesseliminates
unproductive steps,exploresnew measurements,and applies controls for improvement (see
Appendix T).
236
The following shows the application of the model by using the prototype factor set as
a demonstration.As such, thirteen internal and four external factors were identified as
influential factors of the safety mechanism.According to Figure 6-3 (page 231), they
have the form:
Y= F (XI, X2 X13)
...
Y=F(EI... E4)
Yj= 486+ 403*X2 +. 244 * X3 +. 113* X7 +. 133* X12
. .
Where Y is the safety mechanism

Yl is perceivedsafety performance
Xs are internal factors, Es are external factors
X2 is Employee safety behaviour
X3 is Level of operationalsafety in operation and maintenance
X7 is Impact of accidents/incidents
X12is Communication system
As the survey result reveals, there are four influential factors for perceived safety
performance,and it is suggestedthat these should be applied as the starting point for
improvement. One point worth noting is that although "Employee safety concept" is
identified as the biggest predictor of safety performance,it is however found that the
"Employee safety concept" has the highest level of satisfaction, while the
"Communication system" is less satisfactory.Therefore, "Communication system" is
the factor in greatest need of improvement, if budget or resources allocation in a
is
company not enoughto support all improvement plans at one time.
By applying the Six Sigma approach, factors can be broken down to several
manageableprojects, collecting the improved results and achieving a better perceived

safety performance. The details of the application of organisational factors using the
Six Sigma approachare describedas follows:
237
Referring to the denotation in previous page, break Yl down to a number of Xs of a
manageable size for improvement. As such,
Y1= F (X2, X3,X7,X12)
X2, X3, X7,X12are the focus or results of the processto be improved. Each X contains
a Six Sigma project, which aims to explore possibilities for improvement.
Take communication system(X12)for example:
X12=F (Zi, Z2, Z3, Z4....)
where X12is the communicationsystem

Zs are the actionsneededto improve X12
Each Six Sigma project encompassesthree aspects, as follows:
1. Who: cross-functionteam members(the major players)

2. What: objectives, measurements } DMAIC process
_...
3. How: analysis, improvement, control tools
The DMAIC processesof X12could be detailedin Table 6-25.
238
(7uiptc'r 6 4uulv. ves and Discussions
X12 =F (Z1, Z2, Z3, Z4.... ), Zs will be identified in the "Analyse" phase.
Table 6-25 The DMAIC Process of X12 "Communication system"
Phase Action Goal

Define What is important to the communication system
Define "X12"
(critical factors)?
(D)
Measure What and how is the process of communication

system performing? Measure "X12"
(M) How reliable is the obtained data?
What are the critical defects causing variation?

To what degree?
For instance:
Analyse Z, = Knowledge management system Find & Measure
Z2 = Performance evaluation system the "Zs"
(A)
Z3 = Existing resources (bulletin, company mail,
etc) and communication channels
incorporation
Z4= Rewarding system
How to fix the critical defect (Zs)?

What variation in critical Xs can be removed?
Improve For instance: Improve the "ZS"
(I) How much percentage of employee satisfaction/
benefit and communication channels can be
increased by building a KM website?
How can the improvement be maintained? Control "ZS"so that

Control there is no variation in
For instance : the "X, 2"
(C)
Maintain the website and free flow of channels
It is worth noting that, Six Sigma is not panacea that can cure all the problems.
Having demonstrated the Six Sigma methodology, there are also other change
management tools that can assist practitioners in identifying these actions (sub-factors)
which warrant the greatest amount of attention in improving the organisational

factors.
239
240
CHAPTER 7
Conclusions and
Recommendations
"But People who arefreed from the confinesof their box on the organization chart,
whosestatus rests on real world achievement... those are thepeople who develop
the self-confidenceto be simple, to share every bit of information available to them,
to listen to thoseabove,below and around them and then move boldly. "
Jack Welch,1989
7.1 Conclusions - Thesis Aims Revisited
Looking back to Chapter 1, five objectives have been set out. By summarising the
findings which are aligned with the objectives for the reader,this chapteraims to bring
togetherthe knowledge gainedthroughout the investigationsthat formed the substance
of this study. The achievement of this thesis and the findings, as well as its
will
recommendations, be discussedin the following sections.
7.1.1 Objective One
To evaluate current airline safety management systems and become a
reference text which can be used by academics and industry.
Starting from Chapter2, this thesis opensby defining safety as "risks are minimised to
an acceptable level" and introducing the safety interactions in the air transport system,
given that the safety of air travel and its inherent challengeshave beenwith us sincethe
n_ 241
Chapter7 Conclusions
andRecommendations
first passenger flight. By identifying the interaction of safety significant activities and
external influence within the system, which are involved in conducting a safe flight, it
was found that the external impacts are mainly from three aspects: technical innovation,
growing traffic, and the regulatory environment. The innovation of technology

increases the reliability and productivity of aircraft, which in turn stimulate the growth
of air traffic. Nonetheless, it is the airline's responsibility to provide transportation as
well as safety, each is mutually dependent on, and worthless without the other.
Especially with the increasing competition in the airline industry, the safety records of
airline not only are used as a basis for passengers for deciding which airline to choose,
but also directly associates with the airline's reputation within the media and public.
Meanwhile, given the complexity of aviation environment, the regulators play an

important role in airline operation. These all exert influence on the airline safety
management system (SMS).
The subsequent literature review probes the evaluation of airline safety management
system and clearly compares the empirical situations systematically, which identifies
and rationalises the areas for improvement in the airline industry. It is concluded that
is
safety supported by organisational structure and regulations; fundamentally it is the
concept and behaviours that define safety policy and principles, which influence the
outcome of SMS. In addition, SMS must represent a pure closed loop. It must have
facility for setting process and standards and subsequent reviews, and then adjustment
is revised process or new process. The judge needs to consider what is acceptable and
when there is a threat requiring change. The coverage of literature review and extent of
evaluation is sufficient depth and quality to be used as a reference by academics and the
industry. Therefore, the objective one - to evaluate current airline safety management
systems and become a reference text which can be used by academics and industry, is
achieved.
242
Chapter 7 Conclusionsand Recommendations
7.1.2 Objective Two
To investigate retroactive and proactive approaches to safety and their

application within the airline industry.
The evaluation of current airline safety management system leads to two main
approaches to enhance and improve airline safety services in the present circumstances.
One approachis that of retroactive measures;the other is theproactive measures.
Traditionally retroactive approach to safety
The main tasks of the secondobjective are to firstly review current after-fact measures
to safety, including accident investigation, human factors, and risk management-
evaluating the existing risk control tools to maintain safety quality. Within a complex
system like the airline industry, it is necessary- and usually taken for granted- to trace
back the causalfactors leading to the accident.By identifying the strategiesthat would
enableus to copewith the deficiencies,the control systemof airline safety servicescan

be improved. The value of accidentinvestigation lies in learning the relevant lessonsfor
future prevention. As such, accident investigation is an appropriate tool to find out
unanticipated failures in technology, and serves the purpose of contributing to the

human error database.Recent incident reporting systemsare also emphasised.Their
value lies mainly in understanding system and operational human performance,

pointing out causesfor concern and signal before
weaknesses the systembreaks down.
Other impacts of accidents include the improvement of safety managementand risk
management. The concept of the safety management system is now extended to

integrate existing knowledge of safety managementand provide a systematic view of
"know-how", to which this industry has woken. Risk management tools are also
developedin order to collect safety information andprevent the identified errors,which
were discovered in the human factor study. As human imperfection remains a fact of
life, error management is developed to observe training (analyse workplace)
behavioursin an attempt to help to understandoperationalhuman performance.
243
andRecommendations
Causes of accidents serve as contributing and primary factors to this approach, while
risk management, human factors and error management are the diagnostic tools.
However, the low number of accidents makes it difficult for these after-fact measures to
determine patterns and to establish what to do next. Although incident investigation can
have both reactive and proactive concept, a few researchers have proposed that
adopting a proactive approach would be more effective than simply reacting to
accidents. As such, contemporary proactive approaches to safety are investigated.
Proactive Approaches to Safe
The main tasks in this part are to evaluatecurrent proactive concept and measures,to
develop a proactive approachto safety from a cultural and organisationaldimension,
and investigate the differences between retroactive and proactive approaches.The

differencebetweenretroactive and proactive safety managementis in the treatmentof
the contributing causesof accidentsand their underlying factors to prevent accidents
from occurring again.The traditional approachto studying aviation safetyhas involved
the analysis of accident data. Researchhas focused on finding factors that link to, for
example, pilot-error accidents through systematic accident investigations. However,
accidents still happen and the accident rate has not been significantly improved for
severaldecades,
which points to the need for a new approachin safety management.
Proactive safety management moves away from individual error as the focus of
understanding accident causation, to the organisational approach. This method assigns
the responsibility from one person at the sharp end to all systems within an organisation.
As such, the current contemporary approaches investigated in this thesis are designed to
aim at a systematic approach to the identification and prevention of accidents. The
concept is also designed to ensure continuous improvement by analysing whether the
values of an airline's culture are consistent with good safety practices. Most important
of all, the proactive approach aims to change the organisational culture surrounding
errors or inadequate recognition, which degrade safety, and uncover the latent
organisational conditions stemming from organisational culture.
Chapter 2 presents the results of an exploration of empirical and theoretical findings
244
andRecommendations
regarding the approaches to safety within the airline industry. The results also show the
importance of and need for proactive approaches. To this end, the second objective- to
investigate retroactive and proactive approaches to safety and their application within
the airline industry is achieved.
7.1.3 Objective Three
To develop a model for a `safety mechanism' in the context of proactive

safety management
It is evident from the research findings that proactive safety is an important but still
relatively undefined task in the airline industry. What it lacks is a model to
conceptualise the safety mechanism in proactive safety management, which
encompasses the culture and climate considerations and organisationally influential
factors. Therefore, the aim to establish a framework for the study of a proactive
approach and for the evolution of airline safety management system by the
development of a safety mechanism model, has been targeted and developed after
investigating both approaches to safety.
Chapter3 therefore describesthesemethodologiesapplied in the study.Thesemethods
provide a framework which allows airline safety managementsystemto be analysedin
a systematic manner, especially as the main emphasis was on the fundamental
objectives of the evaluation, i. e. to obtain a critique of some of the main problems

relating to airlines' safety management system, and justify the best solutions for
overcoming them.
This proactive safety mechanismmodel is built upon the existing knowledge of what is
thought to contribute to an effective proactive safetyby adding an increasedknowledge
of the organisationalfactors, which serveto influence the proactive safety mechanism,

and which will serve to be the predictors of the safety mechanismand performanceof
airline safety services.As such, this part is

of study exploratory in focus and qualitative
in nature.
245
Chapter 7 Conclusions and Recommendations
In Chapter 4, it introduces that interviews were firstly undertaken to gain the empirical
knowledge and experiences necessary to develop a safety mechanism for proactive
safety management. Combined with the academic studies in previous chapters, it is
found that management systems and programmes must provide an effective safety
framework because it is the importance of safety to the organisation, and the workers'
perception of the value of safety, that "manipulate" safety performance.
A safety mechanism model consisting of organisational climate, organisational culture,
safety climate, safety culture, safety philosophy and decision-making has been
developed. This model is neither a model of proactive safety management nor an
organisational safety culture/climate model. It is a model of the evolution of the airline
safety services (performance) with proactive safety management in SMS. By means of

defining the criteria and interaction between each layer, the discussion provides an
appreciation of how the safety mechanism should fit within in the airline industry, and
build upon the concepts which have been prominent within the literature to this point in
time. Therefore, the third objective - to develop a model for a `safety mechanism'in the
is
context of proactive safety management achieved.
7.1.4 Objective Four
To verify organisational factors that affect the safety mechanism and

investigate the relationship between the factors.
The model also focuseson the influential factors which impact upon the development
of a safety mechanism.Given the need for an examination of the structureunderlying

the model, Chapter5 thereforepresentsthe result of model test, including a casestudy
and safety survey, Chapter 6 analysesand discussesthe results in order to verify the
fourth objective.
246
Case study
The casestudy presentedin Chapter5 provides a useful illustration of the myriad forces
which conspire to define a system's safety mechanism. Given the complexity of the
case,the casestudy provides the readerwith a greaterappreciation of the explanatory
power of the model in relation to how the safetymechanismhasdevelopedand how the
safety managementsystemand serviceshave evolved.
Survey Results
The main objective of this safety mechanism survey is to take a closer look at the
relationships betweenthe forces hypothesisedto be at work, provide an assessmentof

theseforces, verify the influential factors and validate the hypothesisedfactor structure
in the safety mechanism.In short, it aimed to provide a framework for understanding
the factors which underlie the developmentof a safety mechanismacrossregions. By
the survey results, the primary contribution which this model makesto the literature is
to expandthe conceptof proactive safety from a conceptualisationof what an effective
organisational culture or safety culture should be to a comprehensivemodel which

recognisesthe fact that the safety is
mechanism a complex, multi-dimensional construct,
which evolveswith time accordingto a wide variety of organisationalfactors within the
internal and externalworking environment of the airline industry. By doing so, possible
remedial action for the evolution of airline safety managementsystemand servicescan

be recommended.
247
To this end, four hypotheseswere set to be explored and testedin this thesis.They are
discussedin Chapter6:
Hypothesis 1: The safety mechanism is the composite of organisational

climate/culture, safety climate/culture, safetyphilosophy, and decision-making.
The definition of each layer will be redefined (except decision-making) to suit
the airline industry.
The first hypothesisis supportedby the literature review and empirical findings.
Hypothesis 2: The safety mechanism model will be demonstrated as a
multi-dimensional construct, which will be influenced by a complex
factor-structure. Meanwhile, this factor structure will also reflect the current
environmental factors in the airline industry.
The second hypothesis is supported by the case study and survey results,
although a slight difference between the hypothesised factor structure and
survey results was found out.
Hypothesis 3: The organisational factors mediate the relationship between the

safety mechanism and safety performance, and act as predictors of safety
performance.
The third hypothesisis supportedby the literature and survey results.
Hypothesis 4: The safety mechanism model is seen as being critical to the

development of proactive safety management; however, the airline safety health
and performance still needs the coordination of proactive and reactive safety
management.
The forth hypothesis is supportedby the combined finding from the results of
interview, casestudy and safety survey.
248
As such, the fourth objective - to verb organisational factors that affect the safety
mechanism and investigate the relationship between the factors is achieved after these
four hypothesesare supportedby the researchfindings.
7.1.5 Objective Five
To make recommendations concerning how airlines can fit the proposed

safety mechanism model to the industry for the evolution of a safety
management system.
In Chapter 6, the limitations and application of the safety mechanismmodel are also
discussed to validate the model. Moreover, section 7.2 provides the final
recommendationsin order to fit in the airline industry and make contribution for the
evolution of an airline safety managementsystem. The objective five is then achieved.
249
7.2 Recommendations
The following are the recommendationsin view of:
1. Future research areas:
Firstly, consistent monitoring of safety performance measurementis necessary to

justify the model, becausethe essenceof proactive safety managementis to measure
safety performancein operation on a regular basis, instead of measuringerrors or the

occurrencesof incidents or accidents,which are used as in
measurement retroactive
safety management.
Secondly, it would be interesting to conduct other factor-structures on a regional or
country -wide scale. As Reason (1995a) has stated, "there is unlikely to be a single
universal set of indicators for all types of hazardous operations, one way of
communicating how safety health can be assessed is by listing the organisational

factors that are currently measured." Therefore, a listing of various organisational
factor sets, plus ANOVA analysis of factors, can provide a clear and complex view from
different angles and concerns.
Thirdly, asmentionedpreviously, there are severalchangemanagementtools in current
use in industry, and documentedin the literature. It is worth exploring tools other than
Six Sigma to integrate the identified organisationalfactors into airline operations,to
identify the actions (sub-factors)required for the improvement in daily operations(or
on a regular basis), and to suit the different needsof different companiesin different
regions.
Lastly, although a proactive safety conceptis emphasisedin this thesis, there is still a
needto developthe retroactive approachesto safety,the conceptsandmethodologiesof
which are describedin Chapter 2. Becauseretroactive safety has

management proved
fruitful for the entire airline industry, its value is without doubt. For example,perhaps
accident investigation must improve in its global consistency. Organisations must

however learn form the misfortune even if the accidentreport is not aimed at them.
250
2. The airline safety management system:
Due to the variability of the airline industry; individual airlines should focus on
conducting safety surveys and their analysis to decide which of the factors are most
critical to them. It is worthy noting that although English is an international language, it
is recommended to use their native languages to design the questions in order to ensure
the respondents fully understand the meaning and implication, when conducting the
safety survey. By doing so, the most critical (important) areas in the airline can be
identified (it is even better to conduct deep dialogue interviews to augment the survey
analysis).
One particular factor was also identified as critical to perceived safety performance,
namely the barrier of organisational learning. This fact has two implications: (1) if the
barriers confronted by the company are not removed, safety performance is unable to
improve. So barriers of organisational learning are critical to airline safety
performance. (2) to remove the barriers of organisational learning relies on change
management, as discussed in Section 2.3.2.4.
The latter implication echoesandjustifies the tools adoptedby the researchto translate
the factors into the company changemanagementtools. Yet it is worth noting that in
Section 2.3.2.4, changemanagementis stated as the result of organisationallearning,
which is attributed to feedback in risk management,or broadly speaking, the safety

management system. In other words, it is "mistakes" that drives the learning and
change. This approachmatches up with the retroactive concept. Compared with the
changemanagementtools adopted in the safety mechanism model, the latter help to
identify the defects proactively and correct them in advance before they become
"mistakes". Both of them, however, are equally important to airline operations. As
such, airline managersneed to evaluate and distinguish them carefully in order to
choosesuitable change tools

management for airline operation.
Moreover, the suggested change management tools require the cross function
co-operationbetweendifferent departments,similar to BA's exampleregarding culture
251
andRecommendations
change (Appendix G). The thesis has also discussed the need for a complex interaction
across all departments when addressing flight safety risks and business risks within the
airline industry, and the interdepartmental training programmes associated with
proactive training for employees in responding to assaults or potential assaults in

Chapter 2. To this end, the systemic approach requires that organisational change is
brought about by an improvement in organisational factors. Once this is done, the
proactive safety mechanism will become stronger and stronger.
Furthermore, one point worth emphasising is the fact that the identified internal
organisationalfactors are all found to be significantly correlatedwith the impact of the

regulatory environment. The role of the regulator and its relationship with airlines are
described in Chapter 2 and Chapter 4 (the results of the interviews), where also
identifies that airline safety climate is governed by the regulatory environment and
regulators.It is obvious that the recent deregulation,globalisation and privatisation of

airlines have strengthenedthe role andresponsibilitiesof regulators,who provide many
functions to the airline, especially in terms of safety monitoring and safety inspection.
With their help, management's role is to develop a supportive system to identify
barriers proactively and retrospectively so that all employeescan strive for continuous
improvement.The aim of evolution within airline safety managementsystemcan then
be achieved.
Lastly, in order to fulfil the recommendationsto the airline industry and particularly to
safety managers, a framework is developed to recommend how an airline safety

management system may be benchmarked and improved by incorporating the research
results from the thesis.
In Chapter 4, the results of the interviews summariseda common framework of risk
assessmentin the airline industry, as presentedin Figure 4-2 (page 120). Basedon this
finding, Figure 7-1 adds the main results of this research containing the extent of
previous chaptersas well as the application and recommendationsmentioned in this

chapter.The establishmentof this figure provides the industry with a quick and clear
picture of the structure regarding retroactive and proactive managementin airline
252
safety managementsystem,which also allows various airlines to assessthemselvesby

comparingthemselveswith other airlines.
To the end, by further providing this framework, it is hoped that the airline industry and
particularly safety managers will become more aware of their positions and the
competitive advantages of improving safety when conducting airline safety

managementsystemand safety performance.
253
Figure 7-1 A Generic Framework for Retroactive and Proactive Approach
to Safety within Airline SMS
Airline safety management system

" Policies and principles/ Documentation/Accountabilities/ Training/ Communication
" Role of regulator: (Form safety climate)
Retroactiveapproach Proactive approach
Accidents/Incidents investigation: Cultural dimension

" Investigatestructural failure Organisationalclimate/culture,
" Identify human safety culture
erroneousbehaviour (human factors)
Risk..Mana
.....................
Hazards identification
- investigation
Incident
- Reporting system Result of incidents
- Auditing
Identify process/procedurerisks investigation
Assess risk and risk analysis

Risk assessment, Evaluate data, Trend analysis
Risk control (defences) and recovery

