International Journal of Advanced Science and Technology

Vol.81 (2015), pp.35-48

http://dx.doi.org/10.14257/ijast.2015.81.04

Classification of Software Risks with Discriminant Analysis

Techniques in Software planning Development Process

Abdelrafe Elzamly1, Burairah Hussin2, Samy S. Abu Naser3, Mohamed Doheir4

1
Department of Computer Science, Faculty of Applied Sciences, Al-Aqsa
University, Gaza, Palestine
2,4
Information and Communication Technology, Universiti Teknikal Malaysia
Melaka (UTeM)
3
Faculty of Engineering & Information Technology, Al-Azhar University, Gaza,
Palestine

Abstract
The aim of this study is to propose the discriminant analysis (DA) techniques to
classify and manage risks in software planning development process. These techniques
are used to test the controls to each of risks to determine and classify if they are effective
in mitigating the occurrence of each risk planning factor. We classified risks to three
levels by predicted group membership. Also top ten risks planning software development
were mitigated by using controls based on discriminant analysis techniques.

Keywords: Risk Management, software Planning Risks, controls, Discriminant

Analysis Techniques, predictive risk model

1. Introduction
Despite much research and progress in the area of software project management,
software development projects still fail to deliver acceptable systems on time and
within budget [1]. Much of the failure could be avoided by managers pro-actively
maintenance and dealing with risk factors rather than waiting for problems to occur
and then trying to react. Due to the involvement of risk management in monitoring
the success of a software project, analyzing potential risks, and making decisions
about what to do about potential risks, the risk management is considere d the
planned control of risk. Integrating formal risk management with project
management is a new phenomenon in software engineering and product management
community. In addition, risk management methodology that has five phases: Risk
identification, risk analysis and evaluation, risk treatment, risk controlling, risk
communication and documentation relied on three categories or techniques as risk
qualitative analysis, risk quantitative analysis and risk mining analysis throughout
the life of a software project to meet the goals [2]. Today, we must think of risk is a
part of software project process and is important for a software project survival.
Risk management is a practice of controlling risk and practice con sists of processes,
methods, and tools for managing risks in a software project before they become
problems [3]. The objective of this study is: To classify the software planning risks
of software development in the software development organizations, to mode l the
activities performed for mitigating the risks planning software development which
identified.

2. Literature Review
Khanfar et al., [4], the new technique used the chi-square (χ2) test to control the
risks in a software project. However, we also used new techniques which are the

ISSN: 2005-4238 IJAST

Copyright ⓒ 2015 SERSC
International Journal of Advanced Science and Technology
Vol.81 (2015)

regression test and effect size test proposed to manage the risks in a software project
and reducing risk with software process improvement [5]. Also we improved quality
of software projects of the participating companies while estimating the quality–
affecting risks in IT software projects. The results show that there were 40 common
risks in software projects of IT companies in Palestine. The amount of technical and
non-technical difficulties was very large [6]. Furthermore, we used the new stepwise
regression technique to manage the risks in a software project. These tests were
performed using regression analysis to compare the controls to each of the risk
factors to determine if they are effective in mitigating the occurrence of each risk
factor implementation phase [7]. In addition, we proposed the new mining technique
that uses the fuzzy multiple regression analysis techniques to manage the risks in a
software project. However, these mining tests were performed using fuzzy multiple
regression analysis techniques to compare the risk management techniques to each
of the software risk factors to determine if they are effective in mitigating the
occurrence of each software risk factor[8]. This paper aimed to present new
techniques to determine if fuzzy and stepwise regression are effective in mitigating
the occurrence software risk factor in the implementation phase [9]. Additionally,
we proposed artifact model of the software risk management for mitigating risks. It
has the five levels to mitigate risks through software project [10]. Previous studies
had shown that risk mitigation in software projects classified into three categories–
namely, qualitative, quantitative, and mining approaches. Firstly, quantitative risk is
based on statistical methods that deal with accurate measurement about risk or
leading to quantitative inputs that helped to form a regression model to understand
how software project risk factors influence project success. Furthermore, qualitative
risk techniques lead to subjective opinions expressed or self-judgment by software
manager using techniques namely scenario analysis, Delphi analysis, brainstorming
session, and other subjective approach to mitigate risks. There are several software
risk management approaches, models, and framework according to a literature
review.

