SOFTWARE ENGINEERING TEAM CLIMATE:
FACTORS AND MEASUREMENT
BY
ARJUMAND BANO SOOMRO
A thesis submitted in fulfilment of the requirement for the degree of Doctor of Philosophy in Information Technology
Kulliyyah of Information and Communication Technology International Islamic University Malaysia
SEPTEMBER 2018
ii
ABSTRACT
Software engineering mainly deals with software development teams, which comprise of a group of two or more people working together cohesively to produce goods or services. Each team possesses a particular working environment or team climate which can be defined by its domain specific requirements. Software engineers are different from the personnel in other domains in terms of their needs and individuality due to the nature of resources required by software engineers based on the project objectives. The team climate is known to be a significant factor in predicting team performance. Therefore, it is necessary to understand the right team climate composition for software development teams. This research has been conducted to identify significant factors that contribute to the team climate of software engineering teams and the development of an instrument to measure it. The research was quantitative, survey-based study where questionnaire has been used as an instrument for the data collection. The instrument was developed based on the significant team climate factors identified from the systematic literature review. The factors were further investigated in terms of their association with different group theories and models. The target population of this study were software professionals affiliated with software industry. The context of this study was not limited to any particular country or region. In finalizing the instrument, the analysis has been done into two phases.
Firstly, Principal Component Analysis (PCA) has been used as the variable reduction technique on data set gathered from 138 software professionals. The results of PCA helped produced the proposed SETC instrument. The resultant instrument contains 31 items which can be classified under four major constructs namely i) Project team characteristics (PTCh) ii) Team cohesion (TC) iii) Job satisfaction (JS) and iv) User representativeness (UR). The correlation among constructs showed significant results that is strong positive relationship and the construct validity showed the nonconvergence among the constructs. The SETC instrument evaluation has been performed by conducting another survey, using Team Performance (TP) as the dependent variable. The impact of SETC constructs on TP was measured by executing PLS-SEM (partial least squares structural equation modelling) algorithm on data set of 91 respondents to test the hypothesis. The results showed that the measurement and structural model estimates fulfilled the convergent validity criteria and the discriminant validity results were found satisfactory. This affirms the constructs are theoretically connected and each construct is truly distinct from other constructs empirically. Hypothesis tests involved the structural relationships among constructs, where the results showed positive relationships between TP variable and the SETC