Internal surveys,
Risk programmes
"
..................................................................................................................................................................
Human and risk vs. Human, risk and culture vs.
Retroactive programme Proactiveprogramme
254
Develop safety survey based on the proactive

safety mechanism: Organisational
climate
Organisationalculture
Safety climate
Safety culture
Safety philosophy
Decision making
List Organisational factors:

Internalfactors-
1: Employeesafety attitude & behaviour
2: Employeesafety concept
3: Level of operational safety in
operation and maintenance
4: Corporate safety policy
5: Personnel - quality of working life
6: Employment of risk programme
7: Impact of accidentrncidents
8: Financialconcern
9: Procedures and documentation
10: Commercial cost pressures
11: Organisationalstructureand managementcommitment
12: Communicationsystem
13: Necessityof safety reports
Externalfactors-
1: Influences of region and country
2: Public and the media influence
3: Impact of regulatory environment
4: Involvementof investmentcommunity
Feed back and organisational change Change analysis

" Build safety case Incorporate suitable changemanagement
tools: e.g. Six Sigma, etc.
FOrganisational
learning and change management
e.g. Re-definepolicies and principles, evaluation system,
training process,accountabilities,etc
Safety Health= Retroactive approach+ Proactive approach

(safetyperformance) (retroactive+proactivesafetymanagement)
q.
255
Finale
In an attempt to provide transport and safety services, the airline industry involve an
array of equipment, high technology and supporting services, etc. The complexity of
this structure provides many opportunities for errors, some of which even result in
serious incidents or accidents. Though aircraft accidents occur very infrequently, they
undoubtedly have a high impact and pose a severe threat to airline viability. To prevent
them from recurring, the industry throughout the world has invested substantial
amounts of time and money in discovering the causes of accidents and incidents. In
recent years, the airline industry has begun to analyse hazardous situations and to
correct them before they result in accidents or incidents. However, even with this effort
and the advance of technology, accidents still occur, and with each accident, public
fears about air safety are fuelled.
Accidents/incidentsbring risk and fear into our lives, so people aremotivated to satisfy
thesesafety needsand searchthe reasonswhy they went wrong, which is the sprit of
retroactive safety management. If this need is not met (e.g. to improve the reliability of
aircraft), people will be stuck at this level, and will feel that somethingis lacking. Yet,
the cause of dissatisfaction is easy to identify (e.g. causesof accidents/ incidents,
human factors, etc.).
On the other hand, proactive safety management is like reaching the top level of
self-actualisation. At this level, people start to pursue the self-actualisation needs,
which are "the desire to become more and more what one is, to become everything that
one is capable of becoming. " As though built into the human gene, everyone in the
airline has the necessary concepts, attitudes, behaviours, knowledge, decision-making,
etc. orientated towards safety.
One metaphor is that retroactive approaches to safety are like plastic surgery. One can
actually become more beautiful after the surgery but the fundamental genetic makeup is
never changed. The way one thinks and behaves still remains the same as before.
Offspring will inherit this genetic makeup, which may become a family trait.
256
Conversely,proactive approachesto safety are more like genetic engineeringbefore a

baby is even born, in order to create the right personality. This is the essenceof
Therefore
proactive safety management. in summary,the ultimate contribution of this
researchis to provide airlines with reliable data, applicable referencesand practicable
methodology to enable their safety managementsystem to evolve at a fundamental,
"genetic" level.
257
258
References-
Aarons, R. N., Safety Cultures Vital in Business Aviation. Business & Commercial
Aviation, September 1998,9.
Abeyratne, R. I. R., The regulatory management of safety in air transport. Journal ofAir
Transport Management. 4(1), 1998,25-37.
Adams, G B. and Ingersoll, V. H., Painting over old works: The culture of organisations
in an age of technical rationality. In: Organisational symbolism, Turner, B. A., (Eds),
Berlin: Walther De Gruyter, 1989.
Airbus Industrie, Global Market Forecast 1995-2014: a long range view. Blagnac: Airbus
Industrie, 1995.
Airline Business, The world airline ranking. 17(9), September 2001,57-92.
AIR SAFETY WEEK, Next year "crucial " for safer skies accident reduction program.
15 (14), April 2,2001,1-4.
Air Transport Intelligence (ATI), Online Database.http://www. rati.com
Air Transport World,World airline report. 38(7), July 2001,49-150.
Akhurst, D. and Vivian, M., Themanagementof systemsfor safety and the virtual airline.
Flight Safety Foundation/InternationalFederationof Airworthiness in conjunction with
the InternationalAir TransportAssociation 50"' Annual International Air Safety Seminar.
Arlington, Virginia, November 6,1997.
Al-Harabi, A., Aviation maintenancein multicultural settings: the challengesof cultural

tolerance and of employees'maintenanceresourcemanagementhumanfactors
awareness, Ph. D. Thesis, Cranfield University, College of Aeronautics,Air Transport,
2001.
Arbon, E. R., Mouden, L. H. and Feeler,R. A., ThePractice of Aviation Safety.Flight

Safety Foundation, FSF/CP-90/12,June 1990.
Bacchi, M., Cacciabue,C. and O'Connor, S., Reactiveand proactive methodsfor human
factors studies in aviation maintenance.9`hSymposiumon Aviation Psychology,Ohio,
USA, 27 April -1 May 1997.
Bailey, C. W., and Peterson,D., Using perception surveysto assesssystem effectiveness.

Professional Safety,February, 1989,22-26.
Barnett, A. and Higgins, M. K., Airline safety: the last decade.ManagementScience.

35(l), 1989,1-21.
259
References
Barnett,A. and Wang,A., Airline Safety: TheRecentRecord.MassachusettsInstitute of

Technology,Cambridge,Massachusetts,1998.
Beaty, D., The naked pilot: the human factor in aircraft accidents. Shrewsbury: Airlife,
1995.
Becker,T. A., PassengerPerceptionsof Airline Safety: Marketing Safety Records.Flight

SafetyFoundation,Flight SafetyDigest, October 1992.
Berendsen,R., TheRole of International Standardsand Guidelinesfor Achieving System

Safety.3rdAnnual Conferenceon Aviation Safety Management,London, 22-23 May,
2000.
Bisson, T., A Corporate Approach to Safety Management. Flight Safety Foundation 50th
IASS, Washington DC, U. S., 4 November, 1997.
Boeing, unpublished.Boeing SafetyProgram Model. Seattle,Boeing Airplane Group.
Boeing, Current Market Outlook, Boeing Airplane Group, June2001.
Braithwaite, G R., and Caves,R. E., Airline safety - somelessonsfrom Australia. The
Aeronautical Journal, January 1997.
Brown, R. L and Holmes, H., The use of a factor-analysisprocedurefor assessingthe

validity of an employeesafety climate model. Accident Analysis and Prevention, 18(6),
1986,455-470.
Budworth, N., Strengtheningthe voice of the health of safetyprofessional in thepolicy

process. Proceedingsof the RoSPAAnnual Congress, 1996.
Bureau of Air Safety Investigation (BASI), An Evaluation of the BASI-INDICATE Safety

Programme, Canberra, 1998.
Cabrera, D. D., Isla, R., and Vilela, L. D, An evaluation of safety climate in ground
handling activities. In: Soekkha, H. M. (Ed. ), Aviation Safety, Proceedings of the
IASC-97 International Aviation Safety Conference, Netherlands, 27-29 August, 1997.
Chang, Y. 'n%! NM (in English: Air Transportation Management), 1St

edition, Taipei: Ding Hang, 1998,2-14.
Chang, Y. C., Influence of airline ownership rules on aviation policies and carrier
strategies, PhD thesis, Cranfield University, 2002.
Chapman, I., Maintenance Error Management Theory, MEMS Workshop, UK CAA,

2000.
260
References
Choularton, R., Complex learning: Organisational learning from disasters. Safety Science,
39 (2001), 61-70,2001.
Civil Aviation Safety Authority (CASA), Aviation Safety Management- An operator

guide to building a safety programme, Australia, 1998.
Collins ConciseDictionary & Thesaurus,New edition, Glasgow,UK: Harper Collins

Publishers, 1995.
Cooper, M. D., and Philips, R. A., Validation of a safety climate measure. Paper presented
at the British Psychological Society, Annual Occupational Psychology Conference,
Birmingham, 3-5 January, 1994.
Cox, S. and Cox, T., The structureof employeeattitudesto safety: a Europeanexample.

Workand Stress,5(2), 1991,93-106.
Curtis, T., Airline accidents and media bias: New York Times 1978-1994, (1997).
http: //airsafe. com/nyt bias. htm
Dannatt, R., TheRole and Effectivenessof GovernmentRegulation ofAir Transport

Safety.4thAir Transport ResearchGroup Conference(ATRG), 2-4 July, Amsterdam,
Netherlands,2000.
Darlington, R. B., Factor Analysis. http://comp9.psych.comell.edu/Darlington/factor.htm,

2002.
Dedobbeleer,N., Beland, F., and German,P.,Is there a relationship betweenattributes of

construction sites and workers safetypractices and climate perceptions ? In : Das, B.
(Ed.), Advancesin Industrial Ergonomics and Safety II, London : Taylor and Francis,
1990,725-732.
De Cock, G, Bouwen, R., de Witte, K., Organisatieklimaat:Een opdrachtvoor het

personeelsbeleid?Praktisch Personeelsbeleid,Capita Selecta 16,1986,1-20.
Degani, A. and Wiener, E. L., Philosophy,policies, proceduresand practices:The four

`P's of flight deck operations.In: Aviation Psychology in Practice, eds.Johnson,N.,
McDonald, N. and Fuller, R., Aldershot, UK: Ashgate, 1994,44-67.
Doganis, R., Theairline businessin the twenty-first century. 1" edition, London:
Routledge,2001.
Doganis, R., Airline trends, challengesand prospects.Visiting lecture note. Air transport
Group, College of Aeronautics, Cranfield University, November 2001a.
.... 261 ýý....... _..__._..

References
Doganis, R., Flying off course, the economics of international airlines. 3Tdedition,
London: Routledge,2002.
Donald, I. And Cantor,D., Attitudes to Safety: PsychologicalFactorsand the Accidents

Plateau.Industrial Reviewand Report, 347,1993,3-8.
Dumas,R., Safety and Quality: the human dimension,ProfessionalSafety,32 (12),

December,1987.
Edkins, G. D., The INDICATE safetyprogram; evaluation of a method to proactively

improve airline safety performance.SafetyScience.30,1998,275-295.
Edwards,C. J., Developing of safety casewith an aircraft operator. IBC Conferenceon

Aviation SafetyManagement,London, 20-21 May, 1999.
Edwards, C. J., Developing of a hazard model for an aviation safety case. FSF, IFA &
IATA Joint Meeting on "Enhancing safety in the 21st century", Rio de Janeiro, Brazil,
November, 1999a.
Earnest,R. E., Characteristicsof proactive & reactive safety systems.ProfessionalSafety,

42 (11), 1997,27-29.
FederalAviation Administration (FAA), Human FactorsGuide for Aviation Maintenance,

1995.
FederalAviation Administration (FAA), FederalAviation Requirements(FAR) Part 121,

1996.
FederalAviation Administration (FAA), StrategyDocument, 1997a.
FederalAviation Administration (FAA), Human FactorsIssuesin Aviation Maintenance

and Inspection,Aviation Medicine, 1997b.
Ferry, S. T., Modern accident investigation and analysis. The USA: Wiley & Sons, 1988.
Fischoff, B., Acceptable risk: a conceptualapproach.Risk: Health, Safety& Environment,

5 (1), 1994,1-28.
Fitts, P. M. and Jones, R. E., Analysis of factors contributing to 460 `pilot error"
experiences on operating aircraft controls (TSEAA-694-12). Wright-Patterson AFB, OH:
Aero medical Laboratory, 1947.
Flight Safety Foundation (FSF), The Dollars and Sense of Risk Management and Airline
Safety, Flight Safety Digest, May 1999a.
262
References
Flight Safety Foundation (FSF) Icarus Committee, Airline Safety: Airline Management
Self-Audit, Flight Safety Digest, May 1999b.
Flin, R., Meams, K., O'Connor, P., and Bryden, R., Measuring safety climate: identifying
the common features.SafetyScience,34(2000), 2000,177-192.
Furnham,A., and Gunter,B., Corporate assessment:auditing a company'spersonality.

London: Routledge, 1993.
Geller, E. S., Ten Principles for Achieving a Total Safety Culture. ProfessionalSafety.
September,1994,18-24.
Glennon, D. P.,Measuring organizationalsafety climate. Australian SafetyNews, 23-28

January 1982.
Glendon, A. I., Stanton, N. A., and Harrison, D., Factor analysing a performance shaping
concepts questionnaire. In: Roberton, S, A. (Ed. ), Contemporary Ergonomics 1994:
Ergonomics for all. London: Taylor& Francis, 1994,304-345.
Glendon,A. I., and Stanton,N. A., Perspectiveson safety culture. SafetyScience,34,

2000,193-214.
Glendon,A. I., and Litherland, D. K., Safety climate factors, group differencesand safety
behaviour in road construction.SafetyScience,39(2001), 2001,157-188.
Global Aviation Information Network (GAIN), Operator's Flight SafetyHandbook. FAA,

Issue 1, June 2000.
Goglia, J., NTSB Report, the FourteenthAnnual Symposiumon Human Factors in

Aviation Maintenance,Vancouver,March 38-30 2000.
Green,R. G., Muir, H., James,M., Gradwell, D. and Green,R., Human Factors for
Pilots. 2nded., London: Avebury, 1996.
Grimaldi, J. V. and Simonds,R. H., SafetyManagement,Homewood, Illinois: Richard D.

Irwin Inc., 1984.
Grose,V. L., Managing risk: systematiclosspreventionfor executives.Englewood, Cliffs,

NJ: Prentice Hall, 1987.
Guldenmund,F. W., The nature of safety culture: a review of theory and research.Safety
Science,34 (2000), 2000,215-257.
Hague,P. N., QuestionnaireDesign. London: Kogan Page, 1993.
......,ý_
a .. _. 263 _,_.... _ _..... __. .
References
Hair, J. F., Anderson, R. E., Tatham, R. L., Black, W. C., Multivariate Data Analysis with
Readings. 4th edition, New Jersey: Prentice-Hall, 1995.
Hale, A. R. and Glendon, I. A., Individual Behavior in the Control of Danger. Amsterdam:
Elsevier, 1987.
Hamilton, C., "Back to Basics" SafetyAudits, 10`hEASS "Managing Aviation Safety-

Back to Basics", Flight Safety foundation (FSF), Amsterdam,Netherlands,March 1998.
Hammer,W., Handbook of systemandproduct safety.New JerseyUSA: Prentice

Hall, 1972.
Hampden-Turner,C., Corporate Culture, London: Piatkus, 1994.
Handy, C. B., Understanding Organisations, 3rdEd., London: Penguin Books, 1985.
Hart, C., The Global analysisand information network (GAIN),

http://www.nasdae.faa.gov/GAIN/presentation/Chris Hart sneech.htm
(30 January,1999)
Hawkins, F. H., Human Factors in Flight. Aldershot: Ashgate,2nd edition edited by

Harry W. Orlady,1993.
Hayward, B., Culture, CRM and aviation safety,Aviation Safety for the 21st Century in
The Asia Pacific Region, Asia Pacific Regional Seminar.The Australian and New
Zealand Societiesof Air Safety Investigators(ANZSASI), Brisbane,Australia, 29-31
May 1997.
Head,T., Report into Glider conspicuity trail held at RAF Bicester in June and
October 2002. Human Factor Group, Cranfield University, 2003.
Heinrich, H. W., Petersen,D., and Roos,N., Industrial accidentprevention: A safety

managementapproach. 5`hed., New York: McGraw-Hill, 1980.
Heimreich, R. L., Managing human error in aviation. Scientific American. 276( 5), 1997,
62-67.
Heimreich, R. L., Culture threat, and error: assessingsystemsafety.Safety in Airlines-

The ManagementCommitment, Joint conferenceof the Royal Aeronautical Society and
the Guild of Air Pilots andAir Navigators, 3-4 June 1999,London, UK. 3.1-3.10.
Heimreich, R. L., A closer inspection: what really happensin the cockpit, Flight safety
Australia, January-February2001,32-35.
264
References
Heimreich, R. L. and Wilhelm J. A., CRM and Culture: National, Professional,

Organ isational, Safety. 9"' Symposium on Aviation Psychology, Ohio, USA, 27 April -1
May 1997.
Herrero, S. G, Saldana,M. A., Campo,M. A., and Ritzel, D. 0., From the traditional
concept of safety management to safety integratedwith quality. Journal of Safety
Research,33 (2002), 1-20,2002.
Ho, L. C., A critical analysis of airline safety management with reference to pilots and
aviation authority officers. Ph.D. thesis, Cranfield University, UK, 1996.
Hobbs, A., Maintenance human factors: Learning from errors to improve systems, in B. J.
Hayward and A. R. Lowe (eds) Aviation Resource Management, Vol. 1, Aldershot:
Ashgate, 2000,347 - 355.
Hofstede, G, Culture's Consequences:International Differencesin Work-RelatedValues.

Beverly Hills, CA: Sage,1980.
Hofstede, G, Cultures and Organisations: Software of the mind: Intercultural

Cooperations and Its Importance for Survival. London: McGraw-Hill, 1994.
Hollnagel E., Human Reliability Analysis-Contextand Control. London: Academic Press,

1993.
Holtom, M., FOQA: Aviation's Most Importance Safety Tool. "Enhancing Safety in the
2 1s`Century", FSF,IFA and IATA Joint Meeting, Rio de Janeiro,Brazil, November 1999.
Hood, C., Jones,D. K., Pidgeon,N. F., Turner, B. A. and Gibson, R., Risk management.
In: Risk: Analysis,perception and management,London: The Royal Society, 1992.
Hsu, Y. L, An Analysis of the Impact ofAccidents on Airline Performance. MSc thesis,

Cranfield University, UK 1999.
International Air TransportAssociation (IATA), Safety in Operations,The 20thTechnical

Conferenceof IATA, Istanbul, 10-14November 1975.
International Air TransportAssociation (IATA), PassengerForecast 1998-2002,Geneva,

Switzerland, 1998.
International Air TransportAssociation (IATA), Aviation Information and Research,

September2001 Industrie Briefing, Geneva,Switzerland, 2001.
International Air TransportAssociation (IATA), PassengerForecast 2001-2005,Geneva,

Switzerland, 2002.
,.,._._... 265
References
International Civil Aviation Organisation ( ICAO), Accident Prevention Manual ICAO,

Montreal, Canada,1984.
International Civil Aviation Organisation(ICAO), Human Factors Digest No.], ICAO,

Montreal, Canada,1989.
International Civil Aviation Organisation(ICAO), Circular 247, Human Factors Digest

No.10: Human Factors, Managementand Organisation, ICAO, Montreal, Canada1994.
International Civil Aviation Organisation(ICAO), Human Factors Training Manual.

(Doc 9683) ICAO, Montreal, Canada,1998.
International Federationof Airworthiness (IFA), Engineering Solutions to Human

Problems,videos and training package,1998.
Janic, M., An assessmentof risk and safety in civil aviation. Journal ofAir Transport
Management,6 (2000), 2000,43-50.
Johnson,D., Turning safety on its head.Industrial Safety& Hygiene News. October

1994.
Johnson,G, and Scholes,K., Exploring Corporate Strategy.2nd Ed., New York London:
Prentice-Hall, 1988.
Johnston, N., CRM: Cross-cultural perspectives. In: Cockpit resource management.

Wiener, E. L., B. G, Kanki and Heimreich, R. L., (Eds.) SanDiego, CA: Academic Press.
1993.
Johnston, N., Managing risk in flight operations. In: Applied Aviation Psychology, Ed.
Hayward, B. J., H. and Lowe, A. R (Eds.). Aldershot: Ashgate, 1996,1-19.
Joint Airworthiness Association (JAA), Joint Airworthiness Requirements(JAR) 25.1309

Large Aeroplanes,July 1996.
Joint Airworthiness Association (JAA), Joint Airworthiness RequirementsJAR-OPS 1

CommercialAir Transportation(Aeroplanes).March, 1998.
Kuang Fu Publishing, SI AN O,3<A (in English: Thosewhich transportationcreates

history), 1988.
Lauber, J., Safety cultures and the importanceof human factors. The CRMAdvocate, 94
(4), 1994,1-3.
Lauber, J., In: The Fallible Human.AEROSPACE,August 1996.
266
References
Laurent, A., Human Resource Management in International Firms, Ed., Evan, P., London:
Macmillan, 1990.
Lawrence, B. M., Managing Safety Through the Aircraft Lifecycle- An Aircraft

Manufacturers Perspective. Aviation Safety Management Conference, 20-21 May,
London, 1999.
Learmount,D., Safety.Flight International, 2 Jan,2001,64.
Lee, T., Perceptions,attitudesand behaviour: the vital elementsof a safety culture. Health
and Safety, October, 1996,1-15.
Lowrance, W. W., Of acceptablerisk. William Kaufmann 1976.
Manzella, J.C., Achieving safety performanceexcellencethrough total quality management.

