AFY533E Research

and Analysis Methods

Yurdanur Ünal
Objectives
 Be acquainted with the scientific method.
 Understand some of the different domains
and disciplines for scientific research of
emergency management.
 Understand the role of other scientific
disciplines in the practice of emergency
management.
 Understand one experiment demonstrating
the scientific method.
Research
Science
The aim of science is to provide new and useful information in the
form of data which can be verified in a way that allows other
researchers can make similar observations and obtain similar
results.
Research points out the relationships between events and the things
that affect them, and describes them to others.
Publications
Science has progressed over the centuries because researchers have
been able to trust the validity on their predecessor’s work and have
known about it because of publications.
Development
In the last century, scientific development has increased dramatically in
many disciplines because of a better understanding of the scientific
method, vastly improved research facilitates, and much better and
faster communication of results.
Research
 Emergency Management as an academic discipline is a new field
and those developing the academic aspects are drawing on many
other fields—some scientific, and others more related to technology
and management.
The scientific method
is a structured way of gathering, analyzing and drawing conclusions.

The “classic experiment” is typically conducted in a laboratory where

conditions can be tightly controlled.
The object
to determine the effects of a single independent variable by comparing its
effects on two groups by isolating the effect of the independent variable:
1. the independent variable – experimental
2. not receiving the independent variable – control
Hypothesis is generally stated before an experiment is performed,
which is believed to describe underlying behavior to be investigated.
The experimental method provides the most reliable source of scientific
information and is the preferred method of science.
 Relationship between Social Structure and
Disaster (from Kreps, 1989)
1. Magnitude of impact
2. Scope of impact
Events - Impacts
3. Length of warning
4. Duration of impact
Disaster Domains Social Structure

Physical, temporal and social

dimensions that can serve as
Social Units - Responses
either independent or dependent
variables. İmportant bridging cocept is the
For classification domain.

Focus on physical and temporal e.g. Floods have greater scope of

similarities and differences of impact than tornadoes.
impact in comparison.
Social Domains of Disaster
(from Kreps, 1989)
Domains
Care of fatalities
Hazard-vulnerability analysis
Providing victim basic needs
Maintenance of standby human and
material resources Damage needs and assessment
Disaster preparedness, planning and Damage control
training Restoration of essential services
Public education Public information
Hazard mitigation-structural Traffic control
Hazard mitigation-nonstructural Law enforcement
Insurance of predictions and warnings Local governance
Dissemination of predictions and Coordination and control
warnings Reconstruction planning
Evacuation Reconstruction of physical structures
Protective action Reestablishment of economic
Mobilization of emergency personnel functioning
and resources Resumption of other social institutions
Search and rescue Determining liability for the event
Medical care
Purpose of Research

Purpase might be more than one of them

 Exploration
 Description: what, where, when and how
To describe situations and events
Scientific descriptions are more acurate and precise.
 Explanation: why
Exploration

A researcher examines a new interest – the subject of the

study is new.
Focus groups – small group discussions
Done for 3 purposes
1. to satisfy he researcher’s curiosity  better understanding
2. to test the feasibilty  more extensive study
3. To develop methods to be used for successive research
 Exploration
Examples:
1. Taxpayer dissatisfaction
How wide spread it is?
What levels and degrees of support are there within the community?
How is the movement organized?
What kinds of people are active in it?
2. College’s graduation requierements
History of requirements,
Learn the reasons
3. Focus groups : Channeling
Who are they?
Where do they come from?
Why are they doing this?
How can we tell if they are real or fake?
4. Why people still live flood prone areas?
Who are they?
What are their education level?
What are their social status? etc.
 Exploration
Show
Where research begins.
Whereas researchers working from the deductive theories have the
key variables in advance
Identification some of the possible variables.
Classification
gender
education level
religious background
origin
 Grounded theory sources
 Exploration
Shortcomings
Seldom provide satisfactory answers but might suggest which
research methods could provide definitive answers