constructs except for the TC. The effect size shows that the three constructs (JS, PTCh and UR) have significant effect on the TP. This research contributes in providing insights of individuals’ working experience as a team, which would be helpful in understanding the team climate dimensions among software professionals. The main contribution of this research was the development of SETC constructs and SETC instrument. The instrument can be used in unfolding the success and failure reasons of software projects in development process and potentially useful for successful team building.
iii
ثحبلا ةصلاخ
ABSTRACT IN ARABIC
جمابرلا ريوطت قرف عم يساسأ لكشب تايمجبرلا ةسدنه لماعتت تيلا
رثكأ وأ ينصخش نم نوكتت
لمعلل خانم وأ ةنيعم لمع ةئيب قيرف لك كلتيم .تامدلخا وأ علسلا جاتنلإ محلاتم لكشب ًاعم نولمعي في ينفظولما نع تايمجبرلا وسدنهم فلتيخ .لالمجبا ةصالخا تاجايتحلاا للاخ نم هفيرعت نكيم يعاملجا متهايصخشو متهاجايتحا ثيح نم ىرخلأا تلاالمجا ذو
كل جاتيح تيلا دراولما ةعيبط ببسب ىلع ًاءانب ا نهو
.عورشلما فادهأ نم ،كلذل .قيرفلا ءادبآ ؤبنتلا في مهم لماع يعاملجا لمعلا خانم نأ فورعلما نم
ديدحتل ثحبلا اذه ءارجإ َّتم .جمابرلا ريوطت قرفل يعاملجا لمعلا خانلم ةبسانلما ةبيكترلا مهف يرورضلا تيلا ةمهلما لماوعلا .اهسايقل ةادآ ريوطتو تايمجبرلا ةسدنه قرفل يعاملجا لمعلا خانم في مهاست
ناكو
عملج ةادآك نايبتسلاا مدختسا ثيح ، ةيئاصقتسا ةسارد لىإ دنتست ةيمك ةسارد نع ةرابع ثحبلا جارم للاخ نم اهديدتح َّتم تيلا ةمالها ةيخانلما لماوعلا لىإ اًدانتسا ةادلآا ريوطت َّتم .تناايبلا ةيجهنم ةع
اهطابترا ثيح نم لماوعلا في قيقحتلا نم ديزلما ءارجإ َّتم امك .لالمجا اذه في ةقباسلا تاساردلل لىإ ينبستنلما تايمجبرلا فيترمح نم نوكتت ةساردلا هذه نم ةفدهتسلما ةنيعلا تناك .ةفلتمخ جذانمو تيارظنب دلب يأ ىلع ةساردلا هذه قايس رصتقي لم .تايمجبرلا ةعانص .ةنيعم ةقطنم وأ
دنع تاسمللا عضو
تناوكملل يساسلأا ليلحتلا مادختسا َّتم ،ًلاوأ .ينتلحرم ىلع ليلحتلا ءارجإ َّتم ،ةادلآا ىلع ةيرخلأا ( نم اهعجم َّتم تيلا تناايبلا ىلع ةيرغتلما لازتخلاا ةينقتك ) PCA
138 جئاتن .تايمجبرلا فيترمح نم
PCA ةادآ جاتنإ في تدعاس
SETC قلما
ىلع ةتجانلا ةادلآا يوتتح .ةحتر اهفينصت نكيم اًفنص 31
و ةيسيئر تاينب عبرأ تتح ( عورشلما قيرف صئاصخ ًلاوأ :يه
قيرفلا كساتم ًاينثا ،) PTCh (
TC ،)
( يفيظولا اضرلا ًاثلثا JS
)،
( مدختسلما ليثتم ًاعبار و UR
.) ةمهم جئاتن تابيكترلا ينب طابترلاا رهظأ
يجإ ةقلاع تدجو تيلاو ةادآ مييقت ًّتم .تابيكترلا ينب قفاوتلا مدع ءانبلا ةيحلاص تتبثأو مهنيب ةيوق ةيبا
SETC ( قيرفلا ءادأ مادختسبا ىرخآ ةيئاصقتسا ةسارد ءارجإ قيرط نع
َّتم .عبتا يرغتمك ) TP سايق
بيكرت يرثتأ SETC
ىلع TP ةيمزراوخ ذيفنت للاخ نم PLS-SEM
ىرغصلا تاعبرلما جذانم(
ل ةيئزلجا نم تناايب ةعوممج ىلع )عبرمل 91
.ةيضرفلا رابتخلا بيجتسم تاريدقت نأ جئاتنلا ترهظأو
و سايقلا زيامتلا ةحص جئاتن نأو ةبراقتلما ةيحلاصلا يرياعم تفوتسا يلكيلها جذومنلا
تناك اذه .ةيضرم
امك .ةيبيرجتلا ىرخلأا نىبلا نع زيمتم ميمصت لك نأو ًيارظن ةطبترم ةينبلا نأ دكؤي تارابتخلاا تنمضت
يرغتم ينب ةيبايجإ تاقلاع دوجو جئاتنلا ترهظأ ثيح ،تابيكترلا ينب ةيلكيلها تاقلاعلا ةيضرف TP
ـلا تابيكرتو SETC
ءانثتسبا TC
. ( ةثلاثلا تابيكترلا نأ يرثأتلا مجح حضوي JS
و
PTCh
و
iv
UR ىلع يربك يرثتأ اله ) TP
معلا ةبرجتل ىؤر يمدقت في ثحبلا اذه مهاسي . ،قيرفك دارفلأا ىدل ل
ةيسيئرلا ةهماسلما تناك .تايمجبرلا في ينصصختلما ينب لمعلا قيرف داعبأ مهف في اًديفم نوكيس ام وهو ةادآو تابيكرت ريوطت وه ثحبلا اذله SETC
حاجنلا بابسأ نع فشكلا في ةادلآا مادختسا نكيم .