Professional Safety,42(5), 1997,26- 28.
Marquardt, M. and Reynolds,A., The Global Learning Organization. Gaining

CompetitiveAdvantage Through Continuouslearning. Burr Ridge, III: Irwin
ProfessionalPublishing, 1993.
Maurino, D. E. Cross cultural perspectivesin human factors training: Lessonsfrom the

ICAO Human FactorsProgramme.TheInternational Journal ofAviation Psychology,
4(2), 1994,173-181.
Maurino, D. E., Reason,R., JohnstonN., and Lee, R. B., BeyondAviation Human

Factors. London: Ashgate Publishing, 1995.
Maurino, D. E., In: The fallible human,AEROSPACE,August 1996.
Maurino, D. E., CRM. Past, Presentand Future. Paperpresentedin 9thSymposiumon

Aviation Psychology,Ohio, USA, 27 April -1 May 1997.
Maurino, D. E., Tired of sweepingup at the end of the parade,Flight SafetyAustralia,

January-February2001,36-39.
McDonald, N. and Ryan, F., Constraintson the developmentof safety culture: A

preliminary analysis. The Irish Journal of Psychology, 13(2), 1992,273-281.
McFadden,K. L., and Towell, E. R., Aviation human factors: a framework for the new
millennium. Journal ofAir TransportManagement,5,1999,177-184.
McIntyre, G. R., Patterns in safety thinking. Aldershot: Ashgate, 2000.

.
267
References
McIntyre, G. R., The application of systemsafety engineeringandmanagement

techniquesat the US FederalAviation Administration (FAA). SafetyScience,40,2002,
325-335.
Merritt A., Replicating Hofstede: A study of pilots in eighteen countries. Paper presented
in 9th Symposium on Aviation Psychology, Ohio, USA, 27 April-1 May 1997.
Merritt A. and Heimreich, R. L., Creating and sustaininga safety culture: somepractical
strategies.In: Applied Aviation Psychology,Ed. Hayward, B. J., H. and Lowe, A. R.
Aldershot: Ashgate, 1996,20-26.
Morley, F. J. J., Ripples in a pond: a comprehensive,generalisedmodel of the evolution

of safety culture. PhD thesis, Cranfield University, 1999.
Morrison, S and Winston, C., Air Safetyderegulationand public policy. TheBookings

Review,6,10-14, Winter, 1988.
Moses, L.M., and Savage,I. (eds.), TransportationSafetyin an Age of Deregulation. New

York: Oxford University Press,1990.
Muir, H. and Thomas,L., Passengersafety in very large transport aircraft. The

Aeronautical Journal, April, 2003,201-206.
Nagel, D. C., Human error in aviation operations.In: Humanfactors in aviation. Ed.

Nagel, C., and Wiener, E. L., SanDiego: Academic Press,Inc., 1988.
National TransportationSafety Board (NTSB), Aircraft Accident Report (AAR-97/06):

In-Flight Fire and Impact with Terrain, ValuJetAirlines, Flight 592, DC-9-32, N904 Vi
Everglades,Near Miami, Florida. May 11,1996. NTSB, Washington,1997.
National TransportationSafety Board (NTSB), Aircraft Accident Report (DCA82AAO11),

Air Florida, Flight 90, B737, Washington,January 13,1982. NTSB, Washington,1983.
Neal, A., Griffin, M. A., and Hart, P. M., The impact of organizationalclimate on safety
climate and individual behavior. Safety Science,34,2000,99-109.
Niskanen,T., Safety climate in the road administration.SafetyScience17,1994,237-255.
Overall, M., Managing SafetyManagement.Effective Safety,IBC Conferencefor Health

& Environment Management,London, 20-23 September,1999.
Panes, J., Evolution of the Aviation Safety Paradigm: Towards Systemic Causality and.
Proactive Actions. In: Applied Aviation Psychology, Ed. Hayward, B. J., H. and Lowe, A.
R. Aldershot: Ashgate, 1996,39-49.
268
References
Pasman, H. J., Risk informed resource allocation policy: safety can save costs. Journal of
Hazardous Material, 71 (2000), 2000,375-394.
Perrow, C., Normal Accidents: Living WithHigh-Risk Technologies.New York: Basic

Books, 1984.
Petersen,D., What should be use and why. Professional Safety,43 (10), 1998.
Pidgeon,N., Safety culture and risk managementin organization.Journal of

Cross-Cultural Psychology22(1), 1991,129-140.
Pidgeon,N., and O'Leary, M., Man-madedisasters:why technology and organizations

(sometimes)fail, SafetyScience,34,2000,15-30.
Profit, R., Systematic safety management in the air traffic services. London: Euromoney,
1995.
Rasmussen,J., Trendsin human reliability analysis.Ergonomics, 28 (8), 1985,

1185-1196.
Reason,J. Human Factor. UK: CambridgeUniversity Press,1990.
Reason,J., Comprehensiveerror managementin aircraft engineering.Human Factors in

Maintenance.TheAerotech 94 Congress.London: Institution of Mechanical Engineers,
1994.
Reason,J., A systemsapproachto organizational error.ERGONOMICS. 38( 8), 1995a,

1708-1721.
Reason,J, In: Beyond aviation humanfactors: safety in high technologysystems.Edited

by D Maurino, N Johnston,J. Reasonand B. Lee, Aldershot: Avebury Aviation, 1995b.
Reason,J., Managing the risks of organizational accidents,Aldershot: Ashgate, 1997.
Robbins, S. P.,Essentialsof Organisational Behaviour. 3rdedition, New Jersey:Prentice-

Hall, 1992.
Rohner,R. P.,Towardsa conceptionof culture for cross-culturalpsychology.Journal of

Cross-Cultural Psychology, 15,1984,111-138.
Rose, L. N., Profitability and product quality: Financial indicators and airline accident
performance. Sloan School of Management, Massachusetts Institute of Technology, MA.
1988.
Rose, L. N., Fear of flying? Economic analyses of airline safety. Journal of Economic
Perspectives, 6(2), Spring 1992,75-94.
.. W..,__.. 269 ..... ýý... ,

References
Sadgrove,K., Completeguide to businessrisk management,Aldershot: Gower, 1996.
Sales, E., Prince, C., Browers, C. A., Stout, R. J., Oster, R. L. and Cannon-Bowers, J. A.,
A Methodology for enhancing crew resource management training. Human Factors.
41(l), 1999,161-172.
Savage,J., The Useof Flight Data in Airline SafetyManagement,Joint conferenceof the

Royal Aeronautical Society and the Guild of Air pilots andAir Navigators- Safety in
Airline- The Managementcommitment. 3-4 June, 1999.
Sayce,A., Interview notes of Safety Regulation Group, UKCAA. December2001.
Schein, E. H., Organisational culture and leadership. 2°d edition, San Francisco:
Jossey-Bass, 1992.
Schneider,B., OrganisationalClimates: an essay.Personal Psychology28,1975,

447-479.
Shrivastava, P., A Typology of Organisational Learning Systems. Journal of Management

Studies, 21(1), 1983,7-28.
Sharples,C., An airline's view of airline safety management,Focus on Commercial

Aviation Safety,24, Autumn 1996,12-16.
Shorrock, S. T and Kirwan, B., Developmentand application of a human error

identification tool for air traffic control. Applied Ergonomics,33 (2002), 2002,319-336.
Smircich, L., Conceptsof culture and organisatioanlanalysis.Administrative Science

Quarterly, 28 (1983), 1983,339-358.
Smith, C., High risk business.Airline Business,12(11),November 1996,68-69.
Smith, T. A., Will safety be ready for workplace 2000. Professionalsafety, 41(2), 1996,
37-38.
Smith, D., Making senseof disaster : issuesin organisational learning. IBC Conference
on Aviation Safety Management, London, 20-21 May, 1999.
Stacey,R. D., Strategic Managementand Organisational Dynamics. London: Pitman,

1993.
Taylor, A. F., Human Factors. Lecture note from the short course"Accident
Investigation", 1999.
270
References
Taylor, F. and Y. L. Hsu, Evaluation of reactive and proactive approaches to safety within
the airline industry, paper presented at the Fourth Pacific International Conference on
Aerospace Science and Technology (PICAST 4), Taiwan, 21-23 May, 2001a.
Taylor, F. and Y. L. Hsu, The Impact ofAccidents on Airline Performance. Paper

published and allowed to present in the 9th WCTR Conference, Seoul, July 22-27 2001b.
TheEconomist,How safeis your airlines? 342 (7999). January 11th-17`h,1997,15.
TransportCanada,Risk Managementand Decision-making in Civil Aviation, TP 13095,

Canada,April 2001a.
TransportCanada,Introduction to SafetyManagementSystems,TP 13739,Canada,April

2001b.
Turner, B. A., Exploring the Industrial Subculture.London: Macmillan, 1971.
UKCAA (Civil Aviation Authority in United Kingdom), Human Factors and Aircraft
MaintenanceHandbook. Issue2, UK, August 2000.
UKCAA (Civil Aviation Authority in United Kingdom), CAP 712: SafetyManagement

Systemsfor CommercialAir Transport Operations.UK, May 2001.
Warner,F., Introduction. In: Risk: Analysis,perception and management,London: The

Royal Society, 1992.
Watkins, M., Investigation of the Impact of the Airbus A3XX on Airport Operation. MSc
Thesis, Cranfield University, 1997.
Wells, A., CommercialAviation Safety,Blue Ridge Summit, Pa: TAB Books, 1991.
Westrum,R., Cultures with requisite imagination. In: Wise, J., Hopkin, V. D. and Stager,
P. (Eds.), Verification and validation of complexsystems:Humanfactors issues.Berlin:
Springer-Verlag,1993.
Westrum,R., Organisationaldynamics and safety. In: McDonald, N., JohnstonN. and

Stager,P. (Eds.), Application of psychology to the aviation system.Aldershot, UK:
Avebury, 1995.
Westrum,R., Human factors expertsbeginning to focus on organizational factors in

safety,ICAO Journal, October 1996.
Whorton, J. W. and Worthley, J. A., A perspectiveon the challenge of public management:

Environmental paradox and organisationalculture. Academyof ManagementReview,6,
1981,357-363.
271
References
Wiener, E. L., Intervention Strategies for the Management of Human Error, Flight Safety
Digest, February, 1995.
Wiener, E. L., Chidester, T. R., Kanki, B. G, Palmer, E. A., Curry, R. E. and Gregorich, S.
E., The impact of cockpit automation on crew coordination and communication: I.
Overview, LOFT evaluations, error severity and questionnaire data. NASA Report
CR-177587, Moffett Field, CA, NASA-Ames Research Center, 1991.
Williamson, A. M., Feyer,A. M., Cairns,D., and Biancotti, D., The developmentof a
measureof safety climate: the role of safety perceptionsand attitudes.SafetyScience25,
1997,15-27.
Woods,H. R., Aviation SafetyPrograms. USA: IAP, Inc., 1992.
Zohar, D., Safety climate in industrial organizations:theoretical and applied implications.

Journal ofApplied Psychology,65(1), 1980,96-102.
272
Bibliography
Air Transport World, Sorting Out the Answers. 30(4), April 1993,82-84.
Chan, H., Can airline public relations activities suppress the effect of negative
publicity caused by safety incidents? MSc thesis, Cranfield University, 2001.
Maslow, A. H., Motivation and Personality. 2nded., New York: Hamper and Row,
1970.
Ale, B., Smith, E., and Pitblado, R., Safetyaround airports- Developmentsin the
1990sand the Future Directions. P Annual Conferenceon Aviation Safety
Management,London, 22-23 May, 2000.
Bailey, J., Could you pleaseconfirm......? Airline Business,15(1), 1999,36-37.
Barnett, A., Abraham, M., Schimmel, V., Airline safety: some empirical findings.
ManagementScience,25(11), 1979,1045-1056.
Boeing, Statistical Summary of Commercial Jet Airplane Accidents - Worldwide

Operations 1959-2002, Boeing Airplane Group, May 2003.
Braithwaite, G R., Caves,R.E., and Faulkner, J.P.E., Australian aviation safety -

observations from the `lucky' country. Journal ofAir Transport Management, 4,
1998,55-62.
Cooper, D. Measurement of Safety Climate: A component analysis, paper

presented at IOSH meeting, February 1, London, 1995.
Cooper, W. E., CFIT: How do we terrain-proof our pilots. In: Applied Aviation
Psychology, Ed. Hayward, B. J., H. and Lowe, A. R. Aldershot: Ashgate,
1996,89-96.
Crosby,P. B., Lets talk quality. New York: McGraw-Hill, 1989.
Deacon,A., The role of safety in total quality management.TheSafetyand Health

Practitioner, 12(1), 1994,18-21.
DO 178B/ED-12B Software Considerations in Airborne Systems and Equipment

Certification, 1992.
Doganis, R., The airport business. London: Routledge, 1998.
273.., ý_ý,
Bibliography
Donoghue, J. A., Safety responsibility. Air Transport World, 32 (2), 1995.
Edkins, G. D., Reducingoperating coststhroughproactive safety: Lessonsfrom

Australia. Paperpresentedat the InternationalAir TransportAssociation "Inflight
Management99' Conference,Amman, Jordan3-5 May 1999.
Evan, J. R. and Lindsay, W. M., The management and control of quality. 3rd
edition, New York: West Publishing, 1995.
Fewings, R., Introduction to Airports, Lecture note. Air Transport Group, College
of Aeronautics, Cranfield University, 2001.
Garvin, D. A., How the Baldrige Award really works. Harvard BusinessReview,
69 (6), 1991,80-95.
Goetsch, D. L., Occupational safety and health for technologist, engineers and
managers. New Jersey: Prentice-Hall, 1999.
Goldbaum,E., Airline & Safety: Computerproject aims to prevent errors,

Reporter-StateUniversity of New Yorkat Buffalo, 28(14), December,1996.
Guild, S., Picking up the pieces.Airline Business,11(3),March 1995.
Hampden-Turner,C., Corporate Culturefor CompetitiveEdge.A user'sguide.

London: the Economist Publications, 1990.
Hudson, P.,Reason,J., Wagenaar,W., Bentley, P.,Primrose,M. and Visser,J.,

Tripod DELTA : proactive approachto enhancedsafety.Journal of Petroleum
Technology,46,1994,58-62.
Hudson, P.,van der Graaf, G C., andVerschuur,W. L. G, Perceptionsof

Proceduresby Operators and Supervisors.HSE Conferenceof the Society of
PetroleumEngineers,Caracas,1998.
Hughes,C. S., Managing Quality. 79th Annual Meeting of Transportation

ResearchBoard (TRB), WashingtonDC, 9-13 January,2000.
Keating, S. D., How can aviation insurance underwriting be made more

profitable?
MSc thesis, Air Transport Group, Cranfield University, 1994.
Kotler, P., Armstrong, G, Saunders. J., and Wong V., Principles of marketing. 2nded.,
London: Prentice Hall Europe, 1999.
274
Bibliography
Kotter, P. J. and Heskett, L. J., Corporate culture and performance. New York:
The Free Press, 1992.
Lee, P. Y., The Price of Saving Passenger Lives with Reference to Asia Pacific and
Latin America. MSc thesis, Air Transport Group, Cranfield University, 2000.
Lyle, C., Guiding international civil aviation into the 21" century. Presentationof
ICAO's StrategicAction Plan, Montreal, 22 May 1997.
McDonald, N., Corrigan, S. Daly, C. and Cromie, S., Safety managementsystems

in
and safety culture aircraft maintenanceorganisations.SafetyScience,34(2000),
2000,151-176.
McVicker, S., Practical Risk Assessmentfor Use in Safety Management Systems

for Small-to Medium-sized Airlines. 10thEASS -Managing Aviation Safety-Back
to Basics, Amsterdam, Netherlands, March 1998.
Morrell, P.,Airline Finance. 2nd ed., Aldershot: Ashgate,2002.
Moses, L. M., and Savage,I., The effect of airline pilot characteristicson

perceptionsof job safety risks. Journal of Risk and Uncertainty, 2,1989,335-351.
Nelms, D. W., PassengerViolence. Air Transport World, 35 (3), 1998,35-42
Philips, E. H., Cargo Carriers Seek Stronger Ties With MRO Providers, Aviation
Week & Space Technology, August 27,2001.
Porter, L. J. and Parker, A. J., Total quality management - the critical success
factors. Total Quality Management, 4 (1), 1993,13-22.
Roland, H. E., and Moriarty, B., System Safety Engineering and Management. 2nd
edition, New York: John Wiley & Sons, Inc., 1990.
Sagan,S. D., The limits of safety: organisations, accidentsand nuclear weapons.

New Jersey: PrincetonUniversity Press,1993.
Sage, A. P. and White, E. B., Methodologies for risk and hazard assessment: a
survey and status report. IEEE Transaction on System, Man, and Cybernetics
SMC-10,1980,425-441.
Salazar,N., Applying the Deming philosophy to the safety system.Professional

Safety.34 (12), 1989,22-27.
275
Bibliography
Samson, D. And Terziovski, M., The relationship between total quality
managementpractices and operationalperformance.Journal of Operations

Management,17(1999), 1999,393-409.
Saunders,I., Managing quality and risk. Journal of Occupatioanl Health and

Safety, 11(6), 1995,579-586.
Scherkenbach,W. W., Deming 'sRoad to Continual Improvement.Knoxville, TN :

SPCpress, 1991.
Schwartz,D., From safety indexing to risk management,FSF/ICARUS Committee

Report, 1998.
Senge, P. M., Thef fth discipline fieldbook: strategies and tools for building a
learning organization. London: Nicholas Brealey, 1994.
Shaw, S., Airline Marketing and Management. 3rd cd., London: Pitman, 1990.
Siomkos, G. J., Managing airline disaster:the role of consumersafety perceptions

and sense-making.Journal ofAir TransportManagement,6(2), 2000,101-108.
Slovic, P.,Perceptionof Risk. Science,vol 236.1987,280-285.
Smith, D., The Kegworth aircrash-a crisis in three phases? Disaster Management,
4(2), 1992,63-72.
Smith, T. A., Managing for continuousimprovementof safety performance-Focus

on the customer.Professional Safety,40,1995,18-23.
Taylor, A. F., Airline Crisis Management Training by Simulation. SAGSET

seminar paper, September 1997.
Transport Canada, Promotional video, 1998.
Yeahiya,M., Aviation insuranceand the trend in the aviation insurancemarket.