3. Top 10 Software Planning Risks in Software Development Process:

We displayed the top software planning risk factors in software development
project lifecycle that most common used by researchers when studying the risk in
software projects. However, the list consists of the 10 most serious risks to a project
ranked from one to ten, each risk's status, and the plan for addressing each risk.
These factors need to be addressed and thereafter need to be controlled. These
software project risks are illustrated [11]: Low key user involvement, Unrealistic
schedules and budgets, Unrealistic schedules and budgets, Unclear / misunderstood /
Unrealistic / change scope and objectives (goals), Insufficient/inappropriate staffing,
Lack of senior management commitment and technical leadership, Poor /inadequate
planning and strategic thinking, Lack of effective software project management
methodology, Change in organizational management during the software project,
Ineffective communication software project system, Absence of and historical data
(templates).

4. Risk control Techniques

Through reading the existing literature on software risk management approach
and methodology, we listed thirty control factors that are considered important in
reducing and modeling the software risk factors identified in planning software
development; these controls are [12]–[19]:

36 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

C1: Using of requirements scrubbing, C2: Stabilizing requirements and

specifications as early as possible, C3: Assessing cost and scheduling the impact of
each change to requirements and specifications, C4: Develop prototyping and have
the requirements reviewed by the client, C5: Developing and adhering a software
project plan,C6: Implementing and following a communication plan, C7:
Developing contingency plans to cope with staffing problems, C8: Assigning
responsibilities to team members and rotate jobs, C9: Have team-building sessions,
C10: Reviewing and communicating progress to date and setting objectives for the
next phase, C11: Dividing the software project into controllable portions, C12:
Reusable source code and interface methods, C13:Reusable test plans and test cases,
C14: Reusable database and data mining structures, C15: Reusable user documents
early, C16: Implementing/Utilizing automated version control tools, C17:
Implement/ utilize benchmarking and tools of technical analysis, C18: Creating and
analyzing process by simulation and modeling, C19: Provide scenarios methods and
using of the reference checking, C20: Involving management during the entire
software project lifecycle, C21:Including formal and periodic risk assessment,
C22:Utilizing change control board and exercise quality change control practices,
C23: Educating users on the impact of changes during the software project, C24:
Ensuring that quality-factor deliverables and task analysis, C25: Avoiding having
too many new functions on software projects, C26: Incremental development
(deferring changes to later increments), C27: Combining internal evaluations by
external reviews, C28: Maintain proper documentation of each individual's work,
C29: Provide training in the new technology and organize domain knowledge
training, C30: Participating users during the entire software project lifecycle.

5. Empirical Strategy
Data collection was achieved through the use of a structured questionnaire for
estimating the quality of software through determine risks that were common to the
majority of software projects in the analyzed software companies. Top ten software
planning risks and thirty control factors were presented to respondents. The method
of sample selection referred to as distribution personal regular sampling was used.
This procedure is appropriate when members of homogeneous groups (such as
software project managers, IT managers) are difficult to locate. The seventy six
software project managers have participated in this study. The project managers that
participated in this survey are coming from specific mainly software project
manager in software development organizations. However to describe “software
Development Company in Palestine” that have in-house development software and
supplier of software for local or international market, we depended on Palestinian
Information Technology Association (PITA) Members’ webpage at PITA’s website
[PITA 2012 www.pita.ps/], Palestinian investment promotion agency [PIPA 2012
http://www.pipa.gov.ps/] to select top IT manager, software project managers. In
this paper, we Discriminant Analysis techniques to classify software planning risks
in software development process. A Clearly, Discriminant Analysis (DA) is used for
predicting group membership based on a linear combination of independent
variables. Wilks Lambda (^) is used to test the efficiency of Discriminant Analysis
(DA) function. This measure indicates the significant difference between the target
groups. The Discriminant function is expressed as follows.
DF = w1 C1 + w2 C2 + … +wi Ci + a (1)
Relationships between Risks and Control Variables
These tests are used the discriminant analysis techniques to compare the controls
to each of the risk planning software development factors to determine and evaluate

Copyright ⓒ 2015 SERSC 37

International Journal of Advanced Science and Technology
Vol.81 (2015)

if they are effective in mitigating the occurrence of each risk factor. Wilk’s Lambda
is used to test if there is relationship between the discriminant function and the
independent variables (controls).