نأ نكيم تيلاو ريوطتلا ةيلمع في تايمجبرلا عيراشلم لشفلاو حجنا قيرف ءانبل ةديفم نوكت
.
v
APPROVAL PAGE
The thesis of Arjumand Bano Soomro has been approved by the following:
_____________________________
Norsaremah Salleh Supervisor
_____________________________
Azlin Nordin Co-Supervisor
_____________________________
Norbik Bashah Idris Internal Examiner
_____________________________
Uffe Kock Wiil External Examiner
_____________________________
Syed Ahmed Aljunid External Examiner
_____________________________
Saim Kayadibi Chairman
vi
DECLARATION
I hereby declare that this thesis is the result of my own investigations, except where otherwise stated. I also declare that it has not been previously or concurrently submitted as a whole for any other degrees at IIUM or other institutions.
Arjumand Bano Soomro
Signature ... Date ...
vii
COPYRIGHT
INTERNATIONAL ISLAMIC UNIVERSITY MALAYSIA
DECLARATION OF COPYRIGHT AND AFFIRMATION OF FAIR USE OF UNPUBLISHED RESEARCH
SOFTWARE ENGINEERING TEAM CLIMATE: FACTORS AND MEASUREMENT
I declare that the copyright holders of this thesis are jointly owned by the Student and IIUM.
Copyright © 2018 Arjumand Bano Soomro and International Islamic University Malaysia. All rights reserved.
No part of this unpublished research may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without prior written permission of the copyright holder except as provided below
1. Any material contained in or derived from this unpublished research may only be used by others in their writing with due acknowledgement.
2. IIUM or its library will have the right to make and transmit copies (print or electronic) for institutional and academic purposes.
3. The IIUM library will have the right to make, store in a retrieved system and supply copies of this unpublished research if requested by other universities and research libraries.
By signing this form, I acknowledged that I have read and understand the IIUM Intellectual Property Right and Commercialization policy.
Affirmed by Arjumand Bano Soomro
……..……….. ………..
Signature Date
viii
Dedication
To my beloved family and friends
ix
ACKNOWLEDGEMENTS
First of all, I am gratified with the core of my heart to Almighty Allah who made it possible to complete this thesis.
I am also grateful to my supportive supervisor Assoc. Prof. Dr. Norsaremah Salleh and co-supervisor Assist. Prof. Dr. Azlin Nordin who have continuously encouraged me throughout my research. I am especially thankful to my main supervisor Dr. Norsaremah who guided me with great patience and keep persuaded me during my research. She made me learn many things as I enrolled in PhD program at IIUM with weak research background. Thank you very much Dr. Norsaremah.
I must acknowledge my work to my dear husband Khurram Ahmed Qureshi, son Moazzam Jan Qureshi, sister in law Mehrunnisa Bozdar and my niece Amber Sharif Bozdar. Without their support, concern and love, it was impossible for me to complete my PhD studies. I am among those who strongly believe in du’a (pray) of parents and I believe their du’a enlightened the journey of my research, Ammi Farhat Jehan Begum Bashi and Abbu Mohammad Shafi Soomro.
I highly indebted to the members of my research unit at IIUM, Sr. Norzariyah, Sr. Marwa, Br. Imam and Dr. Younis, their presence made me to share and exchange the thought provoking ideas for the solution of my research problem. Dr. Younis, Br.
Imam and Sr. Zariyah thanks a lot for your support during my thesis submission as it was far difficult for me to do it without your support.