MSc thesis,Air Transport Group, Cranfield University, 1991.
276
APPENDIX A Definitions of Accident/ Incident
In the aviation industry, the most widely used definition is the one developed by the
International Civil Aviation Organisation (ICAO, Annex 13,1994). In order to
effectively prevent accidents and promote aviation safety, ICAO revised the content of
annex 13 in March of 1994. It suggests its state members to define "serious incident"
and to thoroughly investigate them. The current version revised in 2001 has clearly
defined the serious injuries.
AIRCRAFT ACCIDENT:
An occurrenceassociatedwith the operation of an aircraft which takes place

between the time any person boards the aircraft with the intention of flight
until such time as all such personshave disembarked,
in which:
(a) a person is fatally or seriously injured as a result of:

" being in the aircraft, or
" direct contact with any part of the aircraft, including parts which
have become detached from the aircraft, or
" direct exposure to jet blast;
EXCEPT when the injuries are from natural causes, self inflicted or inflicted
by other persons, or when the injures are to stowaways hiding
outside the areas normally available to the passengers and crew; or.
(b) the aircraft sustains damage or structural failure which:

" adversely affects the structural strength, performance or flight
characteristics of the aircraft and
" would normally require major or replacement of the affected
component.
EXCEPT for engine failure or damage, when the damage is limited to the
engine, its cowlings or accessories; or for damage limited to
propellers, wing tips, antenna, tires, brakes, fairings, small dents
or puncture holes in the aircraft skin; or
(c) the aircraft is missing or is completely inaccessible.
Note 1. For statistical uniformity only, an injury resulting in death within

thirty days of the date of the accident is classified as a fatal injury
by ICAO.
Note 2. An aircraft is considered to be missing when the official searchhas
been terminated and the wreckage has not been located.
277
Appendix A Definitions ofAccident/Incident
INCIDENT:
An occurrence,other than an accident,associatedwith the operationof an

aircraft which affectsor could affect the safetyof operation.
SERIOUS INCIDENT:
An incident involving circumstancesindicating that an accident nearly occurred.
Note 1. The difference between an accident and a serious incident lies only
in the result.
Note 2. Examples of serious incidents can be found in Attachment D of
Annex 13 and in the ICAO Accident/Incident Reporting Manual
(Dot 9156)
SERIOUS INJURIES:
An injury which is sustained by a person in an accident and which:

a) requires hospitalisation for more than 48 hours, commencing within
seven days from the date the injury was received; or
b) results in a fracture of any bone (expect simple fractures of fingers,
toes, or nose); or
c) involves lacerations which cause sever haemorrhage, nerve, muscle
or tendon damage; or
d) involves injury to any internal organ; or
e) involves second or third degree burns, or any burns affecting more
than 5 percent of the body surface; or
involves verified exposure to infectious substances or injurious
radiation.
278
APPENDIX B Impact of Accident on Airline Operation & Finance
Performance
B. 1 The Immediate Impact on Operation Performance

- Crisis Management
Crisis management is the "first aid" of the breach in the airline's safety services relationship,
because following an accident the airline must deal effectively and courteously with its
passengers and crew, with the victims, both alive and dead and also with their friends and
relatives. At the same time it will have to deal with the media and with business, political and
other pressures resulting from the accident (Taylor, 1997). These are the immediate effects
resulting from an accident. They not only instantly affect the airline's operation but also may
have a significant influence in the following undertaking.
The effects resulting from an accident i. e. death, injuries and property damage, the
relationship with media, public image and load factor will be discussed in the following
sections.
B. 1.1 Death, injuries, and property damage
When an airline is created, it is obligated to take on the responsibilities of transportation and

safety, which are worthless without each other, just like the industry's coins of value. As such,
once there is a failing on each side, the airline must be responsible for the outcome and
dealing with the following details: the death, injuries, and damaged property, which are the
immediate effects, resulting from the accident and the most threatening crisis for any airline.
Once an accident has occurred, the airline duty Operation Control Manager receives
information that something has gone wrong. According to the call-out list and designated
process, key personnel are informed with great urgency. Upon the notification of an accident,
various departments are involved and emergency facilities are activated in accordance with
Emergency Response Procedures (ERPs). Emergency Response Manuals vary with different
airlines and countries. However, in order to achieve an effective management, emergency
response procedures should at least cover the following: (see Figure B-1)
Figure B-1 Emergency Response P
Accident
Headquarter Aý,,
Relief Aircraft Local

CMC icident support Tea
LACC FRRC
SRC RA
ICC MIC PIC
TEC
Source: Adapted from Taylor, 1997
279
Appendix B Impact ofAccident on Airline Operation & Finance Performance
(1) The Headquarters Response - Crisis Management Centre (CMC) includes

a. Media Information Centre (MIC)
b. PassengerInformation Centre(PIC)
c. Incident Control Centre (ICC)
(2) The Local Response
a. The local Accident Control Centre (LA CC)
b. The Survivors Reception Centre (SRC)
c. The Friends and Relatives Reception Centre (FRRC)
d. TheReputationArea (RA)
e. Liaison with local hospitals and hotels
(3) The mobile response is involved in the activation of
a. the ReliefAircraft
b. the Incident Support Team(IS7)
In order to assist with local response,airlines have plans to dispatch an aircraft to carry
personnel and specialiststo the sceneof the accident. Most of the IST membersare trained
volunteerswho provide support to the airline local staff, and with the handling of the incident
and those involved.
The Emergency Response Procedures will inevitably affect the normal aircraft and aircrew
scheduling with dramatic changes. Due to the damage or the loss of the accident aircraft, the
airline has to reschedule the timetable. For example, an aircraft may need to be taken out of
normal service as a relief aircraft; or the airline may need to lease another aircraft from
elsewhere as a replacement. Ground staff may also be asked to perform additional work to
handle trivial tasks during the critical period. These are classified into the uninsured costs of
the accident. As to the public, they need to be informed about the changed time-tables as soon
as possible in order to reschedule their own plans.
B. 1.2 The relationship with the media
Few news eventshave such a powerful draw as aviation disasters.It is the dramatic nature of
aircraft accidentsthat attracts the media and grabsthe headlines.Data reveals that from 1978
to 1994, the New York Times disproportionately reported fatal events involving jet aircraft
and fatal eventsin the U. S. or involving US carriers (Curtis,1997).
Once an accidenthas occurred, the first task for the newsroom is to despatchreportersto the
sceneso that they can collect all possible information and pictures describing the event and its
immediate aftermath.Consequently,inaccurateinformation may be reported to the public due
to lack of verification.
As such, more and more airlines are recognising that dealing with the media is also a major
part of crisis managementbecausethe way in which airlines respond to media enquiries will
determinethe tone in which an accidentis reported.If the media is not supplied with constant
factual information then it is possible that conjecturesand rumours will make the front page
of many newspapersfor days or weeks. Bailey (1999) describedan example of a TWA 800
explosion: the company's initial public responseantagonisedthe media and maladroit public
comments fed the growing media portrayal of an airline which was uncaring and
unresponsive to the feelings of the victims' families. The public outcry led to the introduction
of the Airline Disaster Family Assistance Act, later extended to include foreign carriers. This
has imposed a significant financial and logistical burden on airlines operating to, from and
within the US.
Hence, it is of vital importance to deal with the media properly, which not only helps the
to communicate positively with the public, but creates a win-win scenario. The power
airlines
of media can never be underestimated.
280
B. 1.3 Reputation, public image and load factor
While the accidentairline is likely to suffer the most, rebuilding confidencequickly is crucial.
Becausethe load factor is always in direct proportion to its public image and reputation.For
example, following the A300 crash at Nagoya in April 1994, China Airlines experienceda 20
percent decline in the number of passengersduring May and did not notice an increaseuntil
December.
Another example is that of British Midland. In the aftermath of the Boeing 737 accident in
1989 near the UK's East Midlands airport, the chairman of British Midland, Sir Michael
Bishop, went to the scenequickly and spoke to the media as well as expressingconcern and
sympathy for victims' families. According to the observation of Guild (1995), British Midland
suffered no subsequent loss of traffic on the route from London Heathrow airport to Belfast.
Five years later, British Midland claimed to be the market leader on that route.
Dealing with its clients, the passengers,and the crew, as well as their relatives, friends and the
media is never easy. Besidesthat, an airline must continue to be fully operational for months
and years to come and ensurethat, despite any accidents.This, of course,makesthe recovery
from accidentharder.
A Boeing spokesmanonce said: "Ours is a businessthat is based almost entirely on public

confidence." Indeed, in today's competitive market, a carrier perceived as being not as safe as
others is less favourable to the travelling public. As the airline businessis highly dependenton
public confidence, so the question is: How to recover the airline's public image should an
accidentoccur?There are various answersto this question.
Take US Air for example. Following an accident in 1994, and in order to rebuild the public's
image, a full page advertisementrevealed the fact that the company planned to hire outside
organisationsto scrutinise safety and in addition appoint a new vice president for safety.This
resulted in controversial discussions.Some aviation consultantsthought the more quickly the
airline returnedto normal operation the better, without the need to draw public attention to the
event.Nevertheless,somepeople arguedthat USAir did the right thing to inform the public of
their plans and were searchingfor a solution to remedy the situation.
There can be no right or wrong regarding the strategiesof coping becausedifferent reactions
and strategies vary with different types of airline and their nationality. One implication worth
noting is that few managershave had the opportunity of acquiring the necessaryknowledge to
deal with such a crisis since an aircraft accidentis a rare occurrence.Therefore, Taylor (1997)
suggeststhat it is not sufficient to know that there is a chapter on accidentssomewherein a
manual. There is a clear need for intensive training, and without this a crisis could very easily
becomea catastrophe.
B.2 The Impact on Financial Performance- Costs of Accident
It seemsfairly simple that accident costs are the sum of all reportable damage, injury, and
illness costs.Peopletend to just look at the costs of accident in terms of monetary costs which
will affect airlines directly. In fact, accident costs vary largely from airline to airline and
country to country. Their sumsare not entirely determinedby economic consideration.
Costs of aircraft accident are classified into two categories:Insured Cost and Uninsured Cost
(or Hidden cost). The former are the actual claims paid by insurance companies,commonly
compensationpayments and medical expenses.The latter are those costs not related to the
compensation payments directly and hidden in some departments of the company such as
damagedreputations, lower productivity, etc. which can be neither quantified nor estimated.
Therefore, the monetary value is not always the most critical factor.
281... ý_ý
From the above, there are two fundamental points that should be noted in relation to the cost
of accidents: firstly, there are economic consequences of aviation safety; secondly, the costs
and benefits of safety cannot be measured only in economic terms.
Although the economic considerationof accidentcostsis not the first priority for every airline,
accident costs could have an influence on the airlines' financial performance.These costs can
individually and collectively drain the company's financial reserves.Those which are tangible
and able to be measured by monetary value, such as loss of business, the expenditure of a
crisis management centre, costs of accident investigation, etc. will affect the airline's cash
flow and profitability directly.
Meanwhile, those costs that stem from the intangible uninsured costs also have a significant
impact. A damagedreputation, lower brand loyalty, decreasingcompany morale, and so on,
would shakethe confidence of stockholdersand employees,and lower the company's stock
price. These intangible costs can even acquire greater importance than direct financial effects
measuredby accountingmethods.In somecasesthey have causedthe collapseof a company.
Therefore, given insured costsare coveredby proceedsfrom insurancefor the crashedaircraft,

the accidentcompanycannot expectto recover all the uninsuredcosts as the premiums are too
high to be afforded, particularly when safety costs, like the safety improvement and safety
prevention cost (investment),are included.
282
APPENDIX C Quality Management versus Safety Management
C.1 The Concept of Quality
One of the most influential individuals in the quality revolution has been Dr. Edwards
Deming. It was in 1950 when Japanhad a weak economy and a reputation for manufacturing
cheapand low-quality goods, that a group of visionary scientists,engineers,and businessmen
brought to Japan an American managementconsultant, Dr. Deming. He not only taught the
Japaneseindustry how to use a tool called statistical process control to achieve continuous
improvement in quality but he also brought a philosophy for the total managementof a
company.
Thirty years later Japanhad become one of the world's greatestindustrial powers, having has
achieved a reputation for quality that was unsurpassed.Witnessedby businessand industries
in other countries, in 1980s"Quality" grew in popularity for companiesto focus on. Quality
awards, such as the Malcolm Baldrige National Quality Award (MBNQA) in the US, were
consequentlybeing establishedto provide quality criteria in severalindustrialised countries.
Particularly over the last 10 years, industry's commitment to quality has significantly affected
the activities and tasks performed to create products and provide services (Manzella, 1997).
Almost any production company has a quality control or quality assurancedepartment,which
has allowed companiesto increasequality and productivity with less supervision.
In the early stages of development, the concepts of quality and excellence have focused
primarily on external customers - providing "quality" products, improving customer
satisfaction and building better customer relationships. Gradually, the managementsystem
exists not to ensurethe management'sorders are executedbut to help employeesand remove
barriers that prevent them from doing the job (Smith, 1996).
Understanding and recognising quality in civil aviation is important. There are several reasons.
Firstly, quality influences travel demand and market share from a customer's point of view.
Secondly, the performance of carriers is of interest to the regulators, carriers themselves and
the public. Knowing the information and position can help to enhance quality of carriers.
Lastly, the outcomes of a specific aspect of quality, such as air safety, are engaging people's
curiosity,
C.2 Linking Quality with Safety
A remarkable evolution in quality assurance specifications has taken place during the last 35
years (Hughes, 2000). This evolution has taken the form of various quality-improvement
initiatives, which have produced several forms of specifications, each an improvement on the
previous one:
Statistically orientatedend-result specifications (1960s);

Statistical quality assurance(1970s);
Total quality management(1980s)
In other words, the evolution of quality management shows three stages: quality control,
quality assuranceand total quality. Similar stages can be found in SMS, including safety
control, safety assuranceand total safety (Herrero et al., 2002). The objective of quality is to
improve the quality of the product through the detection and elimination of defects. It is
similar to the objective of safety which is the reduction of injuries through the elimination of
unsafeactsand work conditions.
One of the first contribution, relating to the integration of quality with safety was by Dumas
283
Appendix C Quality Management versus Safety Management
(Herrero et al., 2002). In 1987, after carrying out a study for more than 5 years, Dumas
discovered that quality programmes and safety programmes have the same components, i. e.
successful safety programmes and successful quality programmes are based on the same solid
foundations (Dumas, 1987).
Manzella (1997) affirms that in order to obtain excellent safety results, one needs to integrate
the safety system into the quality management system. Table C-1 illustrates the similar
elements of quality and safety. He comments that quality and safety principles are essentially
the same. As Crosby (1989) states, "Safety is a great analog for understanding quality.
Everything in safety is about relating to the absolute of quality management. "
Table C-1 The Principles and Relationship of Quality and Safety
SAFETY vs. QUALITY
Goal: Zero accidents Goal: Zero defects
Incidentanalysis Event analysis
Written policies, procedures Documented policies, procedures

and guidelines and work instructions
Safety committees Quality circles, employee
involvement team
Employee participation Empowerment
Statistical analysis Control charts, statistical process

control
All accidents are preventable All non-conformances are preventable
Source:Manzella, 1997
C.3 From TOM to Safety
Following the evolution of quality management, TQM has been a popular intervention all
around the world, especially in industrialised countries (Garvin, 1991; Evans and Lindsay,
1995; Samson and Terziovski, 1999). It is a management framework set up to deliver
self-regulatory compliance.
Since 1992, the UK's occupational health and safety has been viewed as an integral part of
TOM, which is based upon two respects: the model, HS (G) 65, and the norm, BS (British
Standards) 7850 (Deacon, 1994) (see Figure C-1). HS (G) 65 is used for safety and health
management, while BS 7850, the quality regulation added to the traditional concept of TQM
that satisfaction of the client, the safety, the health, the environment and the managerial
objectives are checks to each other.
284
Appendix C Quality ManagementversusSafetyManagement
Figure C-1 Model of Integrated Quality Safety
TQM
HS(G)65 BS7850
Health and safety Quality
BS: British Standards are developed and maintained by BSI British Standards, which is the
UK's National Standards Body.
HS (G) 65: Published by HSE (Health & Safety Executive, British agency) as practical guide for
directors, managers,health and safety professionalsand employee
Source:Deacon, 1994
Although there are always going to be debatesabout how to categoriseelements of TQM

(Samsonand Terziovski, 1999) or there is less agreementas to what the key elementsof TQM
are and what the critical factors are that influence the TQM implement process (Porter and
Parker, 1993),there is a widespreadconsensusthat TQM is a way of managing organisations
to improve their overall effectiveness(Porter and Parker, 1993) and TQM has demonstrated
that it is an effective way of maximising the long-term competitivenessof a company and can
improve the effectivenessof the programmesof safety and health (Goetsch, 1999).
To this end, JAR - OPS 1 (aeroplanes) and 3 (helicopters) demand the delivery of safety and
airworthiness. Based on the ICAO recommended practice (Annex 6 part 1), JAR- OPS states
that an operator shall establish an accident prevention and flight safety programme, which
may be integrated with the quality system, including programmes to achieve and maintain risk
awareness by all persons involved in operations. They instruct the operator to design and run
a "quality system" with its "quality assurance programme" to demonstrate regulatory
compliance. In addition, the ISO 9000 international standards can also help to implement a
quality system. It offers some useful advice that procedures should be documented only where
a lack of documentation may detract from quality.
With the well-defined managementstructure, TQM is actually no more than a formalised

method of communication to ensurethe right measuresare taken at the right time to satisfy
JAR-OPS 1&3 requirements and the company's intentions for compliance, as well as the
various constructrequirementsfor Health & Safety,the Environment and so on.
The famous PDSA (Plan, Do, Study and Act) quality cycle, also called Deming Wheel (see
Figure C-2), provides the tools needed to accomplish continuous improvement in quality,
productivity and safety.
285,
Figure C-2 Using PDSA for Safety
Incorporate improvement Find safety problem

in the system Define it - common or special
Look for anothersafety team work on solutions
Operational definition of safe
problem
design an improvement
Evaluate fording Implement the design

Does safety improve? on a small scale
Do it again - improve
plan Source:Adapted from Smith, 1995
Scherkenbach (1991) provides a useful outline of how to operationally define PDSA in eight
steps:
I. Plan: Develop a plan to improve

Step 1: Identify the opportunity for improvement
Step 2: Document the present process
Step 3: Createa vision of the improved process
Step4: Define the scopeof the improvement effort
II. Do: Execute the plan

Step5: Over a period, pilot the proposedchangeson a small scalewith customers
III Study: Study the result

Step6: Observewhat you learnedabout the processimprovement
IV. Act: Adjust the process,based on new knowledge

Step 7: Operationalisethe new mix of resources
Step 8: Repeatthe cycle
These steps are, in'fact, the processesof TQM. They describeshow this cycle works in a
particular company, bringing into its scope all existing documentation and management
practices and making improvements.Most important of all, the implication from this cycle is
what the tools of the Quality Assurance Programme are and how to use them to achieve
(measure)quality.
Salazar (1989) points out two tools exist that can be used to measurethe quality of a safety
programme:
1. Safety inspectionsthat identify the practices,behaviours,and unsafeconditions; and
2. Safety audits that identify the actions carried out by top managementof the company
that affect positively the systemof safety.
286
C.4 Total Safety Management (TSM) versus TQM
TSM is safety managementwritten and practiced using the principle of TQM (Herrero et al.,
2002). The similarities and differences between TQM and TSM are listed in Table C-2.
According to Goetsch(1999), TQM makes everybody involved in the progressof quality, and
TQM also makes the Director of Quality act as both coordinator and assistant.TSM makes
sure everybody is involved in the topics of safety, and the functions of the Safety Directors
would be those of coordinating the processesand facilitating the necessaryresources.
Table C-2 Differences and Similarities between TQM and TSM
TQM TSM
Know the processes Know the sources of risks
Minimise the errors Minimise the risks
Center on prevention Center on prevention
Reduce variations Reduce the uncertainty
Deepen in the satisfaction of the client Deepen in the safety of the workers
the organisation and the clients
The problems are caused more by The accidents and injuries are caused
the system than by the individuals more by faulty planning than by the people
Source:Saunders,1995
Manzanedo (1994) advancsthe concept of TSM, indicating that many concepts,principles,

rules, technical and objectives of the TQM could move the TSM with a single change of
quality for safety (quoted from Herrero et al, 2002). This is reflected in the organisational
structure (see Figure C-3) suggested by FAA (GAIN, 2000), where the Flight Safety Officer
has a similar position to the Quality Manager.When the managementfunctions of safety and
quality are the same,these two positions can be combined in one, as some airlines do. Also
CAP 712 (UKCAA, 2001) indicates that in most small and medium sized companies it is
expected that the flight safety and quality tasks will have many common points and there
could, therefore,be no objection to the combination of the roles in one staff member.
287 ...
Figure C-3 An Example of Organisational Structure
Chief Executive Officer
Flight Safety Officer I11 QualityManager
Operations II Maintenance II Others
Note:
Safety & Quality functions may be combined under the same managementfunction.
Formal Reporting
Formal Communication
Source:GAIN, 2000
288
APPENDIX D Risk Analysis Techniques
There are various techniques to perform risk analysis. The following are some tools
frequently usedin the airline industry:
1. Probabilistic Risk Assessment(PRA)

PRA quantifies the probabilities and consequencesassociated with accidents and
malfunctions by applying probability and statistical techniques as well as various
consequenceevaluation methods.
2. Hazard Mode and Effect Analysis (HMEA)

This in-depth risk analysis has also been called Reliability Analysis for Preliminary
Design (RAPD), Failure Mode and Effect Analysis (FMEA), Failure Mode, Effect, and
Criticality Analysis (FMECA), and Fault Hazard Analysis (FHA). Basically it is a tabular
analysis used to analyse the effects of system and sub-system failures on a system's
operation.This bottom up analysisis a simple method and useful for complex systems.
3. Fault Tree Analysis (FTA)
FTA is a graphical method commonly used in engineering and systems safety engineering.
It is used to assess a system by identifying an end event and examining the range of
potential contributory events. FTA documents qualitatively the potential causal chains
leading to the head event, and accommodate quantitative analysis of the probability of the
head event.
4. Event Tree Analysis

The purpose of Event Tree Analysis is to organise, characterise,and quantify potential
accidentsin a methodical manner by modelling the sequenceof events that result from a
single initiating event. This analysis is used when time sequenceis important and is a
useful tool for analysing emergencyresponseto systemfailures.
5. Flight Operation Risk Assessment(FORAS)

FORAS is a tool that will assessthe accident/incident risk associated with a flight
operation. FORAS is designed to give safety managers and other users a quantitative
assessmentof specific risk for an operation,broken down into a variety of subgroups,by
fleet, region, route or individual flight. This assessment is performed using a
mathematicalmodel which synthesisesa variety of inputs, including information on crew,
weather, managementpolicy and procedures,airports, traffic flow, aircraft and dispatch
operations.The system will identify those elements that contribute most significantly to
the calculatedrisk, and will be able, in somecases,to suggestpossible interventions.Two
risk categorieswere identified in the first stageof FORAS: Controlled Flight into Terrain
(CFTT)and in-flight injuries due to atmosphereturbulence.
6. Risk Analysis Matrix (RAM)

Using RAM, it is possible to standardise the qualitative risk assessments and categorise
the hazard using the criteria the airlines consider important. Although there might be
different formats of RAM, the essence is the same. In RAM, risk is calculated as a
function of both the likelihood of occurrence and the severity of the likely outcome. The
Risk Matrix shown in Figure D-1 is based upon the one regularly used by British Airways
Safety Services in their monthly safety bulletin `Flywise'.
289
Appendix D Risk Analysis Techniques
Five categoriesof risk are identified:
A: Severe- an incident requiring the highest priority for resourcesand action.