R1: Risk Of ‘Low Key User Involvement’ Compared To Controls.

Table 1. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square Df Sig.
1 through 2 .599 36.867 6 .000
2 .923 5.777 2 .056

Table 2. Classification Results for Risk 1 using Discriminant Function

Predicted Group
Membership
Cluster Number of Case for R1
Low Medium High Total
Training Count Low
3 0 1 4
Medium
1 4 13 18
High
0 2 52 54
% Low
75.0 .0 25.0 100.0
Medium
5.6 22.2 72.2 100.0
High
.0 3.7 96.3 100.0
Testing Count Low
3 0 1 4
Medium
1 4 13 18
High
0 2 52 54
% Low
75.0 .0 25.0 100.0
Medium
5.6 22.2 72.2 100.0
High
.0 3.7 96.3 100.0

a Cross validation is done only for those cases in the analysis. In cross validation,
each case is classified by the functions derived from all cases other than that case.
b 77.6% of original grouped cases correctly classified.
c 77.6% of cross-validated grouped cases correctly classified.

Table 1 and Table 2 show that discriminant function 1 explains 86.6% of the
variance between the risk groups while discriminant function 2 only accounts fo r
13.4% of the variance. Also it indicates the canonical correlation of the discriminant
functions to the independent variables. Functions 1 and 2 have positive correlation
(r = 0.592, 0.278) hence both function 1 is important for the classification of the
independent variables to risk groups. Eigen value is significant for the discriminant
function 1. The chi-square values (χ2 = 36.867) which is a statistics for measuring
these tests of significance of the Eigen values. However, Wilk’s Lambda is used to
test if there is relationship between the discriminant function and the independent
variables. Associated with each Wilk’s Lambda is a chi -square statistics to measure
the significance of this relationship. If this chi-square statistic corresponding to
Wilk’s Lambda is statistically significant it concluded that a relationship exists
between the discriminant function and the independent variables. The result shows
there is significant relationship between the discriminant function 1 and the
independent variables of c1, c21, c28 related groups. The output for significance
tests and strength of relationship statistics for the discriminant analysis is shown in
Table 1. The coefficients for building the classification models are presented in
equation 2. The output for classification of groups is shown in Table 2. The
classification results allow us to determine how well we can predict group
membership using a classification functions. The results in equation the

38 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

discriminant function equation for predicting the classification of risk 1 with risk
management techniques in software devotement project are given as:

DF1 = 1.635*C1 + 1.336*C21 +0.649*C28 -10.085 (2)

R2: Risk of ‘Unrealistic Schedules and Budgets’ Compared to Controls.

Table 3. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square Df Sig.
1
.784 17.737 2 .000

Table 4. Classification Results for risk 2 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R2 Low Medium High Total

Original Count Low
1 0 4 5
Medium
0 0 20 20
High 1 0 50 51
% Low
20.0 .0 80.0 100.0
Medium
.0 .0 100.0 100.0
High 2.0 .0 98.0 100.0
Cross- Count Low
1 0 4 5
validated(a)
Medium
0 0 20 20
High 5 0 46 51
% Low
20.0 .0 80.0 100.0
Medium
.0 .0 100.0 100.0
High 9.8 .0 90.2 100.0
b 67.1% of original grouped cases correctly classified.
c 61.8% of cross-validated grouped cases correctly classified.