Thank you so much.
x
TABLE OF CONTENTS
Abstract ... ii
Abstract in Arabic ... iii
Approval page ... v
Declaration ... vi
Copyright ... vii
Dedication ... viii
Acknowledgements ... ix
List of Tables ... xiii
List of Figures ... xvii
List of Acronyms ... xix
CHAPTER ONE: INTRODUCTION ... 1
1.1 Background of the Study ... 1
1.2 Statement of Problem ... 6
1.3 Research Questions ... 8
1.4 Research Objectives ... 8
1.5 Significance of the Study ... 8
1.6 Contributions ... 10
1.7 Limitations of the Study ... 12
1.8 Thesis Organisation ... 13
CHAPTER TWO: LITERATURE REVIEW ... 15
2.1 Introduction ... 15
2.2 Review Questions ... 16
2.3 Identifying the Relevant Literature ... 17
2.3.1 Search strategy for primary studies ... 17
2.3.2 Primary Search Process ... 19
2.3.3 Strategy for Secondary Search Process ... 20
2.3.4 Selection Criteria ... 21
2.4 Quality Checklist and Procedures ... 22
2.4.1 Study Quality Checklist ... 22
2.5 Data Extraction Strategy ... 23
2.5.1 Data Extraction Process ... 24
2.6 Data Synthesis and Findings ... 24
2.6.1 Primary Studies Overview ... 25
2.7 Answering Review Question No. 1 ... 27
2.8 Answering Review Question No. 2 ... 30
2.9 Answering Review Question No. 3 ... 33
2.10 Answering Review Question No. 4 ... 35
2.11 Discussion On Slr’s Findings ... 38
2.11.1 Implications ... 39
2.11.2 Threats to Validity of the SLR’s Findings ... 42
2.11.3 Recommendations Based on SLR’s Findings ... 43
2.12 Recent Published Studies Not Included In the SLR ... 44
2.13 Groups... 46
xi
2.14 Group Developement ... 48
2.15 Group Dynamics ... 49
2.16 Group Characteristics ... 49
2.17 Group Interaction ... 50
2.18 Group Structure ... 50
2.19 Group Interdependence ... 51
2.20 Group Goal ... 51
2.21 Group Cohesiveness ... 53
2.22 Software Project Teams ... 53
2.23 Summary ... 54
CHAPTER THREE: RESEARCH METHODOLOGY ... 57
3.1 Introduction... 57
3.2 Research Philosophy ... 57
3.3 Research Design ... 59
3.3.1 Research Process ... 60
3.4 Population and Sample ... 62
3.4.1 Population ... 62
3.4.2 Sample ... 63
3.4.3 Sampling Frame ... 64
3.5 Instrumentation ... 64
3.5.1 Item creation... 64
3.5.2 Measurement and measurement scale ... 65
3.6 SETC constructs ... 66
3.6.1 Project Team Characteristics... 69
3.6.2 Team Cohesion... 72
3.6.3 Job Satisfaction ... 74
3.6.4 User Representativeness ... 76
3.7 Instrument Pre-testing ... 77
3.7.1 Instrument Revision ... 79
3.8 Piloting of instrument ... 80
3.8.1 Method ... 82
3.8.2 Response rate ... 82
3.8.3 Sample size... 82
3.8.4 Subjects ... 82
3.8.5 Reliability of Scale ... 83
3.9 Data Collection Procedures ... 84
3.10 Data Analysis Procedures ... 86
3.11 Summary ... 86
CHAPTER FOUR: DATA ANALYSIS AND RESULTS ... 88
4.1 Introduction... 88
4.2 Preliminary data analysis ... 88
4.2.1 Response rate ... 88
4.2.2 Missing data ... 89
4.2.3 Normality ... 91
4.2.4 Normality check at construct level ... 96
4.2.5 Outliers ... 97
4.2.6 Outliers’ treatment ... 100
xii
4.3 Demographics ... 102
4.4 Multivariate Analysis (MVA)... 104
4.4.1 PCA based on eigenvalues ... 105
4.4.2 PCA with a fixed number of factors ... 109
4.4.3 Naming Latent Components ... 115
4.4.4 Normality checks for Latent components ... 121
4.4.5 Bivariate correlation among components... 122
4.4.6 Construct Validity ... 123
4.5 Summary ... 125
CHAPTER FIVE: SETC Evaluation ... 127
5.1 Introduction... 127
5.2 Hypothetical model of SETC and Team Performance (TP) ... 127
5.3 Evaluation of Hypothetical model ... 130
5.4 Outliers Examination ... 131