Problem is their representativeness – sample problem

 Description
To decribe the situation and events.
observe  describe
Scientific observation is careful and deliberate,
Scientific decriptions are typically more acurate and
precise.
e.g. National Cencus records.
Crime rates of a city
Gender profile of a population
Qualitative studies aim at decription.
 Researchers continue to examine why the observed
patterns exist and what they imply.
 Explanation
Describing studies anwers the questions
what
where
when
how
However, explatory research anwers
why
e.g. Why some cities are higher crime rates?
Why still some people live in regions prone to natural
hazards.
 The logic of nomothetic explanation
A nomothetic explanation is probabilistic and usually not complete.
Criteria
1. Variables must be correlated
2. The cause takes place before the effect
3. The variables are nonspurious
Corelation: relationship between two variables
Time Order: the cause precedes the effect in time.
Nonspurious: the effect cant be expalined by third variable. Coincidental
statistical correlation.
False Criteria
Complete causation:
Exceptional cases
Majority of cases
Necessary and Sufficient Causes:
N:must be present to effect to follow,
S:if it is present, guarantees the effect.
 Units of Analysis
The what and whom being studied
 Individuals
 Groups
 Organizations
 Social Artifacts
e.g. Mariages – marriage partners; crimes-criminals;
corporations – corporate executives.
 The time dimension
 Time sequence of events
 Generalizability of research findings
Cross-Sectional Studies: observations of a sample.
Exploratory and descritive studies are cross sectional.-
snapshot.
Longitudinal studies: observations of the same phenomena
over an extended period of time. Field research
projects.
 Trend studies
 Cohort Studies: age group
 Panel Studies
Scientific disciplines which researchers studying
emergency management might encounter

New technologies are now being widely used in the

practice of emergency management -- an inter-
disciplinary dimension.
For example, geographic information systems (GIS).
Increase of flood risk due to urbanization- use of satelate
images.
Bangladesh e-Journal of Sociology. Vol. 1. No. 2. July, 2004.
Disaster Research: Exploring Sociological Approach to Disaster in
Bangladesh
Mahbuba Nasreen
Survey of the general literature on disaster, Alexander (1993) identified six
schools of thought on natural hazards and disaster studies:
•the geographical approach,
•the anthropological approach,
•the sociological approach,
•the development studies approach,
•the disaster medicine approach and
•the technical approach.
The geographical approach (pioneered by Barrows, 1923 and White, 1945)
deals with the human ecological adaptation to the environment with special
emphasis on the 'spatio-temporal‘ distribution of hazard impacts, vulnerability
and people's choice and adjustment to natural hazards. Social science methods
are widely used in this approach.

The anthropological approach (Oliver-Smith, 1979, 1986; Hansen and Oliver-

Smith, 1982) emphasizes the role of disasters in guiding the socio-economic
evolution of populations. Anthropologists adopting this approach search for
reasons why communities in the 'Third World‘ fail to provide basic requirements
for their people's survival.

The sociological approach (Dynes, 1970; Qurantelli, 1978; Mileti, Drabek and
Haas, 1975; Drabek and Boggs, 1968; Drabek, 1986) discusses vulnerability
and the impact of disaster upon patterns of human behavior and the effects of
disaster upon community functions and organization.
Oliver-Smith (1996) developed three general themes as the major trends in
anthropological research in disaster: behavioral response approach, social change
approach, and political economic/ environmental approach. Oliver-Smith argues that
disaster in developing world occur at the interface of society, technology and
environment and is fundamentally the outcomes of the interactions of these
characteristics. He has also reported that although occurrence of disaster is
frequent, theoretical work in disaster research is limited.

The development studies approach (Davis, 1978; Knott, 1987) discusses the
problems of distributing aid and relief to 'Third World' countries and focuses on
refugee management, health care and the avoidance of starvation.

The disaster medicine and epidemiology approach (Beinin,1985) focuses on the

management of mass casualties. It also includes the treatment of severe physical
trauma and other diseases which may occur after a disaster.

The technical approach (Bolt et al. 1977; El-Sabh and Murty, 1988) focuses on
geophysical approaches to disaster such as studied in seismology, geomorphology
and volcanology and seeks engineering solutions.
Among these approaches two disciplines, geography and sociology, have
dominated the field of disaster research since the 1950s and have emphasized the
environmental and behavioral aspects of disaster.
Drabek's (1986) findings on existing sociological literature are the significant
contributions to the conceptual typology of sociological disaster research. He
identified different areas of concern in disaster research such as planning, warning,
evacuation, emergency, restoration, reconstruction, perceptions and adjustments.
He discussed sociology of disaster under four major headings: preparedness,
response, recovery and mitigation. However, most of the approaches and
sociological research on disaster have been formulated and conducted for the
developed world (especially the USA). Their application to developing areas is
problematic and very limited. Moreover, there is almost no discussion of the gender
response to disaster under any theoretical approach. In fact, only recently
sociologists turned their attentions to the larger questions of social change related
to disaster or the pre-impact conditions in disaster areas as sources of post-impact
changes