It indicates the canonical correlation of the discriminant functions to the

independent variables. Functions 1 has positive correlation (r = 0.464), hence
function1 is important for the classification of the independent variables to the risk
groups. The chi-square values (χ2 = 17.737) which is a statistics for measuring these
tests of significance of the Eigen values. The result shows there is significant
relationship between the discriminant function 1 and the independent variables of c5
related groups. The discriminant function equation 1 for predicting the classification
of risk 2 with risk management techniques in software devotement project are given
as:

DF1 = 2.366*c5 -6.599 (3)

R3: Risk of ‘Misunderstood / Unrealistic Scope and Objectives (Goals)’ Compared

to Controls.

Table 5. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square df Sig.
1
.892 8.336 2 .015

Copyright ⓒ 2015 SERSC 39

International Journal of Advanced Science and Technology
Vol.81 (2015)

Table 6. Classification Results for risk 3 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R3 Low Medium High Total

Original Count Low 0 1 1 2
Medium 0 8 17 25
High 0 4 45 49
% Low .0 50.0 50.0 100.0
Medium .0 32.0 68.0 100.0
High .0 8.2 91.8 100.0
Cross- Count Low
validated(a) 0 1 1 2
Medium 0 8 17 25
High 0 4 45 49
% Low .0 50.0 50.0 100.0
Medium .0 32.0 68.0 100.0
High .0 8.2 91.8 100.0
b 69.7% of original grouped cases correctly classified.
c 69.7% of cross-validated grouped cases correctly classified.

The canonical correlation of the discriminant functions to the independent variables.

Functions 1 has positive correlation (r = 0.328), hence function1 is important for the
classification of the independent variables to the three risk groups. The chi-square values
(χ2 = 8.336) which is a statistics for measuring these tests of significance of the Eigen
values; hence the test of the Eigen values is significant. The result shows there is
significant relationship between the discriminant function 1 and the independent variables
of c15 related groups. The discriminant function equation for predicting the classification
of risk 3 with risk management techniques in software devotement project are given as:

DF1 = 2.756*c15 -7.796 (4)

R4: Risk of ‘Insufficient/Inappropriate Staffing’ Compared to Controls.

Table 7. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square Df Sig.
1 through 2
.542 43.509 10 .000
2
.938 4.511 4 .341

Table 8. Classification Results for Risk 4 using Discriminant Function

Predicted Group
Membership
Cluster Number of Case for R4 Low Medium High Total
Original Count Low
2 0 7 9
Medium
0 12 12 24
High 1 4 38 43
% Low
22.2 .0 77.8 100.0
Medium
.0 50.0 50.0 100.0
High 2.3 9.3 88.4 100.0
Cross- Count Low
validate 2 0 7 9
d
Medium
1 10 13 24
High 3 4 36 43
% Low
22.2 .0 77.8 100.0
Medium
4.2 41.7 54.2 100.0
High 7.0 9.3 83.7 100.0
b 68.4% of original grouped cases correctly classified.

40 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

c 63.2% of cross-validated grouped cases correctly classified.

Table 7 and Table 8 show that discriminant function 1 explains 91.8% of the
variance between the risk groups. Also it indicates the canonical correlation of the
discriminant functions to the independent variables. Functions 1 has positive
correlation (r = 0.650) higher than the chi-square values (χ2 = 43.509) which is a
statistics for measuring these tests of significance of the Eigen values. The result
shows there is significant relationship between the discriminant function 1 and the
independent variables of c1, c3, c4, c12, c13 related groups. The discriminant
function equation for predicting the classification of risk 4 with risk management
techniques in software devotement project are given as:

DF1 = -1.610*c1 -1.112*c3+1.658*c4+1.153*c12+1.198*c13-3.376 (5)

R5: Risk of ‘Lack of Senior Management Commitment and Technical Leadership’

Compared to Controls.

Table 9. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square df Sig.
1 through 2 .764 19.485 4 .001
2 .933 4.999 1 .025

Table 10. Classification Results for Risk 5 using Discriminant Function

Predicted Group
Membership

Cluster Number of Case for R5 Low Medium High Total

Original Count Low 5 1 0 6
Medium 2 9 18 29
High 3 5 33 41
% Low 83.3 16.7 .0 100.0
Medium 6.9 31.0 62.1 100.0
High 7.3 12.2 80.5 100.0
Cross- Count Low
validate 5 1 0 6
d
Medium 4 7 18 29
High 3 5 33 41
% Low 83.3 16.7 .0 100.0
Medium 13.8 24.1 62.1 100.0
High 7.3 12.2 80.5 100.0
b 61.8% of original grouped cases correctly classified.
c 59.2% of cross-validated grouped cases correctly classified.