5.5 Outlier treatment ... 132
5.6 Demographics ... 133
5.7 Path model ... 136
5.8 Measurement model estimates ... 137
5.9 Structural Model estimates ... 143
5.9.1 Computations ... 143
5.9.2 Computation of Effect size (ƒ2)... 146
5.9.3 Computation of Predictive Relevance (Q2) ... 147
5.10 Summary ... 148
CHAPTER SIX: DISCUSSION ... 150
6.1 Introduction... 150
6.2 Overview of the Research Findings ... 150
6.2.1 SETC Constructs and their Measures ... 153
6.3 Threats to validity ... 154
6.4 Implications ... 156
CHAPTER SEVEN: CONCLUSION ... 159
7.1 Introduction... 159
7.2 Systematic Literature Review ... 159
7.3 Research Questions and Hypotheses ... 161
7.4 Results from the first Survey (SETC DEVELOPMENT) ... 163
7.5 Results from the second survey (SETC EVALUATION) ... 165
7.6 Future Work ... 167
7.7 Final remarks ... 168
REFERENCES ... 170
APPENDIX A: IREC APPROVAL LETTER (PDF FORMAT) ... 184
APPENDIX B: DATA EXTRACTION FORM FIELDS ... 186
APPENDIX C: COVER LETTER OF SETC INSTRUMENT ... 187
APPENDIX D: LIST OF STUDIES INCLUDED IN THE SLR ... 188
APPENDIX E: SETC QUESTIONNAIRE ... 191
APPENDIX F: LIST OF PUBLICATIONS ... 196
xiii
LIST OF TABLES
Table 2.1 PICOC for RQs 16
Table 2.2 Major search terms 17
Table 2.3 The alternate search terms 18
Table 2.4 Construction of search strings with OR 18
Table 2.5 Results by Sub-strings 20
Table 2.6 Literature Resources 20
Table 2.7 Selected Primary Studies 21
Table 2.8 Studies distribution as per Quality Score Range 26
Table 2.9 List of Relevant Studies for each RQ 27
Table 2.10 Team Climate Compositions 29
Table 2.11 Personality Model/Instruments Used in Primary Studies 30 Table 2.12 Significant Personality Factors Affecting Software Team
Performance 32
Table 2.13 Significant Team Climate Factors Affecting Team
Performance 35
Table 2.14 Measures used for team performance 36
Table 2.15 Recent studies covering team climate and team performance 45
Table 2.16 Work group models 47
Table 2.17 Tuckman's Model Development Stages 48
Table 2.18 McGrath's Circumplex Model of Group Tasks 52
Table 3.1 A classification of Research Methodologies by Galliers (1991) 59
Table 3.2 Significant Team Climate Factors 67
Table 3.3 Work Group Models 68
Table 3.4 Themes and characteristics related to work group effectiveness 69 Table 3.5 Items used to measure project team member composition 71
xiv
Table 3.6 Items to measure Project Task 71
Table 3.7 Items to measure Team Cohesion 72
Table 3.8 Items to measure Job satisfaction 75
Table 3.9 Qualitative Findings 78
Table 3.10 Experience in software design 83
Table 3.11 Project size vise distribution 83
Table 3.12 Item total statistics 84
Table 3.13 List of Variables 86
Table 4.1 Missing value analysis at Item level 90
Table 4.2 Missing value analysis at Construct level 90
Table 4.3 Little’s MCAR test 91
Table 4.4 Item wise statistics for PTCh construct 93
Table 4.5 Item wise statistics for TC construct 94
Table 4.6 Item wise statistics for JS construct 95
Table 4.7 Item wise statistics for UR construct 95
Table 4.8 Descriptive statistics 96
Table 4.9 z-values 96
Table 4.10 Tests of Normality 97
Table 4.11 Lower and Upper bounds 98
Table 4.12 Project Team Characteristics Extreme Values 98
Table 4.13 Team Cohesion 99
Table 4.14 Job Satisfaction 99
Table 4.15 User Representativeness 99
Table 4.16 Respondents’ Work Experience 103
Table 4.17 Country wise distribution 103
Table 4.18 Project size wise distribution 104
Table 4.19 Pre-Factor Analysis Tests 106
xv
Table 4.20 Communalities 106