B: High- incidents of significant concernwhich take priority over most other
incidents.
C: Medium- incidents requiring the attention and action of a line department.
D: Low- an incident of low concernwhich normally requiresno further action.
E: Minimal- incidents that are of statistical interest only.
Figure D-1 The Risk Matrix
Likelihood of occurrence
LOW MEDIUM HIGH
HIGH C B A
oý MEDIUM D C Bý-
l
LOW E D C
Source:Flywise
7. Risk specific safety index products - performance indicators

A safety rating systemthat assessesthe declining or substandardoperationalperformance
can provide airlines with risk indicators and an early warning signal thus allowing an
early remedial action. For example, Schwartz (1998) suggestsa Risk Index Performance
Indicator for FSF.The risk indexing products include:
1. Primary product
"a set of risk-specific safety indices
2. Secondaryproducts
" associatedtrend measure
" global (organisationaltrend indicator)
And the potential usesof risk index are:

" Performanceanalysis
" Benchmarking
" Relational database
" Performancepredictor
" Safety management
290
APPENDIX E Documentation Pyramid
To achieve the efficiency and safety of an organisation, the operation concepts and
documentation it is very important to maintain consistency and logic. An airline's safety
documentation, can be viewed as a hierarchy containing three tiers: policies, procedures and
records (Ho, 1996), as shown in Figure E-1.
Figure E-1 The Documentation Pyramid
PI ES f- Define goal, mission

and what should be done
Manuals OCED S
f- Define who, when, where
and how to do
RECORDS Evidence for improvement
Source:Adapted from Ho, 1996
Policies
The policy statementis a document stating the company's missions and goals, and defines
what will be done and what should be done. It should be clearly written and easy to
understand.
Procedures
With the statedpolicies, proceduresare made by the usersto operatethe equipment efficiently,
as such proceduresare not inherent in the equipment. Degani and Wiener (1994) arguedthat
procedures are a form of quality assuranceby managementand regulating agenciesover the
operators;they exist to specify unambiguously six things:
1. What the task is

2. When the task is conducted(time and sequence)
3. By whom it is conducted
4. How the task is done (activities)
5. The sequenceof actions
6. What type of feedbackis given (callout, actions etc.)
Since procedures are working instructions, it is noted that they should be designed to be
consistent with the policies. For example, Standard Operating Procedures (SOPs) are a set of
procedures that provide operators with step by step guidance for their task. Standardisation
ensures the best method of operation and makes sure employees behave in a consistent and
predictable way. Following SOPS can maintain the services quality and can achieve a certain
level of safety. It is a part of the assurance system.
Records
Records are the evidencesfor improvement, also serving as one means to check if policies
and procedureshave been followed. Ideally, all departmentsshould review their practices and
procedures periodically to ensure compliance. Although records also show problems
un-reviewed, they do provide valuable and traceable information and data for management,
who can check the need for initiating corrective action.
Manuals
Manuals are the documentsrecording SOPs.They specify the priorities and goals of work and
proceduresfor different departments.
Manuals should be revised periodically in order to meet
the company'sneedsfor safety improvement.
291
APPENDIX F Other MEDA-like Approaches
The Aircraft and Maintenance Safety (ADAMS)

Dispatch consortium has also produced a
paper-based tool, similar to the MEDA form but much expanded (UKCAA, 2000). Other tools
(shown in Table F-1) are computerised versions of the MEDA form, some with data analysis
capabilities.
Table F-1 Other MEDA-like Approaches
Tool Description
TEAM (Tool for Error Analysis in Developed by Galaxy Scientific'', customised
Maintenance) for each airline.
AMMS (Aurora Mishap Management Developed by ex-MEDA and ex-US Air Force
Sv teni) personnel. It is a commercially available

system, designed for use in the transport
industries. PC based but adds the costing
element.
BASIS MEI (BASIS maintenance Error Developed by BA.
Investigation)
MEDA/SEDA This software package is the BF Goodrich
adaptation of the Boeing MEDA software.

U KCAA MEMS FMS (Maintenance Error A customised generic version of the BF
Management System Free MEDA Software) Goodrich software.
Source: UK CAA, 2000
Galaxy Scientific, a software company in the USA were the software supplier to Boeing during the
development of MEDA so TEAM naturally utilises the MEDA tool but provides a personalised
front end for the user (Chapman, 2000).
292
APPENDIX G Cases of Organisational Change - Culture
Transformation
The establishmentof major culture changes appearsto be rare but is feasible (Ho, 1996).
Managementcan direct cultural shift by articulating the desired values, and reinforcing the
proper norms; however, managementmust be sincerein their efforts in this direction. Two of
the most well known examplesare ScandinavianAirlines System (SAS) and British Airways
(BA). Not only were the culture of their companieschanged,but also their fortunes.
G.1 The Case of SAS
In the early 1980s, SAS went through a spectacularturnaround process.The new president,
Jan Carlzon, had discoveredthat the reputation of SAS rested upon the millions of "moments
of truth", i. e. verbal encountersbetween airline staff and passengers,instead of the products
provided, the safety of the aeroplanes,the convenience of schedules and so on. Yet the
"moments of truth' usually last at most less than thirty seconds.So he led the company to
switch from a product-and-technology orientation to a market-and-serviceorientation in the
four years from 1980 to 1983. The background and circumstancesof SAS before and after
this cultural changeare listed in Table G-1.
Table G-1 The Cultural Change Case: SAS
Company feature SAS

Size Medium
Length of major cultural From 1980-1983
change effort
Loss reported before Small loss reported in 1980
change effort (after) (2 percent of revenues)
Leader background Unconventionalinsider
Career path Grew up in SAS, but not
in the core business
Source : Kotter and Heskett, 1992
G.2 The Case of British Airways
Similarly, BA's cultural changewas under the new leadershipof Colin Marshall in 1983. After
witnessing the successof SAS, the new CEO began to put the culture transformation into
practice. Hampden-Turner(1990) notes the following actions and considerationsthat occurred
in BA:
1. Sending out clear signals. Five cultural signals were sent by Marshall:
" Making decisions: making mistakes is more forgivable than not

deciding in the first place.
" Facethe customers:need to start to face outward, toward customers.
" Names,not titles: no one should hide behind titles or job descriptions.
" Cross-functional groups: large committees are replaced by small groups in parallel,
charging them with responsibilities.
" Less hierarchy: reduce the number of levels of hierarchy.
_..., ._.. _. 293..

Appendix G Cases of Organisational Change- culture transformation
2. Cultural research
A major research into customer attitudes towards BA was conducted. The findings
showed that BA was "cold, aloof, uncaring and bureaucratic" to its customers. This was
due to military and technology in its cultural orientation until the early 1980s. In fact, BA
also found that culture is really what customers buy. Production innovation can be rapidly
imitated and copied from other airlines, but culture can not be copied. It has to be learned.
3. Changing the culture through training programmes

The human resources department started to organise and initiate several training
programmes,which pushedhuman dynamics into more positive patterns.Meanwhile, the
empowerment message from Marshall was clearly given by showing that grassroots staff
were the only people who could help the organisation.
4. Extending change throughout the organisation

It was not enough to galvanise the ground staff and the cabin crews alone. Cultural
changesin one section of an organisationmust pervadethe whole organisationin order to
last and be successful.Managerswere trained to createa vision for the whole organisation
so that they could identify with their top managementgroup, followed by a vision for the
group they managed.Subordinateshad to be shown how their job contributed to the larger
objective.
5. Management theory can be applied to real management
BA's version of the sevenSs diagram is shown in Figure G-2. The sevenSs, mobilised by
an overall Strategy, include hiring the right Staff, training them with the right Skills,
managing them in the Style required, Sharing values with them, installing the right
Systemsand improving the Structure.By using such a model, culture and strategycan be
united as a whole.
Figure G-2 BA's Version of SevenSs Diagram
Structure
Staff II Systems
Strategy
Skills ýýý Shared
Style
Source:Hampden-Turner,1990
294
Appendix G Casesof Organisational Change- culture transformation
BA had, eventually, turned itself round and made a great improvement (see Table G-3). The
successful story of BA has now become a touchstone of the cultural change issue. BA
revealed the essence of its success,
which is "what we have got to build are the kind of groups
that nurture individuality and the kind of individuals that can sustain and develop groups."
That clearly showedthe organisationalculture BA would like to build (for all groups) and the
importanceof subcultures(as groups consisting of individuals).
Table G3 The Cultural Change Case: BA
Company feature BA
Size Large
Length of major cultural From 1982-1988

change effort
Loss reported before Significant loss reported in 1981

change effort (after) (7 percent of revenues)
Leader background Outsider
Career path Become CEO at BA in 1983.

Came from Sears Holdings
Source : Kotter and Heskett, 1992
The other key to cultural and organisational changesis to maintain the changes,which was
also achieved by BA, although it did face some dilemmas. While cultural transformation is
taking place, in order to sustain competitive advantage, maintaining changes, providing
feedback and adjusting the changesare necessary.This concept and its associatedprocesses
are now commonly termed the learning organisation. Senge (1994) notes that "Learning
organisations are the places where people continually expand their capacity to create the
results they truly desire, where new and expansivepatterns of thinking are nurtured, where
collective aspiration is set free and where people are continually learning how to learn
together." As such, the best consequenceof cultural changes can result in organisational
learning, and a learning organisation can cultivate a better culture. Organisational change is
then achieved.
295-
Appendix G Cases of Organisational Change- culture transformation
This pagehas beenleft blank intentionally
296
APPENDIX H Four Proactive Safety Programmes
H. 1 Managing Engineering Safety Health (MESH)
1. Origin and objectives

Managing L',nginecring Salety Health (MESH) is a programme created for British Airways
Engineering Company in 1992 by a team lead by Professor James Reason from the University
of Manchester. It is a set of diagnostic instruments for making visible, within a particular
engineering location, the situational and organisational factors most likely to contribute to
human factors problems (Reason, 1994; Maurino et al, 1995).
Designed to assess the safety climate of an organisation, the measures of MESH give an
indication of the system's state of safety (and quality), both at the local workplace level and in
general. It is a system of measuring a number of local and organisational factors and the
interplay between them.
2. Philosophy
According to Maurino et al (1995), the underlying philosophy of MESH is:
High standards of safety, quality and productivity are all dependent on organisational
'health'.
`Health' is assessed and controlled through regular measurements of `vital
signs' at both local and system levels.
M1; S11 identities those 2-3 factors most in need of correction and measures remedial
efforts so that a system's state of health can be assessedand controlled.
MESH is designed to provide the measurement necessary to sustain long-term system
fitness.
The programme identifies three basic groups into which accident-producing factors fall
(Maurino et al, 1995; CASA, 1998). These are shown in Figure H-l:
human fallibility (at the organisational and workplace levels)

Technical and procedural shortcomings
Local hazards.
Figure 11-1 Accident-producing Factors in the Workplace
Human Technical & Local

Fallibility Procedural Hazard
Shortcomings
Note: Grey area is for demonstration Source: Maurino et al, 1995
The content of each of these buckets (the grey area in Figure H-1) will never be empty
completely although they can change from time to time. Imagine that each bucket gives off
particles. The fuller the bucket, the more it gives off. MESH is designed to give up-to-date
indication of the fullness of the buckets, if we assume accidents and incidents arise when
these particles combine by chance in the presence of some weak or absent defence. It does
this by sampling selected ingredients in each bucket (Maurino et al., 1995).
297
Appendix H Four Proactive Safety Programmes
3. Process of assessment
As such, a system's safety health can be assessed by first listing the "ingredient factors",
which are divided into organisational factors and local factors. Exactly what local factors are
assessed depends on the workplace. Different factors can be developed for different
workplaces. They are intended to give a short-term indication of the accident-producing
factors present within a particular workplace.
Table H-1 shows the 12 local factors, which were derived from a survey of the problems
encounteredby maintainers in a line `casualty' hangar. Ideally around 25 percent of the
workforce are required to rate each local factor for the extent to which it causesproblems in a
limited number of recentjobs. Ratings are madeon a weekly basis.
MESH also assessesthe impact of upstream organisational factors upon a particular

workplace, as shown in Table H-2. Those who are in chargeof the departmentor hangar,i. e.
line managers,rate how much each organisational factor has causedproblems in that site.
Theseratings are made quarterly.
Table H-1 Local Factors Measured in an Operational Hangar
1. Knowledge, skills and experience

2. Morale
3. Tools, equipment, parts
4. Qualit of support
Local 5. Fatigue
factors 6. Pressure
7. Time of day/night
8. Environment
9. Computers
10. Paperwork, manuals, procedures
11. Inconvenience
12. Personnel safety features
Source:Adapted from Maurin et at., 1995;Reason1995a
Table H-2 Organisational Factors Measured in Each Workplace
1. Organisational structure
2. People management
3. Provision and quality of tools and equipment
Organisational 4. Training and selection
factors 5. Commercial and operational pressures
6. Planning and scheduling
7. Maintenance of buildings and equipment
8. Communication
Source:Adapted from Maurino et al., 1995;Reason1995a
For convenience, Reason (1995a) has pointed out that assessmentsare made directly on
computers by using the mouse or keyboard. The assessorsare randomly selected and are
anonymous in order to keep MESH as a sampling tool. When logging on to the MESH
programme,assessorsare askedto give their grade,trade and location. MESH employs direct
a 5-point rating of the dimensions with regard to specific locations and tasks. For example,
the questions are phrased as follows: "To what extent has this factor (either local or
organisationalfactor) been a problem in carrying out thesejobs (a previously specified list of
3-5 recently performed tasks)?"
The results of both local and organisational measurementsare converted into bar diagram
profiles, graphically on an x and y-axis, with the x-axis being the factors and the y-axis being
the extent of the problem. Figure H-2 showsa schematiclocal factor profile.
298
Figure H-2 A Local Factor Profile
When completing their rating, assessors are provided with a profiled summary of their own
input together with a cumulated profile for all ratings made over the past four weeks (Maurino
et al., 1995). The whole diagnostic package is implemented within a linked suite of computer
programmes into which all the gathered information and sample results are fed and then
tracked. Their purpose is to identify the two or three factors most in need of remediation. As
mentioned previously, MESH is a sampling tool. It samples jobs and tasks to identify those
factors most likely to come together to cause future incidents. It also tracks the progress of
subsequent remedial actions.
In MESH, the local factor assessments are made at weekly intervals by a randomly selected
proportion of the workplace in each of a variety of workplaces (i. e. operational hangars,
majors overhaul hangars, workshops, etc. ). The organisational factors are assessed at
three-monthly intervals by technical management in each location because these are the
people best placed to judge the impact of "upstream" organisational factors upon the
reliability of their various workplaces (Reason, 1995a).
4. Advantages
Reason (1995a) also suggests that by identifying factors in need of improvement and tracking
the changes over time, MESH enables the maintenance of adequate safety health, comparable
to a long-term fitness programme, in which the focus of remedial efforts switches from
dimension to dimension as previously salient factors improve and new ones come into
prominence.
In terms of training, MI: SII is directly used by front-line personnel for reporting their own
points of view. The MESH end user is able to use the system without needing any further
training because it is characterised by a single phase and tool. The language within MESII is
simple and the items it addresses are so broad and generic that they could be applied in any
other environment.
The advantages of MESI1 can be thus summarised as follows:
Staff are all involved in safety.

Managers can prioritise remedial actions and check upon their impact.
Direct safety resources where they are most needed.
I :ncourage better communications between management and staff.
299
5. Limitations
Although the first impression of MESH appears to be that it is easy, the implementation of
this tool may lead to some difficulties. Caution must be expressed when interpreting the
implied meaning of results when employing a single rating scale method of investigation. Due
to personal tendencies, some people tend to use only the extreme of the scale, while others
tend to use only the middle area.
In addition, when using a computerisedsystem,the "bigger picture" is not always so readily

available. As such, the MESH users have no accessto any overview of the global structure
and the items to be rated are displayedon separaterating pages.
Meanwhile, the factors rated in MESH have very generalmeaningsand can be interpretedand
rated according to different interpretations.Consequently,the results obtained by MESH are
not sufficiently detailed for suggestingproper corrective actions.
MESH has never been applied by low capacity operators,and would be beyond the resources
of smaller operators. Although it has also been implemented by Singapore Airlines
Engineering Company, this programme is not adopted widely by other airlines and has not
achievedthe significant improvementsin safety performancethat were originally expected.A
number of improvementsand modifications are currently being made as a result.
6. Other MESH-Like applications

MESH employs a systemic approach to safety management and is readily adapted to a range
of industries and disciplines. Prior versions of MESH-like instruments have been developed
for the oil and railway industries. Later versions are now being evaluated by the US nuclear
power generation industry (IFA, 1998). Meanwhile, Shell International uses a similar method
in its tanker and exploration operations in the form of Tripod-DELTA. Likewise, British Rail
currently employs a proactive instrument called REVIEW, which has also been used at West
Australian Railways. (In the case of REVIEW and Tripod-DELTA, only organisational factors
are assessed. The judgments are made regularly by supervisors in differing activities and
locations, via a computer programme. Tripod assessments are made quarterly; those for
REVIEW are made at approximately monthly intervals. )
H.2 Identifying Needed Defences In the Civil Aviation Transport