Table 9 and Table 10 show that discriminant function 1 explains 75.6% of the
variance between the risk groups while discriminant function 2 only accounts for
24.4% of the variance. Functions 1 and 2 have positive correlation (r = 0.426,
0.258), hence both function1 and 2 are important for the classification of the
independent variables to the risk groups. The chi-square values (χ2 = 19.485, 4.999)
which is a statistics for measuring these tests of significance of the Eigen values,
hence both test of the Eigen values are significant. The result shows there is
significant relationship between the discriminant function 1 and 2 the independent
variables of c6, c7 related groups. From the results in equation the two discriminant
function equations for predicting the classification of risk 5 with risk management
techniques in software devotement project are given as:

Copyright ⓒ 2015 SERSC 41

International Journal of Advanced Science and Technology
Vol.81 (2015)

DF1 = 2.542*c6 -1.263* c7-3.484 (6)

DF2 = -.032*c6+ 2.149* c7-5.877 (7)

R6: Risk of ‘Poor /Inadequate Software Project Planning and Strategic Thinking’
Compared to Controls.

Table 11. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square df Sig.
1
.914 6.593 2 .037

Table 12. Classification Results for Risk 6 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R6 Low Medium High Total

Original Count Low
0 1 2 3
Medium
0 9 12 21
High
0 7 45 52
% Low
.0 33.3 66.7 100.0
Medium
.0 42.9 57.1 100.0
High
.0 13.5 86.5 100.0
Cross- Count Low
0 1 2 3
validated
Medium
0 9 12 21
High
0 7 45 52
% Low
.0 33.3 66.7 100.0
Medium
.0 42.9 57.1 100.0
High
.0 13.5 86.5 100.0

b 71.1% of original grouped cases correctly classified.

c 71.1% of cross-validated grouped cases correctly classified.

Functions 1 has positive correlation (r = 0.294). The chi-square values (χ2 =

6.596) which is a statistics for measuring these tests of significance of the Eigen
values, hence both test of the Eigen values are significant. The result shows there is
significant relationship between the discriminant function 1 and the independent
variables of c1 related groups. From the results in equation the discriminant function
equations for predicting the classification of risk 6 with risk management techniques
in software devotement project are given as:

DF1 = 2.092*c1 -5.754 (8)

R7: Risk of ‘Lack of An Effective Software Project Management Methodology’

Compared to Controls.

Table 13. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square df Sig.
1 through 2 .684 27.332 6 .000
2 .882 9.056 2 .011

42 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

Table 14. Classification Results for Risk 7 using Discriminant Function

Predicted Group
Membership
Cluster Number of Case for R7 Low Medium High Total
Original Count Low 3 0 0 3
Medium 3 11 14 28
High 3 5 37 45
% Low 100.0 .0 .0 100.0
Medium 10.7 39.3 50.0 100.0
High 6.7 11.1 82.2 100.0
Cross- Count Low
validated( 0 1 2 3
a)
Medium 3 11 14 28
High 3 5 37 45
% Low .0 33.3 66.7 100.0
Medium 10.7 39.3 50.0 100.0
High 6.7 11.1 82.2 100.0
b 67.1% of original grouped cases correctly classified.
c 63.2% of cross-validated grouped cases correctly classified.

Table 13 and Table 14 show that discriminant function 1 explains 68.3% of the
variance between the risk groups while discriminant function 2 only accounts for
31.7% of the variance. Functions 1 and 2 have positive correlation (r = 0.473,
0.344). Eigen value is significant for the discriminant function 1 and 2. The chi-
square values (χ2 = 27.332, 9.056) which is a statistics for measuring these tests of
significance of the Eigen values, hence both test of the Eigen values are significant.
The result shows there is significant relationship between the discriminant functions
and the independent variables of c9, c11, and c24 related groups. The discriminant
function for predicting the classification of risk 7 with risk management techniques
in software devotement project are given as:

DF1 = 2.039*c9 -2.439* c11+1.253* c24-2.440 (9)

DF2 = 0.072*c9+ .888* c11+2.116* c24-8.636 (10)

R8: Risk of ‘Change in Organizational Management During The Software Project’

Compared to Controls.