Table 4.21 Factors extracted by Eigenvalues 107
Table 4.22 Rotated Component Matrix 108
Table 4.23 Communalities 109
Table 4.24 Total variance explained by Four (4) factors 110
Table 4.25 Rotated Component Matrix for Four components 112
Table 4.26 KMO and BTS (final rotation) 113
Table 4.27 Communalities for final 30 items 113
Table 4.28 Extracted four components for 31 items 114
Table 4.29 Final rotated component matrix for 31 items 115
Table 4.30 Reliability statistics (JS component) 117
Table 4.31 Item-total statistics 117
Table 4.32 Reliability statistics (PTCh component) 118
Table 4.33 Item-total statistics 118
Table 4.34 Reliability statistics (TC component) 119
Table 4.35 Item-total statistics 119
Table 4.36 Reliability statistics (UR component) 120
Table 4.37 Item-total statistics 120
Table 4.38 Normality Test 121
Table 4.39 Descriptive statistics of Latent components 122
Table 4.40 z-values for latent components 122
Table 4.41 Spearman’s rho correlation among latent components 123
Table 4.42 AVE for Components in Varimax rotation 123
Table 4.43 Promax Loadings 124
Table 4.44 AVE for Components in Promax rotation 125
Table 5.1 Univariate analysis result for outliers 131
Table 5.2 Respondents working experience 133
xvi
Table 5.3 Country of origin wise distribution 134
Table 5.4 Company location 134
Table 5.5 Project size wise distribution 135
Table 5.6 Construct validity and reliability 139
Table 5.7 Overall SETC and TP model outer loadings 140
Table 5.8 Cross loadings 142
Table 5.9 Discriminant Validity 143
Table 5.10 Testing results of Structural Model Path Coefficients 145
Table 6.1 Removed SETC items 154
xvii
LIST OF FIGURES
Figure 2.1 Formulation of search string 18
Figure 2.2 Sub-strings 19
Figure 2.3 Proportions of Studies according to Publication Type 25
Figure 2.4 Number of Primary Studies by Year 26
Figure 2.5 The Search String 44
Figure 3.1 Alternative Philosophical Paradigm Names Source: (Hussey &
Hussey, 1997) 58
Figure 3.2 Research Questions and Objectives 61
Figure 3.3 Research Flow Diagram 62
Figure 3.4 Proposed SETC constructs 68
Figure 4.1 Formulas for outer bound and lower bound 97
Figure 4.2 Boxplot for Project team characteristic 100
Figure 4.3 Boxplot for Job satisfaction 100
Figure 4.4 Boxplot for User representativeness 101
Figure 4.5 Job satisfaction items 117
Figure 4.6 Project team characteristics items 118
Figure 4.7 Team cohesion items 120
Figure 4.8 User representativeness items 121
Figure 5.1 Team Performance items 128
Figure 5.2 Hypothetical model for SETC constructs and TP 129
Figure 5.3 Boxplots 132
Figure 5.4 Gender wise distribution 135
Figure 5.5 Team climate awareness 136
Figure 5.6 Path Model for SETC and TP 136
Figure 5.7 SETC and TP Model with initial estimates 138
xviii
Figure 5.8 SETC and TP model (Iteration 3) 141
Figure 5.9 SETC and TP structural model 144
Figure 5.10 Cross-Validated redundancy value 148
Figure 7.1 Hypothesis 165
xix
LIST OF ACRONYMS
IS Information System
SLR Systematic Literature Review JS Job Satisfaction
PTCh Project Team Characteristic TC Team Cohesion
UR User Representativeness TP Team Performance
SETC Software Engineering Team Climate MVA Multivariate Analysis
FA Factor Analysis
PCA Principal Component Analysis
SPSS Statistical Package for the Social Sciences
PLS-SEM Partial least squares structural equation modeling
1
CHAPTER ONE INTRODUCTION
1.1 BACKGROUND OF THE STUDY
Software development is a highly collaborative and human-centric task because it mainly relies on software project teams (P.-C. Chen, Chern, & Chen, 2012; Jones &