Oliver-Smith, Anthony. 1986. 'Disaster Context and Causation: An Overview of

Changing Perspective in Disaster Research' in Vinson H. Sutlive et al (eds) Natural
Disasters and Cultural Responses. Williamsburg: Department of Anthropology,
College of William and Mary.
 Emergency Management Issues and
Relationship to Science and Technology
Scientific Discipline or More Information
Emergency Technology
Management Issue
Behavior in Disasters 1.Sociology Kreps (1989)
Flood Predictions 1.Meteorology FEMA Internet site
2.Geographic Information (www.fema.gov)
Systems
3.Geography

Wildfire Spread 1.Physics FEMA Internet site (

2.Engineering www.fema.gov)
3.Forestry US Fire Administration. Wildfire:
Are You Prepared
Hazardous Materials 1.Chemistry Hazardous Materials Accidents:
2.Engineering Fact Sheet and many others on
FEMA web site
 Scientific Discipline More Information
Emergency
Management Issue
Hurricane resistant Engineering FEMA Mitigation Plan
building techniques Public administration (available from FEMA
Internet site)
Post-disaster stress Psychology
EXPERIMENTAL PROBLEM: To investigate the effects of caffeine
(contained in two cups of coffee) on the achievement of college
students on a final examination.
EXPERIMENTAL PROBLEM: To investigate the effects of caffeine
(contained in two cups of coffee) on the achievement of college
students on a final examination.

HYPOTHESIS: Two cups of black coffee taken immediately before

a task requiring mental exertion increase a student’s academic
efficiency.
PROCEDURE:
Two groups of subjects were used. Group 1 consisted of 200
college students who were matched in age, intelligence, sex, and
grade-point average with the 200 students in Group 2.
Subjects in both groups were enrolled in the elementary course in
public administration.
All subjects in Group 1 (experimental group) drank two cups of
black coffee immediately before taking the final examination.
All subjects in Group 2 (control group) were instructed not to take
any stimulants during the day the final examination was to be
taken.
For purposes of analysis, the grades of the students of both
groups were converted into the following numerical equivalents
(grade points): A = 4; B = 3; C = 2; D = 1; and F = 0. The average
grade-point score for each of the two groups was then computed.
RESULTS: The results of the experiment are summarized in the
tables. The table “Data from the Study” indicates the number and
percentage of students in each group obtaining each of the five
letter grades on the final exam in general psychology. The table
“Results of the Study” gives the average grade-point scores of the
two groups. These results indicate that the students in Group 1 did
consistently better than the students in Group 2.
 Data from the Study: Caffeine and Test Performance

Grade Group 1 Group 2

(drank coffee) (did not drink coffee)
Number Number
Percentage Percentage
A 30 15 14 7
B 44 22 22 11
C 82 41 112 56
D 36 18 38 19
F 8 4 12 6
 Results of the Study

Group Average Grade-Point

Score
1.(coffeedrinkers) 2.25
2.(non-coffee drinkers) 1.96

CONCLUSIONS: Comparison of the final examination grades

earned in Introduction to Public Administration by two groups of
college students (a stimulant-taking group and a non-stimulant-
taking group) indicates that taking a mild stimulant, such as two
cups of black coffee, immediately before an examination increases
the academic efficiency and achievement of college students in
the course.
 Exerscise

Design an experiment to test the effectiveness of a training

program to teach local emergency management planners how
to develop local Comprehensive Emergency Management
Plans.
A group of planners is provided with training including model
plans and a four-hour workshop.
Another group receives no such help.
We presume that the group with training will write better plans as
judged by independent ratings of the plans (provided by
experts).
 Exerscise

• independent variable

• dependent variable

• hypothesis

• experimental group

• control group
 Exerscise

• independent variable
(training)
• dependent variable
(ratings of plans by experts)
• hypothesis
(the group with training will develop better plans)
• experimental group
(those receiving training)
• control group
(those not receiving training).
What can go wrong with this experiment?

What are some of the extraneous variables which could

affect the results and make the interpretation unclear?