Table 15. Wilks' Lambda

Wilks'
Test of Function(s) Lambda Chi-square df Sig.
1 .877 9.615 2 .008

Copyright ⓒ 2015 SERSC 43

International Journal of Advanced Science and Technology
Vol.81 (2015)

Table 16. Classification Results for Risk 8 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R8 Low Medium High Total
Original Count Low 0 3 3 6
Medium 0 7 18 25
High 0 3 42 45
% Low .0 50.0 50.0 100.0
Medium .0 28.0 72.0 100.0
High .0 6.7 93.3 100.0
Cross- Count Low
0 3 3 6
validated
Medium 2 5 18 25
High 0 3 42 45
% Low .0 50.0 50.0 100.0
Medium 8.0 20.0 72.0 100.0
High .0 6.7 93.3 100.0
b 64.5% of original grouped cases correctly classified.
c 61.8% of cross-validated grouped cases correctly classified.

Functions 1 has positive correlation (r = 0.350), hence function1 is important for

the classification of the independent variables to the risk groups. The chi -square
values (χ2 = 9.615) which is a statistics for measuring these tests of significance of
the Eigen values; hence both test of the Eigen value is significant. There is
significant relationship between the discriminant function 1 and the independent
variables of c17 related groups. The discriminant function equations for predicting
the classification of risk 8 with risk management techniques in software devotement
project are given as:

DF1 = 2.279*C7 -6.386 (11)

R9: Risk of ‘Ineffective Communication Software Project System’ Compared to

Controls.

Table 17. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square Df Sig.

1 through 2 .654 30.581 6 .000
2 .995 .328 2 .849

Table 18. Classification Results for Risk 9 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R9 Low Medium High Total

Original Count Low
4 0 0 4
Medium
0 0 15 15
High 1 0 56 57
% Low
100.0 .0 .0 100.0
Medium
.0 .0 100.0 100.0
High 1.8 .0 98.2 100.0
Cross- Count Low
4 0 0 4
validated
Medium
0 0 15 15
High 3 0 54 57
% Low
100.0 .0 .0 100.0
Medium
.0 .0 100.0 100.0
High 5.3 .0 94.7 100.0
b 78.9% of original grouped cases correctly classified.
c 76.3% of cross-validated grouped cases correctly classified.

44 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

Table 17 and Table 18 show that discriminant function 1 explains 99.1% of the
variance between the risk groups. Also it indicates the canonical correlation of the
discriminant functions to the independent variables. Functions 1 have positive
correlation (r = 0.586), hence both function1 is important for the classification of
the independent variables to the risk groups. The chi-square values (χ2 = 30.581)
which is a statistics for measuring these tests of significance; hence both test of the
Eigen values is significant. The result shows there is significant relationship
between the discriminant function 1 and the independent variables of c1, C6, and C7
related groups. From the results in equation the two discriminant function equations
for predicting the classification of risk 9 with risk management techniques in
software devotement project are given as:

DF1 = 1.620*C1 +1.620* C6-2.067* C7-3.172 (12)

R10: Risk of ‘Absence of a Historical Data (Templates)’ Compared to Controls.

Table 19. Wilks' Lambda

Test of Function(s) Wilks' Lambda Chi-square Df Sig.
1 .818 14.706 2 .001

Table 20. Classification Results for Risk 1 0 using Discriminant Function

Predicted Group Membership

Cluster Number of Case for R10 Low Medium High Total

Original Count Low
0 1 2 3
Medium
0 9 12 21
High 0 4 48 52
% Low
.0 33.3 66.7 100.0
Medium
.0 42.9 57.1 100.0
High .0 7.7 92.3 100.0
Cross- Count Low
0 1 2 3
validated
Medium
0 9 12 21
High
0 4 48 52
% Low
.0 33.3 66.7 100.0
Medium
.0 42.9 57.1 100.0
High .0 7.7 92.3 100.0
b 75.0% of original grouped cases correctly classified.
c 75.0% of cross-validated grouped cases correctly classified.