Harrison, 1996; Pieterse, Kourie, & Sonnekus, 2006). Since 1960, team, teamwork and team performance have been the topics of study in the domain of software engineering (Shneiderman, 1980; Weinberg, 1998). Jones and Harrison (1960) defined a team or a group as “an arrangement between two or more people to work together so as to produce an identifiable good or service in such a manner that the group members are highly interdependent”. Teams can be defined particularly by three specific key elements: i) common goals, ii) complementary skills and responsibilities, and iii) regular and frequent meetings among team members (Katzenbach & Smith, 2001).
Software groups or project teams are considerably different from work groups in other domains. The main reasons they differ are: i) the nature of resources required from a software engineer, ii) their objectives, and iii) the environment (Dafoulas &
Macaulay, 2001). Faraj and Sproull (2000) referred software project teams as knowledge teams which require expertise as a major resource. According to them, every new project begins with the collection of resources required for the team and then the people with the relevant expertise are gathered. They stated that software development is a knowledge work and its major resource is expertise, specialized skills or knowledge (Faraj & Sproull, 2000).
2
Chen et al. (2012) further extended the concept of software project teams by discussing team members’ composition and the project task. According to them, the team members’ compositions include their educational background, age, gender and flexibility in doing the assigned job. The other characteristic of a software project team is project task, which must be easy enough to be handled by all team members equally. This is only possible when the team members possess a variety of skills. The project task should be handled by all members with foremost importance and finished by them on time; this requires team members’ coordination, cooperation and collaboration.
It has been reported that different groups in an organisation experience their own climate, which can cause success or failure of the work team. It has been revealed that an individual’s behaviour is always under the influence of the team climate (Guzzo & Shea, 1992; Sharma & Gupta, 2012). According to (Anderson & West, 1998) the term team climate can be been defined as shared perceptions and ideas among the individuals working together as a team. Many software engineering research studies adopted this generic concept of team climate and use the Team Climate Inventory (TCI) (Anderson & West, 1998) to measure the team climate (e.g.
Acuña et al., 2015; Ganesh, 2013; Ji & Wang, 2012; Loo, 2003; Stevens et al., 1998;
Sumner & Molka-Danielsen, 2010).
Climates exist at the organisational level as well as at group level (Anderson &
West, 1998), the organisational climate has been perceived by researchers at a broader level that is a combination of attitude, feeling and behaviour of every member of that organisation (Reichers & Schneider, 1990). An organisational climate can be considered as plural by their nature, as it may include many particular reference climates (e.g. climate for safety, climate for motivation etc.) (Schneider & Reichers,
3
1983). In some cases, research studies associated organisational climate with software development procedures instead of team climate (e.g. Sharma & Gupta, 2012).