Environment (INDICATE)
1. Origin and objectives
INDICATE is a safetyprogrammethat hasbeendevelopedin consultationwith the Australian

regional airline industry for proactive purposes.The nameis basedon the underlying purpose
of the programme which is to identify and resolve deficient aviation safety defences (Edkins,
1998). It provides a formal communication channel for aircraft operators to regulations,
policies and standards to the Bureau of Air Safety Investigation (BAST) in Australia. It is also
known as BASI-INDICATE.
2. Underlying philosophy
The basic premise underlying the INDICATE programme is that generally people working
within the aviation industry will report safety hazards if given sufficient opportunity and the
right work environment. However, some individuals are reluctant to report safety hazards for
fear of blame or retribution, especially if the problem reflects negatively on company
management. Alternatively, safety hazards report may be reported but with little feedback
given to the reporter; and some smaller airlines do not have formally appointed operational
safety officers, to whom staff can access directly and can confidentially report safety hazards.
300
Appendix H Four Proactive SafetyProgrammes
Consequently potential safety problems remain undetected.
INDICATE was designed to minimise this type of communication problem by providing a

simple but structuredprocess to ensurethat consistentand high-quality safety information is
disseminatedto all company staff. This is achievedby first educating staff about the concept
of safety defences (BASI, 1998).
Edkins (1998) points out that safety defencesare barriers or safeguardsput in place to protect
a system from both human and technical failure. He presentsa modified version (Figure 2-22
in chapter 2) of Reason's model of organisational accident causation (see Figure 2-20 in
chapter2). 1
Edkins argues that each of the organisation, workplace and person/team components of
Reason'smodel are difficult to identify before an accident becauselatent failures are usually
unforeseeable,workplace factors are dynamic, and errors or violations are unpredictable.This
model implies that the integrity of safety defences can be more accurately determined as they
are more tangible and thus more measurable components within a system. Regularly
evaluating defences provides a tangible means by which latent organisational failures can be
identified. The INDICATE programme has therefore been designed to regularly evaluate
airline safety defences so that the potential risk of an accident can be minimised.
3. Basic elements of the INDICATE programme
The INDICATE programme involves establishing and maintaining the following six core
safety activities (CASA, 1998; Edkins 1998,1999; BASI, 1998):
i. Appointing an operational Safety Manager or Safety Officer who is

available to staff as a confidante for safety related issues
ii. Conducting a regular series of staff meetings to identify safety hazards within
the operation
iii. Establishing a confidential safety hazardreporting system
iv. Conducting regular safety meetingswith management
v. Maintaining a safety information database
vi. Ensuring safety information is regularly distributed to all staff
4. Methodology
Co-operation was agreed with Kendell Airlines, an Australian regional airline, to trail the
INDICATE programme.Commencing in July 1996, it took eight months to complete the trial
and determine whether or not the programme had had a positive influence on the airline's
safety performance.
Since Kendell operates out of two major regional centres, the INDICATE programme was
implemented in one regional centre as an intervention group (INDICATE base - 81 staff)
while the other served as a control group (non-INDICATE base - 72 staff). This enabled a
comparison at the end of the trial period to objectively evaluate any changes in safety
performance across the two bases. Table H-3 summarises the six core elements of the
INDICATE programme as well as showing differences in application acrossthe intervention
and control groups.
301
.
Table H-3 Differences in Application of the INDICATE Programme Across Both

Experimental Groups
Six core safe activities of the INDICATE programme

I. Appointing an operational Safety Manager or Safety Officer who is
available to staff as a confidante for safety Yes No
related issues
ii. Conducting a regular series of staff meetings to identify Yes No
safe hazards within the operation
iii. Establishing a confidential safety hazard reporting system Yes No
iv. Conductingregular safety meetingswith management Yes No
v. Maintaining a safety information database Yes No
vi. Ensuring safety information is regularly distributed to all staff Yes No
A: Intervention group (INDICATE base)

B: Control group (non-INDICATE base) Source:Adapted from Edkins, 1998
5. Evaluation criteria
There are many potential measuresof airline safety performance, including the absolute
number of fatal and non-fatal accidents; fatal and total accidents per million departures;
passengerfatalities per million passengersor per million miles, etc. Regardlessof which
measure is used, it is important that it is examined regularly if a safety management
programmeis to be effective in improving safety performance.Nevertheless,accidentsare so
rare in airline operations that they cannot be used as a statistically reliable index of safety
performance.
As such, the INDICATE programme was evaluated based upon the following five safety
performancecriteria (Edkins, 1998; BASI, 1998).The criteria were used in both intervention
and control groups to determine whether the programmewould achieve an improvement in
airline safetyperformanceover the eight month period. Thesecriteria were:
1. Airline safety culture

2. Airline staff risk perception of aviation safety hazards
3. Staff willingness to report safety hazards
4. Action taken on identified safetyhazards
5. Staff commentsabout safety management
These criteria were chosenbecausesafety culture and hazardsrisk perception have been well
researchedwithin various industries and show a strong relationship to workplace accidents.
However, they are essentially attitudinal measures which should not be relied upon in
isolation. In addition, the remainder of the criteria were included becausethey are more
tangible indicators of the programme's successand were consideredcomplementary to the
attitudinal criteria. All the criteria should provide a comprehensiveevaluation of the validity
of the INDICATE programme(BASI, 1998).
6. Brief description of each criterion
1. Airline Safety Culture

In order to measure changes in safety culture during the eight month trial, the Airline Safety
Culture Index (ASCI) was developed. A questionnaire consisting of 25 positively worded
statements was constructed and each statement requires a response on a five point Likert scale,
ranging from strongly agree to strongly disagree. Based on Brown and Holmes (1986),
Cooper (1995) and ICAO (1994), a number of items were developed regarding the following
core dimensions: management commitment (2 items), management action (6 items),
302
employee commitment (4 items), level of perceived risk (1 item), beliefs about accident
causation (2 items), emergency procedures (1 item), the provision of safety training (2 items),
and safety communication (7 items).
The questionnaire was administered on three occasions to both groups; prior to the
implementation of INDICATE, at the mid term of the trial, and at the end of the trial period.
By comparing the difference in both groups at different periods, it is expected that the
intervention group would demonstratea better safety culture score when compared with the
control group.
1. Staff Risk Perception of Aviation Safety Hazards

ICAO's Accident/ Incident Reporting System was used to compile a list of the most
frequently occurring safety hazardsin commuter/regionalairlines from 1979-1996.The list of
aviation safety hazardswas presentedto staff both in the intervention group and control group
at the beginning and the end of trial. It required respondentsto rate each of the hazards
according to their potential to affect the safety of airline passengers.Each hazard was rated in
accordance with its hazardousness and the likelihood of it occurring within the airline
environment. The questionnaire produced an individual risk perception score for both
hazardousnessand likelihood.
It was expectedthat there would be little difference in risk perception between groups at the
commencementof the trial. Nevertheless,a significant reduction in the mean hazardousness
and likelihood scoresof staff in the intervention group was expected.
2. Staff Willingness to Report Safety Hazards

During the eight month period, both intervention and control groups had chancesto accessa
confidential safety reporting systemto inform managementabout safety hazardsthat they felt
had potentially serious safety implication. Identical information on how to use the reporting
system was given to both groups. It was expected that the intervention group would have
submitted more reporting hazard report forms, comparedwith the control group.
3. -Action Taken on Identified Safety Hazards

It was also expected that the intervention group would have achieved more action on safety
hazards compared with the control group because the implementation of the INDICATE
programmewithin the airline provides a structured framework to addresssafety hazards.
5. Staff Comments About Safety Management

Staff were given the opportunity to complete a free response section in the safety culture
questionnaire before and at the conclusion of the trail. It was expected that staff in the
intervention group would have more positive comments about safety managementcompared
with the control group, at the end of trial. Figure H-5 contains the summary of the
methodological design in the eight month trial.
303
Figure H-5 The Methodological Design in the Eight Month Trail
July 96 July 96 July 96 March 97 April 97
INDICATE Measures: INDICATE Measures:

Group "Safety culture programme "Safety culture
-Risk perception "Risk perception
"Hazard reporting "Hazard reporting Evaluation
"Safety action [4 - "Safety action
Non- INDICATE -Staff comments No "Staff comments
Group programme
7. Implementing INDICATE
Within the company there is a Co-ordinator or Safety Officer who is responsiblefor running
the programme. Depending on the size of the operation, this function may be part of
someone'sexisting duties, or be the responsibility of a dedicatedposition.
Approximately every month a safety meeting was conducted with the managersfrom each
section (technical crew, cabin crew, maintenance crew, ground crew, operations and
union/associationgroups). At these meetings safety issuesraised by staff or management,at
any level, were discussed.
8. Results
The evaluation results reveal that there is a clear difference between the intervention and
control groups.According to Edkins (1998) and BASI (1998), staff in the Intervention group:
are more positive about safety management

are more willing to report safety hazards
have more confidencein management'sability to addresssafety issues
feel there is a better quality of safety communicationbetweenmanagementand staff
and betweendifferent operationaldepartments
9. Benefits
In Australia, it is a legal requirement to report air safety incidents via an Aviation Safety
Incident Report (ASIR). However, there is a recognised problem of under-reporting, which in
part stems from a lack of awareness about what should be reported despite this mandatory
requirement.
The programme provides a simple and structuredprocessto encouragestaff to report safety

hazardsand deficiencieswithin their work area.The safety information databaseand software
allows managementto addressall safety-relatedconcerns.Furthermore, senior management
regularly meet with safety staff to determine what to do about identified hazards.It is clear
that consistent communication of safety-relatedinformation within an organisation is crucial
for improving staff attitudestowards safety.
In addition, INDICATE provides a computerprogrammeto assistthe companyin detailing all

reported concerns.The software programme:
" records the nature of each safety hazard
" records any action (or lack of action) taken on eachhazard
304
" maintains the confidentiality of the reporter

" generatesa recommendationfor either:
-a departmentalmanager
- senior management
- appropriateaviation authorities
Where the information on the hazard is sent dependson the nature of the hazard and who is
best equippedto assistin rectifying an identified problem (CASA, 1998).
As such,a number of benefits of INDICATE are concluded:

1. Encouragesbetter communication between management and employees about
safety.
2. Reveals critical areas for development of procedures or training and for priority
inspections.
3. Provides a framework for feedback as to the efficacy of assumptionsmade about
hardwareor operationalprocedures.
4. Provided a baselinefor managementdecisionsregarding safety issues.
5. Provided a cost-effective safety managementtool.
Edkins (1998) explains that the results of the INDICATE trial suggest that measuring safety
culture provides a useful method for monitoring changes in company safety performance and
may assist in identifying elements of a safety management programme that require
improvement, such as a hazard reporting system.
Most importantly of all, the evaluation of the INDICATE programme illustrates that the
greatest source of variance is not necessarily aircraft equipment or the category of operation,
but the real cost from the safety culture of organisations within the aviation system. A small
to medium size airline, operating within a limited budget, does not have to spend large
amounts of money to improve its own safety culture. The benefits from implementing such
initiatives will ultimately help to improve operational safety and, in some cases, reduce
operating costs.
Currently there are over twenty passengercarrying operatorsof varying sizes both Australia
and overseaswhich have implemented the INDICATE programme,which is easily tailored to
the varying requirementsof different sizesof operation.
H. 3. Proactive Error Reduction System (PERS)
1. Origins
The PERS programme is funded by the Federal Aviation Administration's Office of Aviation
Medicine and devised by Dr. Colin Drury at the University at Buffalo (UB), the State
University of New York (IFA, 1998). The FAA-funded project began in 1989 following a
Congressionalhearing prompted by an incident in which 18 feet of roof pulled away from an
Aloha Airline jet as it was flying over the Pacific Ocean.A flight attendantwas sucked from
the plane and 61 passengerswere injured.
For years, Drury and his co-workers have analysederrors by airline workers in detail. They
are using this knowledge to build practical tools that allow to
users arrive quickly at solutions
to errors madeby airline workers.
- 305
PERS is an error managementsystem which can be used by non-experts in human factors

becauseit is basedon a human-factorsapproachto solving errors. The idea behind this is that
one should not just determine the immediate cause of the error, but examineall the things that
lead up to it. As such, this programme, as shown in Figure 2-23, is structured to use the
repeatingpatterns found in incident data in order to help airlines move from the recognition of
human factors as an issueto practical human factors solutions.
3. Advantages
There are four distinct functions within the programme:

1. A comprehensiveerror managementsystem which can combine many existing data
collection and analysis systems(eg. MESH, MEDA). In other words, it facilitates the
importing of data from other systems such as other audits and MESH, and links
databasesof maintenanceerrors with databasesof known solutions.
2. PERS facilitates the identification of error-pronesituationsby using
- error audit of specific tasks
- error assessment like MESH.
3. PERS supports error management strategies, identifying situations for
design/procedure changes.
4. PERS facilitates error reporting by employing
- error reporting modules like MEDA (an error-reporting system with interfaces to
MEDA)
- critical incident reporting modules.
PERS provides a way for airline personnelto analysean error or potential error, to discover
why it occurred, and then to see how they might go about changing systems,equipment or
work patterns to prevent future errors. PERS not only tells airline workers what to do if an
error occurs,but it also tells them what to do even if what has occurred are not actual errors,
but error-pronesituations.
4. Scenario
For example, supposethat an aircraft was hit by a ground vehicle driven by a mechanic. The
programmewill ask: Why was the mechanicthere?What was the mechanic doing? Why was
he in a hurry?
In one instance,it was found that drivers of ground vehicles often put the vehicles into neutral,
but did not turn them off when they got out of the vehicles for the simple reason that the
engines got so cold that they would be difficult to restart (Goldbaum, 1996). Therefore, by
providing such detailed information about an incident, the airline should be able to pinpoint
and addressthe chain of eventsthat lead up to the error.
306
HA Line Operation Safety Audit (LOSA)
1. Origins and objectives
LOSA was developedby the team of ProfessorR. Heimreich at the University of Texas,USA.
Under LOSA, flaws in human performance and prevalence of error are taken for granted and
the objective becomesimproving the context within which human perform. LOSA ultimately
aims to introduce a buffer zone or time delay between an error and the point at which its
consequencesbecomea threat to safety.The better the buffer or the longer the time delay, the
strongerthe toleranceof the operational context to the negative consequencesof human error
(Maurino, 2001).
LOSA are programmesthat use expert observersto collect data about crew behaviours and
situational factors on normal flights. Observationsgeneratea narrative of the flight classified
by phase, and these are conductedunder strict nojeopardy conditions, which meansthat no
crew are at risk for observedactions. Observers code observedthreats to safety and how they
are addressed,errors and their management,and specific behavioursthat have been associated
with accidentsand incidents.
Data from LOSA provides a picture of system operations that can guide organisational
strategy in safety, operations, and training. Helmreich (2001) points out that a particular
strength of LOSA is that the processidentifies examplesof superior performancethat can be
reinforced and used as models for training. Data collected in LOSA are proactive and can be
usedimmediately to prevent adverseevents.
3. Methodology of LOSA
The critical difference between a LOSA flight and a line check is LOSA's guarantee of
anonymity for the crew. Data are entered into a de-identified database and no crew actions are
reported to management or the regulatory agency.
In LOSA, error is classified as deviation from organisationalor crew exceptionsor intentions.

Errors committed by the flight crew are describedand coded along with actions taken to deal
with the consequences of the errors. Table 2-6 in chapter 2 lists the various errors and
remedial strategiesemployed in LOSA.
4. Advantages of LOSA
One of the important aspects of LOSA is the fact that it captures exemplary as well as
deficient performance, which provides airlines with the areasin which they excel as well as
those in needof improvement.
The other strength is that a database is being developed that allows organisations to compare
their results with other airlines. Such comparisons help in interpreting the significance of the
number of procedural and decision errors observed and the effectiveness of threat and error
counter-measures. The data allow management to prioritise safety initiatives and training
departments can use the information to develop targeted training.
307. _._
Meanwhile, the informative aspect of LOSA data is the ability to link threat recognition and
management with the specific behavioural markers that from the core of CRM. Using
error
LOSA, a model incorporating the Swiss Cheese model has been developed (Helmreich, 1999).
It recognises both overt and latent threats, and how they fit into the management of error and
undesired states.
Thesemarkers emergevery clearly in observerratings of the actions taken by effective crews.

Those who deal proactively with threat and error exhibit the following behaviours:
active captain leadership
-
- briefing known threats
- asking questions,speakingup
- decisions making and reviewed
- operationalplans clearly communicated
preparing/planning for threats
-
- distributing workload and tasks
- vigilance through monitoring and challenging
Although the challenge for the implementation of analysis of normal operations is to

overcome the obstacles presented by a blame-oriented industry, LOSA has been appreciated
by various airlines which are willing to make the investment in conducting the necessary
observations and analyses. In addition, the value of LOSA has also been recognised by ICAO
and has gained support to conduct a "LOSA Week" to promote this programme to interested
countries.
308
APPENDIX I Domino Theory
An accident is the undesired event (unwanted consequence) due to loss of control.

Consequently, it may result in damage to the elements of a system. A domino
sequence originally presented by Frank Bird, Jr. and now updated by Heinrich,
Petersen & Roos (1980) clearly states the accident sequence in a loss of control
system by explaining the five key loss control factors in order of dominoes (see
Figure I-1).
Within this system, control as a function of professional "Management" is optimised

through five established steps that systematically produce desired results. Origin, like
etiology, refers to the sources and its appropriated identification with the basic or
underlying cause serves to reinforce the desire to achieve a more effective control.
Figure 1-1 An Accident Sequence Domino
Management Origins Symptom Conta.Ql Loss

tia
4a pý ýn
n cn OD
4n ýý
c1 C
O y
O
CC
Source: Revised from Heinrich, Petersen & Roos, 1980
The immediate cause is usually a "Symptom" of a deeper underlying problem. Safety

managers and quality control managers frequently refer to these causes as unsafe
acts/conditions or substandard acts/conditions. Historically they are regarded as the
most important factors to attack.
An accident is described as an undesired event that results in death, injury, or property

damage. When considering the broader implications of accidents with "Contact", we
309
AppendixI DominoTheory
can see some more important relationships between the safety, environment, and
exposure factors, and the importance of interface between related human elements in a
coordinatedloss control system.
"Injury" has been frequently used to mean bodily damageor harm through traumatic
accident. Damage, as used in this injury factor, is intended to cover broadly all types
of tangible and intangible property damage. To optimise loss reduction, the safety
manager will also direct substantial attention to control countermeasuresat this last
factor in the sequence.It is also referred to as the post-contactstage.
310
APPENDIX J Interview Questions
+ Safety performance and management
1) What is the organisationalstructureto managesafety?

2) How safetyperformanceis monitored, measuredand audited?
3) Key indicators?Any statisticsare usedto indicate the changesin performance?
4) How to decidethe safety level?
5) What is the safety managementsystemin your company?
6) What are the componentsof a SMS?
+ Hazard discovery and reporting (risk management)
1) The use of risk assessmentmethods or tools?

By whom? How often? Who approves?
2) What stepsare taken in the company to identify and rectify hazards?
Is there a reporting systemin place?
What is the focus of the system?
Who doesthe reported information go to? (organisationallevel or national
level?)
Is the reported information visible usedwithin the organisation?
Are employeesencouragedto report hazards?How?
ºý Safety information
1) Who managesthe safety information?