Functions 1 has positive correlation (r = 0.427), hence Function 1 is used for the
classifying of the controls to the risk groups. The chi-square values (χ2 = 14.706)
which is measuring the significance of the Eigen values; hence both test of the
Eigen values is significant. The result shows there is significant relationship
between the discriminant function 1 and the independent variables of c5 related
groups. From the results in equation the two discriminant function equations for
predicting the classification of risk 10 with risk management techniques in software
devotement project are given as:
DF1 = 2.317*c5 -6.463 (13)

5.3. Identifying Software Planning Risks by using Control Techniques

Table 21 illustrate risks were mitigating by using Control Techniques

Copyright ⓒ 2015 SERSC 45

International Journal of Advanced Science and Technology
Vol.81 (2015)

Table 21. Software Risk Planning Development Factors were Mitigated by

using Control Techniques
module Software Risks Risk Control Techniques

C1: Using of requirements scrubbing, C21:

Including formal and periodic risk
Low key user involvement.
assessment, C28: Maintain proper
1 documentation of each individual's work.
Unrealistic schedules and budgets. C5: Developing and adhering a software
2 project plan.
Misunderstood /Unrealistic scope and C15: Reusable user documents early.
3 objectives (goals).
Insufficient/inappropriate staffing. C1: Using of requirements scrubbing, C3:
Assessing cost and scheduling the impact of
each change to requirements and
specifications, C4: Develop prototyping and
have the requirements reviewed by the
client, C12: Reusable source code and
interface methods, C13: Reusable test plans
4 and test cases.
Lack of senior management commitment C6: Implementing and following a
and technical leadership. communication plan, C7: Developing
contingency plans to cope with staffing
5 problems.
C1: Using of requirements scrubbing.
Poor/inadequate software project
planning and strategic thinking.
6
Lack of an effective software project C9: Have team-building sessions, C11:
management methodology. Dividing the software project into
controllable portions, C24: Ensuring that
7 quality-factor deliverables and task analysis.
Change in organizational management C17: Implement/ utilize benchmarking and
8 during the software project. tools of technical analysis.
Ineffective communication software C1: Using of requirements scrubbing, C6:
project system. Implementing and following a
communication plan, C7: Developing
contingency plans to cope with staffing
9 problems.
C5: Developing and adhering a software
Absence of a historical data (templates).
10 project plan.

6. Conclusions
The paper is the classifying risks of software planning development. These
statistics techniques were used the discriminant analysis techniques, to compare the
controls to each of the risk factors to determine if they are effective in mitigating
the occurrence of each risk factor. However, we classified software planning risks to
high, medium, low by Predicted Group Membership. However, we are referred the
controls in risk management approach were mitigated on software planning risk
software development factors in Table 21. In future work, we can use more
techniques useful to classify and predict software project risk models such as
artificial intelligence techniques.

46 Copyright ⓒ 2015 SERSC

 International Journal of Advanced Science and Technology
Vol.81 (2015)