There are limited researches, which defined team climate factors in the context of software engineering. Ji & Wang (2012) for instance defined team climate by elaborating two factors: i) uncertainty, that team climate can be perceived as
“efficient”, “negative”, or “neutral” etc. and ii) performance, i.e. a good team climate may have a good impact on team performance. In a study conducted by Zhang et al.
(2015), the team climate has been defined in the context of Chinese organisations.
According to them, every team climate possesses its own set of norms, which are under the influence of the organisation’s policies and individual’s behaviour or work practices. In their research, they aligned Chinese cultural connection by including Confucian values and Confucian work dynamism (CWD) in team climate definition and made it more specific for Chinese organisations.
Many research studies till now have highlighted the importance of team climate in relationship with team performance and productivity (Açıkgöz, Günsel, Bayyurt, & Kuzey, 2013; Acuña et al., 2015; Beranek et al., 2005; Ji & Wang, 2012;
Moe et al., 2010; Sharma & Gupta, 2012; Lindsjørn et al., 2016). These studies carried the concept of team climate in various ways without addressing it with any agreed upon team climate factors, particularly for software engineering teams. For example, in a study by Lindsjørn et al. (2016), the authors reported that communication, coordination, balance of member contribution, mutual support, effort and cohesion as the construct for team work quality (TWQ). In addition to that, most of the researches took place in academic settings with undergraduate students instead of in industry by taking software professionals as the subject. Before relating the concept of team climate with any other variables like personality or team performance, it is required to
4
identify the right factors for the software engineering team climate. This can lead to the true understanding of issues related to the software development teams. Therefore, to fill this gap, it is important to identify the software engineering team climate composition. Hence, an extensive literature review has been conducted in a systematic way to identify the significant factors to form the software engineering team climate (SETC) constructs. Acuña et al. (2008) reported that due to the social nature of software development teams, the team climate has a significant impact on team performance. In this regard, these SETC constructs are further investigated in relation to the team performance (TP) construct.
In Information System (IS) projects, the team performance has been defined as the degree to which a team meets or exceeds its goals (Jones & Harrison, 1996). Team performance in IS literature has been measured in various ways at both the individual level and team level and in both behavioural and technical aspects (Guinan et al., 1998; Henderson & Lee, 1992). According to Guinan et al. (1998), the behavioural aspect focuses on team members’ characteristics and capabilities, whereas the technical aspect covers the use of structured methods, production technology and coordination technology. The technical aspect can be covered by three determining factors: i) experience spread (previous experience with the task), ii) team skill (breadth of abilities of team members provided to the team), and iii) managerial involvement (to influence the team members and team effectiveness).
In a review study by Sudhakar et al. (2011), the team performance measurement is reported in terms of subjective and objective measures. The subjective measures cover the technical aspects e.g. function points, object points, use case points, KLOC (Kilo Lines of Code) and defect rates. The perceptual or objective measures include the quality of software, teamwork ability, satisfaction with the end-
5
product and team efficiency. Henderson and Lee (1992) suggested perceptual measures to assess team performance and for this purpose, they collected data from the stakeholders other than the project team. The measures are listed below:
• Efficiency
o The efficiency of team operations o The amount of work the team produces o The team's adherence to schedules o The team's adherence to budget
• Effectiveness
o The quality of work the team produces
o Effectiveness of the team's interactions with people outside of the team o The team's ability to meet the goals of the project
• Time
o The team could have done the work faster with the same level of quality
o The team met the goals as quickly as possible
These measures were adopted in many IS studies (P.-C. Chen et al., 2012;
Jones & Harrison, 1996; Sawyer, 2001). Jones and Harrison (1996) reported that self- assessment of team members in human behavioural sciences are widely adopted.
According to them, the IS project team members’ perceptions of their team is a great predictor of performance. This was affirmed by Guinan et al. 1998, they presented the comparison of behavioural and technical team performance measures by collecting data from both stakeholders and team members. They developed a four-item seven- point scale by following Henderson & Lee (1992). In this research, the work of