2) Any safety publication within the company to reveal the safety information?
3) What is the relationship betweensafety information and discipline procedures?
Any feedbackto employees?
4) Safetymanual? Scope?
5) Any regular safety meeting with other companies?
6) Barriers?
311
Appendix J Interview Questions
+ Safety culture
1) Any cultural approach in safety management?
2) Focus on long term or short term?
3) Any conflicts between safety and production goals?
4) Safety goal? Included in companymission statement?
5) Leadership
6) Communication
+ Industry
1) What is the role of organisationwithin industry (e.g. regulator, operator,service
provider)?
2) What are the main hazardsfacedwithin the industry?
3) What is the degreeof risk?
+ Regulator
1) Do regulatorsprovide sufficient information?
2) What's the role of regulator?
312
APPENDIX K Airline Distribution List
ACES Colombia Air Tahiti Britannia Airways

Aegean Aviation Air Transat British Airway
Aer Lingus Air Vegas British European
Aero Lloyd Air Wisconcin British InternationalHelicopter
Aeroflot Air Zambique,
British Med. Airways
Aerolineas Argentinas Air-India British Regional Airlines
Aerolitoral Airline of South Australia British World
Aeromexico Airlink Pty Ltd Brymon Airways
Aeropelican Airours Internaional BWIAWest Indies Airways
Air 2000 AirTran Airways Canada 3000 Airlines
Air Afrique Alaska Airlines Canada 3000 Cargo
Air Algerie Alitalia Canadian Regional
Air Berlin All Nippon Airways Cape Air
Air Botnia Allegheny Cargolux
Air Caledonie Alliance Cathay Pacific
Air Canada Aloha Airlines CCAir
Air Canada Regional Alpine CCM Airlines
Air China America West Airlines Cebu Pacific Air
Air Dolomiti American Chautauqua Airlines
Air Europa American Eagle Airlines China Airlines
Air France American Trans Air China Eastern Airliens
Air Guinee Amtran China HainanAirlines
Air Hong Kong Arkia Israeli Airlines China Northern Airlines
Air Jamaica Asiana Airlines China Northwest Airlines
Air Kazakhstan Atlantic Coast Airlines China Southern Airlines
Air Liberte Atlantic Southeast Airlines China Xinjiang Airlines
Air Littoral Atlas CityFlyer Express
Air Macau Augsburg Airways CityJet
Air Madagascar Austral Lines Aereas Climber
Air Malawi AustralianAir Express Comair
Air Malta Austran Airlines Comair Limited
Air Mauritius Aviacsa CommutAir
Air MidWest Avianca Condor
Air Moldova AZZURRAir Continental
Air Namibia BankokAirways Continental Express
Air New Zealand Bering Continental Micronesia
Air Nippon Biman Bangladesh Airlines Copa Airlines
Air Nostrum/Iberia Reg Binter Canarias Corporate
Air One bmi british midland Corseair
Air Pacific Bouraq Croatia
313
Appendix KAirlineDistribution List
Air Philippines Braathens Crosair

Air Seychelles Brit Air Crossair
CSA CzechAirlines HawailianAirlines LuthansaCityLine

CyprusAirways HeavyLiftCargo Luxair
Delta Horizon Air Maersk Air
Delta Air Transport Iberia Airlines Maersk Air UK
Deutsche BA Icelandair Malaysia Airlines
DHLAirways Impluse Macair
Dniproavia Indian Airlines MalEv
Domodedovo IndonesiaAir Transport- IAT Mandala

Dragonair Iran Air Mandarian
Eagle Airways JAL Martinair
Eastern Australia JALways Meridiana
easyJet Japan Air System Merlin Express

Egyptair Jet Airways Mesa Airlines
EI Al JetBlue Airways Mesaba Airlines
Emirates JMC Airlines Mexicana
Era Kendell MIAT-MongolianAirlines
Ethiopian Airlines Kenmore Harbor Middle East Airlines
EuroLOT Kenya Airways Midwest Express
Eurowings KLM Monarch
EVAAir KLM cityhopper NationalAirlines
Evergreen InternationalAirlines KLM uk NationalJet
Express Airlines I Korean Air Negeria

Far East Air Transport KrasAir NepalAirways
Kuban Nippon Cargo
FedEx Express
Fine KuwaitAirways Nordeste LA Regionals
Finnair LAB Flying Service Northwest
First Air LAM Mozambique OlympicAirways
Flight West LanChile Oman Air
LAPA PacificAirlines
Flying Colours
Lauda Air Pakistan
Frontier Airlines
Fujairah Air Lauda Air Spa Pakistan International Airlines
Garuda Indonesia Lietuva Palmair Flightline

LineasAereas Allegro Pearl Aviation
Gemini
GhanaAirways Lithuanian PelangiAir
Go Lloyd Aereo Boliviano PGA- Portugalia
Grand Aire Inc LOT Polish Airlines Philippine Airlines
Great Lakes Love Air Polar

GujaratAirways LTU InternationalAirways Premiair
Gulf Air Lufthansa Pulkovo Airlines
Gulfstream InternationalAirliens LufthansaCargo India Qantas
314
Hapag-Lloyd LumbiniAirways QatarAirways

Regionair TAROM
RegionalCompagnieAerienne
Europeene Thai Airways International
Rio-Sul THY Turkish Airlines
Royal Air Maroc Trans States Airlines
Royal Airlines TransAsiaAirways
Royal BruneiAirlines TransaviaAirlines
Royal JordanianAirlines Transbrasil
Royal NepalAirlines Transaero
Royal Phnom Penh Airways Transeast
Ryan InternationalAirlines Transtate
Ryanair Tunisair
SAAirlink TurkmenistanAirlines
Sabena TWA
Sahara Tyrolean Airways
SAS UNI Airways Corps
SAS Commuter United
Saudi Arabian Airlines UPS
Shandon US Airways
Shanghai Airlines Vanguard Airlines
SIA Varig
Sibir Airlines VASP Brazilian Airlines
SilkAir Vietnam Airlines
Skyway Virgin Atlantic Airways
Skyways Express Virgin Blue Airlines
SkyWestAirlines Virgin Express
SouthAfrican Airways VLM Airlines NV
Southern Winds Volare Airlines Spa
Southwest WestJet Airlines
Spanair Wideroe's Flyveselskap
SpiritAirlines World Airways
SpiritAirlines Wuhan
SriLankan Airlines Xiamen Airlines
Sun CountryAirlines Yemenia

Sunshine Express YunnanAirlines
Sunstate
Swissair
Syrianair
TACA International Airlines
TAM Linhas Aereas
TAP Air Portugal
Taquan
315
This page hasbeen left black intentionally
316
APPENDIX L Covering Letter & Questionnaire
Air Transport
Building 115
Group
Cranfield UNIVERSITY
School of Engineering
Cranfield College of Aeronautics

Cranfield, Bedfordshire MK43 OAL
England
Fax +44 (0) 1234 752207
Direct Dial + 44 (0) 1234 754236/7
Dear Sir/Madam,
I am a PhD student, in the Department of Human Factors and Air Transport (School of
Engineering, Cranfield University, UK), conducting research into the airline safety
management. The purpose of my research is to evaluate current airline safety management,
investigate organizational factors affecting flight safety and justify the best practice for safety
improvementswithin the airline industry.
It would be greatly appreciatedif you could spend a few minutes completing the attached
questionnaire. This is
research not sponsored by any outside organization. The information
you provide will be used strictly for the purpose of this study and all replies will be treated as
highly confidential. A summaryof the results can be provided if required.
If you are not the appropriatepersonto fill in this questionnaire,may I ask you to passit on to
the appropriate person since it is an important part of my dissertation. Completed
in
questionnairesmay returned the FREEPOST envelopeprovided. No postageis required.
be
Alternatively, the fax and email addresses,as indicated below, may be used to return the
questionnaires.
Fax: +44 (0)1234 752 207
Email: y. [email protected]
Should you have any questions or require any further information regarding the survey,
please do not hesitateto contact me. It is hoped that you will find time to make a contribution
to this research,which hopefully, will benefit the aviation industry. Thank you very much
indeed!
Yours faithfully
Iris HSU
Departmentof Human Factors and Air Transport
Cranfield University
15 November 2001
317
Ref. No.:
Airline Safety Management

Questionnaire
Please provide your point of view on these questions. All

information given will be treated as strictly confidential. Results of
this survey will be calculated in aggregate form so that neither
airlines nor individuals can be recognized.
Region: Q Africa Q Asia & Pacific Q Latin America & Caribbean

Q Europe Q Middle East Q North America
Position in your airline:
Airline (optional):
Name (optional):
Contact address (optional)
Tel:
Fax:
E.mail:
118
Appendix L Covering Letter & Questionnaire
This survey examines the factors which might influence the way your organization
deals with safety issues. Please respond with reference to your own organization.
For each question, please circle the number which corresponds to the extent to which
you agree or disagree with the statement.
strongly strongly
disagree agree NIA
-
1. There is an organizational awareness towards 1234567
safety in the airline.
2. Employeesknow how to performtheir job in a 1234567
safe manner.
3. Employeesuse correct safety proceduresto 1234567
carryingout the jobs.
4. Employeesensurethe highest levels of safety 1234567
when they carry out the jobs.
5. Employeesbelieveflight safety is an important 1234567
issue.
6. Employees feel that it is important to maintain 1234567
safety at all times.
7. Employeesvoluntarilycarry out the tasks or 1234567
activitiesthat help to improvesafety.
8. Employeesare encouragedto submit ideas to 1234567
improvesafety in the airline.
9. Safety rules can be followedwithout conflicting 1234567
with work practices.
10. Safety problemsare openly discussed between 1234567
employees. -
11. There is good communication between different 1234567
groups in the airline.
12. There is an appropriate Emergency Response
Plan. 1234567
13. After an accident has occurred, appropriate
actions are usually taken to reduce the chance of
reoccurrence. 1234567
14. After an incident has occurred, appropriate
actions are usually taken to reduce the chance of 1234567
reoccurrence.
15. There is a documented business continuity plan 1234567
in the event of accidents.
strongly strongly
16. Changes in working procedures and their effect disagree agree N/A
on safety are effectively communicated to
employees. 1 2 3 4 5 6 7
17. Personnel are confident about their future with
the airline. 1 2 3 4 5 6 7
18. Morale is good. 1 2 3 4 5 6 7
19. Good working relationships exist in the airline. 1 2 3 4 5 6 7
20. Employees' jobs are well defined. 1 2 3 4 5 6 7
21. The size of the airline has an influence on
organizational safety culture. 1 2 3 4 5 6 7
22. The airline's history has an influence on
23. Airline ownership has an influence on
24. An adequate system exists for transmitting
critical information regarding safety within the 1 2 3 4 5 6 7
airline.
25. An adequate system exists for exchanging 1 2 3 4 5 6 7
critical information regarding safety problem with
other airlines.
26. Data* collection, analysis and presentation has 1 2 3 4 5 6 7
an influence on safety performance.
( *Data obtainedfrom the operation,risk assessment,
maintenance,and aircraft manufactures)
27. Management is concerned for the cost more than 1 2 3 4 5 6 7
safety.
28. Safety budget is the first item to be reduced
when commercial pressures emerge. 1 2 3 4 5 6 7
29. Safety rules are adhered to even under cost
pressures. 1 2 3 4 5 6 7
30. There is no conflict between safety and financial
goal. 1 2 3 4 5 6 7
31. Shareholder's welfare and airline's organizational
safety culture are correlated. 1 2 3 4 5 6 7
32. Safety committee has an influence on
33. Senior management commitment plays an
important role in determining the safety 1 2 3 4 5 6 7
performance.
strongly.. strongly
, --------,.
disagree agree NIA
34. The values of managementare identifiedas
being safety orientated. 12 3 4 5 6 7
35. Safety management aims are sufficiently
supported within the airline. 12 3 4 5 6 7
36. Leadership has an influence on organizational
safety culture. 12 3 4 5 6 7
37. There is correlation between the quality
assurance system and organizational safety
culture. 12 3 4 5 6 7
38. An effective ongoing hazard identification
program has an influence on organizational
safety culture. 12 3 4 5 6 7
39. Confidential reports should be properly de-
identified in order to foster organizational safety
culture. 12 3 4 5 6 7
40. There should be a procedure established for
acknowledging safety-related reports. 12 3 4 5 6 7
41. Risk audit, risk assessment, and risk evaluation
have an influence on organizational safety
culture. 12 3 4 5 6 7
42. Potential errors, consequences and recovery
points are identified in training. 12 3 4 5 6 7
43. Training is carried out by individuals with relevant
operational experience. 12 3 4 5 6 7
44. Workload is reasonably balanced. 12 3 4 5 6 7
45. Frustrations that arise from factors outside staff
control can be accommodated without adversely 12 3 4 5 6 7
affecting work.
46. There are adequate opportunities to express the
views about operational problems. 12 3 4 5 6 7
47. Final decisions about safety investment are
made by Chief Executive Officer (CEO). 12 3 4 5 6 7
48. There is an effective mechanism by which the
safety manager or the safety committee can
report to the CEO and make recommendations 12 3 4 5 6 7
for a change or action.
AppendixL CoveringLetter & Questionnaire
49. In the event of CEO making an unfavorable

response to a safety recommendation, there is a
procedure whereby the matter is monitored by
the safety manager or the safety committee until
resolved.
50. Personnel's decision-making is affected by the
organizational safety culture
51. Safety statements and policies of an airline
define the management's intention in safety
matters and airline's commitment to safety.
52. The policy statements define the airline's
fundamental approach towards safety.
53. The airline has a clearly stated set of goals and
objectives.
54. The roles and responsibilities for personnel in the
safety management system are clearly defined
and documented.
55. An effective documentation management system
ensures the availability of procedures.
56. Written work procedures match the way tasks are
done in practice.
57. The recommendations and suggestions from the
industry safety committee are adhered to all the
time.
58. Safety information from aviation safety
authorities is highly valued.
59. The regulations from the aviation safety authority
have an influence on organizational safety culture.
60. The relationship with the Media has an influence
on organizational safety culture.
61. The customer's reaction has an influence on the
organizational safety culture.
62. Consumer habits have an influence on
organizational safety culture.
AppendixL CoveringLetter & Questionnaire
strongly strongly
E___..
disagree agree NIA
63. The perceivedcorporateimage has an influenceon
organizationalsafety culture 1 2 3 4 5 6 7
64. The organization'sinvestmentson other
businessare correlatedwith the organizational
safety. 1 2 3 4 5 6 7
65. The investors'prospectivehas an influenceon
organizationalculture. 1 2 3 4 5 6 7
66. The stock market's reaction has an influence on
organizational culture. 1 2 3 4 5 6 7
67. The country culture and organizational culture
are correlated. 1 2 3 4 5 6 7
68. The country economic influence and
organizational culture are correlated. 1 2 3 4 5 6 7
69. Regional geographical influence and
organizational safety culture are correlated. 1 2 3 4 5 6 7
70. Regional economical influence and
organizational safety are correlated. 1 2 3 4 5 6 7
71. Regional religion influence and organizational
culture are correlated. 1 2 3 4 5 6 7
72. Regional cultural influence and organizational
safety are correlated. 1 2 3 4 5 6 7
73. How would you rate the safety performance of
your company with respect to the rest of your 1234567
industry? Below Above
(1: below Averagefor Industry Average Average
6: Above Average for Industry)
Please offer any suggestions or comments if any:
Thank you very much for your time. Your help has been highly appreciated!!
This pagehasbeenleft blank intentionally

APPENDIX M Participants Comments
1. The answersabove incorporate both flight safety and ground safety.The level
of safety awarenessamong flight crews is higher than the level among ground
crew staffs and management.A lot of progresshas been made in that domains
since the establishmentof Corporate Safety a year ago. Still, a lot of effort is
required to achieve a fully comprehensiveSafety ManagementSystem.
2. Previous poor managementis very hard to overcome. The employeesbecome
very untrusting and unwilling to accept that new managementcares about

safety.
3. Our company is a young one and is the flight safety department.The present
answersare from similar audits we made internally, a couple months ago.
4. Bottom Line - if CEO is highly safety consciousthen so is organization. If he

is not- the culture is not established.
5. We are committed to the safest flight operation possible. Therefore, it is

imperative that we have uninhibited reporting of all incidents and occurrences
that in any way affect the safety of our operations. We urge to every single
employee to help us in providing our outcomes and our employees with the
highest level of flight safety achievable within our working environment.
6. Q'67-72 is my case.I think in some region of the world, the answerwould be
very different.
7. We are a regional airlines serving a domestic/international parent airlines.

Group integrates do sometimes affect/influence issues relating to customer
focus, your survey doesn'taddressthat issue.
325
Appendix M Participants Comments
8. Our CEO is also the accountablemanagementperson agreed with our local
regulatory authority. This also has a big influence in out airlines safety culture.
9. The safety department is only five months old and the above reflects the
existing situation which I hope to improve on safety (health).
10. To question 31,1 am sure shareholdersthink differently - they don't see the
correlation.
11. Question47 investmentdecision doesn'thave to go to CEO that far.
12. In the last 4 questions I was not sure whether regional reflects to a region
within the i.
world, e. middle east, sub-Saharan Africa, etc. or if it means region
within a country.My responseswere basedon regions within a country.
13. The concept that safety should come from the "Top" is completely true.
Another point is that many times and at many organisations. The upper
managementdoes not know well about safety; they are at the kindergarten
level while safety people are at "PhD". Lack of understanding and
communication.
326
APPENDIX N Rotated Component Matrix for Internal &
External Loading Factors
1) Internal Factors -
Variables
Loading Factors
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
V03 0.815 0.170 0.126 0.026 0.146 0.212 0.112 0.074

0.117 -0.027 0.083 0.023 -0.012 0.016 0.041
V04 0.774 0.034 0.103 0.088 0.159 0.069 0.172 -0.059 0.248 0.100 0.061 0.020 -0.050 0.084 -0.014
V07 0.640 0.384 0.080 0.242 0.160 -0.013 0.054 -0.047 -0.085 -0.154 0.172 0.052 0.002 0.064 0.056
V02 0.608 0.170 0.224 -0.010 0.049 0.409 0.078 0.103 0.196 0.096 -0.119 -0.030 0.100 -0.026 0.004
Vol 0.462 0.156 0.158 0.221 0.197 0.198 0.175 0.242 0.361 0.109 0.052 -0.021 -0.057 0.091 -0.051
V09 0.161 0.834 0.170 0.121 0.058 0.043 0.054 0.163 0.241 0.064 0.002 -0.027 -0.002 0.076 -0.016
VOS 0.166 0.830 0.187 0.125 0.065 0.057 0.073 0.158 0.228 0.064 0.015 -0.026 0.007 0.084 -0.020
V06 0.550 0.609 0.089 -0.005 -0.056 0.099 -0.076 0.050 0.170 0.048 -0.025 0.127 0.180
0.196 0.148
V08 0.296 0.537 -0.076 0.049 0.236 0.321 0.030 0.204 -0.081 -0.022 -0.113 0.229 0.138 -0.103 -0.026
V35 0.076 0.434 -0.060 0.317 0.158 0.387 0.130 0.055 0.228 0.279 0.429 -0.030 -0.138 0.026 0.017
V46 0.097 0.106 0.940 0.054 0.118 0.073 0.054 0.041 0.092 0.025 0.042 0.045 0.011 0.017 0.031
V43 0.101 0.082 0.930 0.069 0.116 0.098 0.045 0.035 0.095 -0.011 0.064 0.038 -0.008 -0.019 -0.005
V48 0.138 0.127 0.885 0.051 0.143 0.057 0.035 0.048 0.064 -0.002 0.027 -0.013 0.045 0.096 0.047
V42 0.343 0.086 0.420 0.409 0.206 0.200 0.090 -0.060 -0.031 0.097 -0.039 0.140 0.056 -0.176 0.000
V53 0.049 0.194 0.085 0.821 0.016 0.055 0.178 0.152 0.092 0.088 0.008 0.050 0.094 -0.087 -0.046
V52 0.068 0.019 0.077 0.763 0.034 0.118 0.235 0.208 0.164 0.154 0.043 -0.017 0.166 0.119 -0.088
V55 0.274 0.260 0.069 0.568 0.063 0.152 -0.013 0.081 0.294 -0.018 0.136 0.099 0.022 -0.239 0.106
V51 0.189 0.033 0.133 0.517 -0.114 0.032 0.300 0.345 0.166 0.122 -0.151 -0.023 0.117 0.242 -0.254
V17 0.107 0.036 0.058 0.018 0.825 0.032 0.176 -0.003 0.075 0.170 -0.001 0.212 0.037 0.117 -0.102
V18 0.230 0.051 0.173, -0.013. 0.821 0.066 -0.001 0.067 0.104 0.049 0.169 -0.080 -0.154 0.007 -0.058
V45 0.080 0.286 0.229 0.175 0.619 0.014 0.072 0.068 -0.016 -0.070 -0.060 -0.006 0.010 0.139
-0.293
V19 0.282 0.086 0.222 -0.008 0.599 0.085 0.122 0.253 0.279 0.044 -0.022 -0.057 -0.142 0.176 -0.029
V20 0.299 0.091 0.163 0.314 0.345 0.032 0.176 -0.003 0.075 0.170 -0.001 0.212 0.037 0.117 -0.102
V26 0.199 0.035 0.196 0.100 0.106 0.718 0.106 0.138 0.060 -0.040 0.060 0.124 -0.172 0.050 0.033
V38 0.191 0.096 0.034 0.043 0.285 0.685 0.285 -0.067 0.166 -0.012 -0.026 -0.052 0.188 0.240 0.020
V24 0.180 -0.046 0.415 0.237 -0.100 0.515 -0.100 0.163 -0.071 0.132 0.087 0.189 0.199 0.078 0.021
VII 0.181 0.047 0.064 0.197 0.136 0.155 0.849 0.063 0.041 0.049 0.093 0.039 0.052 0.015 0.141
V12 0.160 0.043 0.061 0.237 0.145 0.172 0.777 0.088 0.052 0.038 0.122 0.012 0.057 0.003 0.109
V13 0.160 0.189 0.224 -0.033 0.277 0.151 0.738 0.115 0.176 0.324 -0.030 0.328 0.151 -0.091 0.160
V10 0.001 0.324 0.124 -0.026 0.113 0.144 0.700 0.166 0.415 0.162 -0.087 0.094 0.185 -0.155 -0.014
V30 0.084 0.152 0.016 0.204 0.102 0.079 0.059 0.876 0.085 0.107 0.061 -0.002 0.033 -0.076 0.034
V31 0.232 0.228 0.105 0.179 0.104 0.081 0.093 0.854 0.078 0.136 -0.051 0.000 0.018 -0.037 -0.012
V54 0.399 0.047 0.064 0.197 0.117 -0.056 -0.117 0.064 0.669 0.107 0.108 0.035 0.014 -0.033 0.132
V56 -0.051 0.043 0.061 0.123 0.129 0.070 0.082 0.126 0.576 0.086 0.270 0.121 0.160 -0.045 0.097
V49 0.165 0.216 0.160 0.325 0.151 0.247 0.190 0.008 0.536 0.087 0.023
0.169 -0.143 0.137 -0.103
0.035 -0.092 0.057 -0.170 -0.234 -0.098 -0.156 -0.818 -0.023 -0.005 -0.031 0.027 -0.109
V27 -0.006 0.231
V28 0.278 0.047 0.089 0.229 0.197 -0.123 0.261 -0.108 -0.044 -0.675 0.161 0.197 -0.104 0.157 0.203
V34 0.301 -0.160 -0.084 0.193 0.188 -0.096 0.241 0.185 0.026 0.534 0.132 0.381 -0.220 0.173 0.056
V29 -0.006 0.250 0.160 0.048 0.296 -0.155 0.153 0.176 -0.014 0.517 0.157 0.413 -0.230 0.175 0.121
V21 0.226 0.189 0.214 -0.025 -0.214 0.300 0.047 -0.144 0.141 -0.142 0.695 0.122 0.070 -0.119 0.002
V22 0.052 -0.082 0.045 0.323 0.189 0.391 0.130 0.098 -0.020 0.046 0.665 -0.047 0.075 0.151 -0.097
V33 0.139 0.067 0.162 -0.168 -0.114 0.011 0.311 0.077 0.236 0.278 0.437 -0.049 -0.100 0.047 -0.003
327
Appendix N Rotated Component Matrix for Internal & External factors
V23 0.325 0.277 0.033 0.260 0.129 0.149 0.214 0.299 0.240 0.008 0.403 0.144 0.072
-0.178 0.002
V25 0.430 -0.006 0.094 0.067 0.298 0.172 0.244 -0.103 0.296 -0.053 -0.026 0.605 0.147 0.179 0.042
0.276 0.046 -0.013 1-0.033 0.277 0.151 0.091 0.037 -0.022 0.071 -0.125 -0.502 0.101 0.180
V50 0.184
V15 -0.035 0.135 0.045 0.177 -0.136 0.065 0.111 0.261 0.223 0.033 -0.069 0.441 -0.023 -0.123 -0.097
V16 0.073 0.189 0.224 0.240 0.012 0.014 0.043 0.115 0.176 0.324, -0.030 0.328 0.151 -0.091 0.160
V39 0.124 -0.029 0.011 -0.064 0.037 0.352 0.036 0.018 0.052 0.084 00095 0.036 0.817 0.079
-0.141
V40 0.035 0.266 0.288 -0.119 -0.094 0.037 0.223 0.156 0.054 -0.281 -0.026 -0.107 0.528 0.345 0.256
V41 0.294 0.169 0.090 0.051 0.049 0.184 -0.041 -0.128 -0.037 -0.095 0.031 0.037 0.181 0.701 0.037
V47 0.161 -0.017 0.060 0.123 0.117 -0.057 -3.557 -0.031 0.094 0.009 -0.103 -0.031 -0.100 0.001 0.820
V32 0.166 0.285 0.040 0.325 0.129 0.080 0.071 0.292 -0.062 0.243 0.182 0.074 0.087 0.268 0.392
2) External Factors -
Factors
Variables 1 2 3 4
70 0.879 0.127 0.004 0.113
69 0.857 0.123 0.000 -0.020
72 0.804 0.154 0.076 0.027
68 0.715 0.164 0.096
-0.230
71 0.685 0.168 0.296
-0.150
61 0.034 0.881 0.003 0.053
63 0.103 0.777 0.035 0.054
62 0.083 0.757 -0.029 0.210
60 0.181 0.546 0.435 0.283
58 0.809
-0.038 -0.005 -0.072
59 0.165 0.756 0.273
-0.019
57 0.097 0.750
-0.051 -0.017
65 0.038 0.226 0.011 0.797
66 0.242 -0.079 0.024 0.788
64 0.049 0.331 0.131 0.555
67 0.180 0.027 0.187 0.219
328
APPENDIX 0 Alpha Values for Internal & External Factors
1) Internal Factors -
Factor
1 2 3 4 5 6 7 8 9 10 11 12 13
V03 815
.
V04 774
.
V07 640
.
V02 608
.
Vol 462
.
V09 834
.
V05 830
.
V06 609
.
V08 537
.
V35 434
.
V46 940
.
V43 930
.
V48 885
.
V42 420
.
V53 821
.
V52 763
.
V55 568
.
V51 517
.
V17 825
.
V18 821
.
V45 619
.
V19 599
.
V26 718
.
V38 685
.
V24 515
.
Vil 849
.
V12 777
.
V13 738
.
VIO 700
.
V30 876
.
V31 854
.
V54 669
.
V56 576
.
V49 536
.
V27 -. 818
V28 -. 675
V34 534
.
V29 517
.
V21 695
.
V22 665
.
V33 437
.
V23 403
.
V25 605
.
V15 441
.
V16 328
.
V39 817
.
V40 528
.
AI ha 85 846 89 83 82 74 759 93 72 04 61 66 54
. . . . . . . . . . . . .
329
Appendix O Alpha Valuesfor Internal and External Factors
2) External Factors -
Factors
Variables
Factorl Factor 2 Factor 3 Factor 4
70 879
.
69 857
.
72 804
.
68 715
.
71 685
.
61 881
.
63 .
777
62 757
.
60 546
.
58 809
.
59 756
.
57 750
.
65 797
.
66 788
.
64 .
555
Reliability 0.8575 0.7854 0.686 0.636