References
[1] A. Elzamly and B. Hussin, “Estimating Stepwise and Fuzzy Regression Analysis for Modelling
Software Design Project Risks,” Asian J. Math. Comput. Res., vol. 3, no. 4, pp. 234–241, (2015).
[2] A. Elzamly and B. Hussin, “An Enhancement of Framework Software Risk Management Methodology
for Successful Software Development”, Theor. Appl. Inf. Technol., vol. 62, no. 2, (2014), pp. 410-423.
[3] J. Sodhi and P. Sodhi, “IT Project Management Handbook”, Management Concepts, (2001).
[4] K. Khanfar, A. Elzamly, W. Al-Ahmad, E. El-Qawasmeh, K. Alsamara and S. Abuleil, “Managing
Software Project Risks with the Chi-Square Technique”, Int. Manag. Rev., vol. 4, no. 2, (2008), pp. 18-
29.
[5] A. Elzamly and B. Hussin, “Managing Software Project Risks with Proposed Regression Model
Techniques and Effect Size Technique”, Int. Rev. Comput. Softw., vol. 6, no. 2, (2011), pp. 250-263.
[6] A. Elzamly and B. Hussin, “Estimating Quality-Affecting Risks in Software Projects”, Int. Manag. Rev.
Am. Sch. Press, vol. 7, no. 2, (2011), pp. 66-83.
[7] A. Elzamly and B. Hussin, “Managing Software Project Risks ( Implementation Phase ) with Proposed
Stepwise Regression Analysis Techniques”, Int. J. Inf. Technol., vol. 1, no. 4, (2013), pp. 300-312.
[8] A. Elzamly and B. Hussin, “Managing Software Project Risks (Design Phase) with Proposed Fuzzy
Regression Analysis Techniques with Fuzzy Concepts,” Int. Rev. Comput. Softw., vol. 8, no. 11, (2013),
pp. 2601-2613.
[9] A. Elzamly and B. Hussin, “A Comparison of Fuzzy and Stepwise Multiple Regression Analysis
Techniques for Managing Software Project Risks : Implementation Phase”, Int. Manag. Rev., vol. 10,
no. 1, (2014), pp. 43-54.
[10] A. Elzamly, B. Hussin and N. Salleh, “Methodologies and techniques in software risk management
approach for mitigating risks: A review”, Asian J. Math. Comput. Res., vol. 2, no. 4, (2015), pp. 184-
198.
[11] A. Elzamly and B. Hussin, “Modelling and Evaluating Software Project Risks with Quantitative
Analysis Techniques in Planning”, Comput. Inf. Technol., vol. 23, no. 2, (2015), pp. 113-120.
[12] A. Elzamly and B. Hussin, “Managing Software Project Risks (Planning Phase) with Proposed Fuzzy
Regression Analysis Techniques with Fuzzy Concepts”, Int. J. Inf. Comput. Sci., vol. 3, no. 2, (2014),
pp. 31-40.
[13] A. Elzamly and B. Hussin, “Managing Software Project Risks (Analysis Phase) with Proposed Fuzzy
Regression Analysis Modelling Techniques with Fuzzy Concepts”, J. Comput. Inf. Technol., vol. 22, no.
2, (2014), pp. 131–144.
[14] A. Elzamly and B. Hussin, “Evaluation of Quantitative and Mining Techniques for Reducing Software
Maintenance Risks”, Appl. Math. Sci., vol. 8, no. 111, (2014), pp. 5533-5542.
[15] A. Elzamly and B. Hussin, “Modelling and mitigating Software Implementation Project Risks with
Proposed Mining Technique”, Inf. Eng., vol. 3, (2014), pp. 39-48.
[16] A. Elzamly and B. Hussin, “A Comparison of Stepwise And Fuzzy Multiple Regression Analysis
Techniques for Managing Software Project Risks : Analysis Phase”, Comput. Sci., vol. 10, no. 10,
(2014), pp. 1725-1742.
[17] A. Elzamly and B. Hussin, “Mitigating Software Maintenance Project Risks with Stepwise Regression
Analysis Techniques”, Mod. Math. Front., vol. 3, no. 2, (2014), pp. 34-44.
[18] A. Elzamly and B. Hussin, “Classification and identification of risk management techniques for
mitigating risks with factor analysis technique in software risk management”, Rev. Comput. Eng. Res.,
vol. 2, no. 1, (2015), pp. 22–38.
[19] A. Elzamly and B. Hussin, “Identifying and Managing Software Project Risks with Proposed Fuzzy
Regression Analysis Techniques: Maintenance Phase”, 2014 Conference on Management and
Engineering (CME2014), (2014), pp. 1868-1881.

Copyright ⓒ 2015 SERSC 47

International Journal of Advanced Science and Technology
Vol.81 (2015)

48 Copyright ⓒ 2015 SERSC