(Alpha)
330
APPENDIX P Mean and Standard Deviation of
Internal and External Factor Scores
1) Internal factors -
Factor 1
Region #Casesavailable Mean Std. Deviation
North America 22 8892058
-0.63922 .
Africa 3 0.35631 1.1738282
Europe 36 0.31140 9568387
.
Asia and Pacific 36 0.07855 1.0234124
LatinAmericaandtheCaribbean 5 -0.30452 4907790
.
Middle East 2 0.02391 5595636
.
Factor 2
Region #Cases available Mean Std. Deviation
North America 22 0.05703 9077366
.
Africa 3 0.29888 2097007
.
Europe 36 0.02356 1.0386296
Asia and Pacific 36 0.13639 8940901
.
LatinAmericaand the Caribbean 5 -1.20469 1.1258093
Middle East 2 -0.94312 2.1768678
Factor 3
North America 22 -. 1167249 1.2608643
Africa 3 4161720 5230922
. .
Europe 36 1434193 8517345
. .
Asia and Pacific 36 -. 1018476 1.0343227
LatinAmericaandtheCaribbean 5 2528859 7532800
. .
Middle East 2 -. 7207907 8191174
.
Factor 4
North America 22 0.27234 8722531
.
Africa 3 0.20191 6216546
.
Europe 36 -0.44308 1.2119959
.
LatinAmericaand the Caribbean 5 0.13786 4579773
.
Middle East 2 -0.07165 6006844
.
Factor 5
North America 22 0.05100 9153141
.
Africa 3 0.01032 5199314
.
Europe 36 1.1199545
-0.05361
5
.
LatinAmericaandtheCaribbean 0.14110 8994528
Middle East .
2 2.4425141
-0.38663
331
Appendix P Means and Standard Deviation of Internal and External Factor Scores
Factor 6
North America 22 0.16137 7284979
.
Africa 3 -0.43425 1.1351994
Europe 36 -0.08888 1.1968939
.
LatinAmericaand the Caribbean 5 -0.09189 1.1633706
Middle East 2 -0.92322 1899493
.
Factor 7
North America 22 0.06720 1.1130961
Africa 3 0.55952 3892361
.
Europe 36 0.07701 8364943
.
Asia and Pacific 36 -0.07538 1.1201511
LatinAmericaand the Caribbean 5 7475811
-0.54081 .
Middle East 2 -0.25566 1.7282853
Factor 8
North America 22 0.32917 9702956
.
Africa 3 0.12134 6369803
.
Europe 36 1.1318192
-0.15066
Asia and Pacific 36 9148090
-0.01434 .
LatinAmericaandtheCaribbean 5 -0.05188 1.1005971
Middle East 2 -0.15911 1319623
.
Factor 9
North America 22 0.17659 9438688
.
Africa 3 0.37270 1.1803107
Europe 36 -0.08588 1.0087299
Asia and Pacific 36 -0.02776 9871342
5
.
LatinAmericaand the Caribbean 0.03291 1.4968176
Middle East 2 9182260
-0.53837 .
Factor 10
North America 22 0.05404 1.1028550
Africa 3 -0.25663 4709795
.
Europe 36 0.10972 1.0229031
.
LatinAmericaandtheCaribbean 5 -0.12281 1.1838685
Middle East 2 -0.29334 1.2140648
332
Factor 11
North America 22 -0.05329 1.0395035
Africa 3 0.81012 5241767
.
Europe 36 -0.07998 1.1712296
Asia and Pacific 36 -0.05820 7823002
.
LatinAmericaandtheCaribbean 5 0.35314 1.1636854
Middle East 2 0.97525 2.382782E-02
Factor 12
North America 22 0.31854 1.0178718
Africa 3 0.34801 6752834
.
Europe 36 0.02561 9357360
.
Asia and Pacific 36 -0.30011 1.0211141
Latin America and the Caribbean 5 0.17671 1.2877234
Middle East 2 0.47327 3039628
.
Factor 13
North America 22 0.02816 8662379
.
Africa 3 0.11566 4541094
.
Europe 36 0.16807 1.0695120
Asia and Pacific 36 -0.21146 1.1172311
.
Middle East 2 0.31609 1977127
.
333
2) External factors -
Factor El
North America 22 -0.09102 9365840
.
Africa 3 0.93939 1280186
.
Europe 36 -0.00948 9958194
.
Asia and Pacific 36 -0.05993 1.1070402
LatinAmericaandtheCaribbean 5 0.28506 9391227
.
Middle East 2 0.12890 1224544
.
Factor E2
North America 22 -0.03430 1.0597765
Africa 3 0.67918 3656843
.
Europe 36 -0.14582 1.1735208
.
LatinAmericaand the Caribbean 5 1.2493247
-0.03162
Middle East 2 1.06694 1.0080965
Factor E3
North America 22 -0.51725 1.1415366
Africa 3 0.78367 5608238
.
Europe 36 9389754
-0.04268 .
.
.
Middle East 2 0.10613 1.2918311
Factor E4
North America 22 -. 1713927 1.1113964
Africa 3 -. 4350394 7676257
.
Europe 36 -. 1214183 1.0891926
Asia and Pacific 36 1732103 8271913
. .
LatinAmericaand the Caribbean 5 5836195 1.1682217
.
Middle East 2 1465730 6180094
. .
334
APPENDIX Q Table of ANOVA and Mean/Standard Deviation
for Internal & External Factors
1) Internal factors -
Factor
Analys is of Variance (ANOVA)
Source Sum of Squares DF Mean Squares F Si
.
Between Groups 13.661 5 2.73 2.99 01
.
Within Groups 89.33 9 91
.
Total 103.00 103

North America 22 4.4182 8093
.
Africa 3 4.9333 8327
.
Europe 36 4.8722 6764
.
.
atin Americaand the Caribbean 5 4.3600 6693
.
Middle East 2 4.2000 1.697
Factor 2
Source Sum of Squares DF Mean Squares F Sig.
Between Groups 10.065 5 2.013 2.12 06
.
.
Total 103.00 10

.
Africa 3 5.5333 2309
.
Europe 36 5.1667 6616
.
Asia Pacific 36 5.1500 6868
.
atinAmericaandtheCaribbean 5 4.3200 0640
.
Middle East 2 4.2000 2627
.
Factor 3
.
Between Groups 3.292 5 65 64 66
. . .
Within Groups 99.70 9 1.017
Total 103.00 10

.
Africa 3 5.4167 5774
.
Europe 36 4.9861 7652
.
.
Latin America and the Caribbean 5 4.8000 5420
.
.
335
Appendix Q TableofANOVA and Mean/StandardDeviationfor Internal &cExternal Factors
Factor 4
Between Groups 11.082 2.21 2.36 04

.
.
Total 103.000 1031

.
Africa 3 5.5000 8660
.
Europe 36 4.5000 9524
.
.
.
.
Factor 5

.
Between Groups 579 5 11 111 99
. . . .
Total 103.00 10

.
Africa 3 3.8333 0408
.
Europe 36 3.8194 9975
.
5
.
.
.
Factor 6

Between Groups 3.465 5 69 682 63
. . .
Total 103.00 10

.
Africa 3 5.1667 5204
.
Europe 36 4.9028 8091
.
.
.
Middle East 2 4.2500 1.4749
Appendix Q TableofANOVA and Mean/StandardDeviation for Internal & External Factors
Factor 7

Between Groups 3.05 61 598 701
. . .
Total 103.00 10

North America 22 4.6818 1.0065
Africa 3 5.0000 1.0000
Europe 36 5.0833 1.1307
Asia Pacific 36 4.8056 1.1166
.
Middle East 2 4.0000 2.8284
Factor 8

Source Sum of S uares DF Mean S uares F Si
.
. . .
Total 103.00 10

Africa 3 4.3333 5774
.
Europe 36 3.8472 1.2468
Asia and Pacific 36 4.0556 1.1388
Middle East 2 3.7500 1.0607
Factor 9

. . .
Total 103.00 10

.
Africa 3 5.2500 4330
.
Europe 36 4.3125 8668
.
.
.
Appendix Q TableofANOVA and Mean/StandardDeviationfor Internal & External Factors
Factor 10

. . .
Total 103.00 10

.
Africa 3 3.5833 6292
.
Europe 36 3.7847 6385
.
.
.
.
Factor 11

. . .
Total 103.00 10

.
Africa 3 5.5000 4330
.
Europe 36 4.9444 6947
.
.
Latin Americaand the Caribbean 5 4.5000 1.1319
.
Factor 12

Between Groups 6.466 5 1.29 1.31 26
.
.
Total 103.00 10

.
Africa 3 4.6667 3819
.
Europe 36 4.2292 6824
.
.
.
Middle East 2 4.0000 1.4142
Appendix Q TableofANOVA and Mean/StandardDeviation for Internal & External Factors
Factor 13

Source Sum of S uares DF Mean S uares F Sig.
. . .
Total 103.000 1 10

.
Africa 3 5.5000 8660
.
Europe 36 5.2361 7120
.
.
.
.
2) External factors -
Factor El

.
. . .
Total 103.00 10

.
Africa 3 4.667 306
.
Europe 36 3.628 1.028
Asia and Pacific 36 3.639 1.113
.
.
Factor E2

. . .
Total 103.00 10

Africa 3 5.0000 2500
.
Europe 36 4.1111 1.1235
: L .
.
.
Appendix Q Table ofANOVA and Mean/Standard Deviation for Internal & External Factors
Factor E3
Between Groups 11.383 5 2.27 2.43 04
.
.
Total 103.00 103

Africa 3 5.5556 5092
.
Europe 36 4.6481 8124
.
.
.
Middle East 2 4.6667 1.4142
Factor E4
Analysi s of Variance (ANOVA)
.
. . .
Total 103.00 10

Africa 3 3.7778 1.0715
Europe 36 3.6759 9904
.
5 4.4667
.
LatinAmericaandthe Caribbean 1.2824
.
APPENDIX R Multiple Regression Result
Self-rated Safety Performance vs. Organisational Factors
Multiple R 733
.
R2 538
.
RZAdjusted 619
.
StandardError 535
.
Analysis of Variance
DF Sum of Squares Mean Squares FP
Regression 4 33.007 8.252 28.822 000
.
Residual 99 28.343 286
.
Variables in Equation
Variable B SEB Beta
Factor 2 403 083 388
. . .
Factor 12 244 073 275
. . .
Factor 7 113 051 164
. . .
Factor 3 133 064 158
. . .
(Constant) 486 413
. .
341
APPENDIX S Correlation of Internal and External Factors
E1 E2 E3 E4
Factor 1 008 216* 421** 146
. . . .
Factor 2 -. 021 272** 314** 141
. . .
Factor 3 -. 059 055 455** 165
. . .
Factor 4 020 163 352** 235*
. . . .
Factor 5 -. 127 107 394** 150
. . .
Factor 6 014 155 327** 155
. . . .
Factor 7 -. 012 164 355** 219*
. . .
Factor 8 -. 050 047 224* 198*
. . .
Factor 9 -. 069 244* 419** 121
. . .
Factor 10 198* 131 207* 226*
. . . .
Factor 11 216* 367** 339** 109
. . . .
Factor 12 -. 105 118 371** 079
. . .
Factorl3 087 017 25* 142
. . . .
* Correlation is significant at the 0.05 level (2-tailed).
Correlation is significant at the 0.01 level (2-tailed).
Internal Factors External Factors

1: Employeesafety attitude & behaviour El: Influencesof region and country
2: Employeesafety concept E2: Public and the media influence
3: Level of operationalsafety in E3: Impact of regulatory environment
operation and maintenance E4: Involvementof investment
4: Corporate safety policy community
5: Personnel-quality of working life
6: Employment of risk programme
7: Impact of accident/incidents
-
8: Financial concern
9: Proceduresand documentation
10: Commercialcost pressures
11: Organisationalstructure &
managementcommitment
12: Communicationsystem
13: Necessityof safety reports
342
APPENDIX T DMAIC Process in Six Sigma
Y=F (X1,.. X)
Where Y is the output (business focus)
Xs are the input (processes)
Phase Action Goal

What are the critical factors to the business
Define focus (from internal and/or external
Define "Y"
(D) customer point of view)?
What and how is the process (The Y)

Measure
performing? Measure the "Y"
(M)
How reliable is the obtained data?
What are the critical defects (Xs) causing

Analyse Find and Measure
deficiency?
(A) the "Xs"
To what degree of the variation?
( Xs: room for improvement)
How to fix the critical defect (Xs)?

Improve
What variation in critical Xs can be Improve the "Xs "
(Iý
removed?
Control Control "Xs" so no

How to maintain the improvement?
(C) variation in " Y"
Source: combined from various authors
343

139739

Uploaded by

Copyright:

Available Formats

139739

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

139739

Uploaded by

Copyright:

Available Formats

CRANFIELD UNIVERSITY

Yueh-Ling Hsu '`,

Airline Safety Management:

AIR TRANSPORT GROUP

Academic Year 2003-2004

Airline Safety Management:

Supervisor: Mr. A. Frank Taylor

This thesis is submitted in partial fulfillment of the requirements

an important but yet relatively undefined task in the airline industry is to

existing knowledge of what is thought to contribute to safety by adding an increased

organisational factors which exert an influence upon the safety mechanism.

Four hypotheses were set out to be tested in an attempt to justify the

performance needs the co-ordination of both retroactive and proactive safety

methodology to enable their safety management system to evolve at a fundamentally

First and foremost,my thanks go to my supervisor,Mr. Frank Taylor, whose technical

project throughout. In particular, thanks and love to my parents, whose endless

1.1 RESEARCH BACKGROUND 1

CHAPTER 2 LITERATURE REVIEW & EMPIRICAL ANALYSIS

2.1.1 Technical innovation and its impact 9

2.2.1 Flight safety risk 18

3.1 INTERviEw 100

3.2.1 Purpose of the model 102

CHAPTER 4 INTERVIEW FINDINGS & MODEL DEVELOPMENT

4.1 INTERVIEW FINDINGS 115

4.1.1 Safety management system 115

4.2.1 Overview and hypothesisof the model 130

CHAPTER 5 RESULTS OF MODEL TEST

5.0 INTRODUCTION 159

CHAPTER 6 ANALYSES AND DISCUSSIONS

6.0 INTRODUCTION 177

CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS

7.1 CONCLUSIONS- THESIS AIMS REVISITED 241

Figure 2-19 Layers of culture 74

Figure 2-21 Proactive defenceevaluation model 87

Table 2-3 Estimatesof Human Error in Different Industries 60

Table 2-5 Definitions of safety culture 82

Table 2-6 Various error types and remedial strategies 92

Table 3-1 The profile of selectedairlines 100

Table 5-2 The Factor structureof a proactive safety mechanism 176

Table 6-2 The underlying nature of factor 1 179

Table 6-15 The underlying nature of factor 14 186

Table 6-17 KMO and Bartlett's Test: External factors 188

Table 6-20 The underlying nature of factor E3 190

Table 6-22 Summariesof the survey results 197

safety mechanism 213

Table 6-25 The DMAIC processof X12"Communication system" 239

AHI Aviation Health Institute

BASIS British Airways Safety Information System

CRM Crew ResourceManagement

CST Central StandardTime

DFDR Digital Flight Data Recorder

ERCs Error Reducing Conditions

FSF Flight Safety Foundation

FOQA Flight OperationQuality Assurance

GPWS Ground Position Warning System

GMT Greenwich Mean Time

IATA InternationalAir TransportAssociation

ILU Institute of London Underwriters

LAUA Lloyd's Aviation Underwriters Association

LOSA Line Operation SafetyAudit