What are the risk factors that have a high impact on the performance objective 'Time' in DAD projects. What are the risk factors that have a high impact on the performance objective, 'Cost' in DAD projects. What are the risk factors that have a high impact on the performance objective, 'Quality' in DAD projects.
This helped us to identify the risk factors that greatly affect the "time" of the project. The average of the mean rankings for the outcome of the project 'Time' for these risk factors (Table 1) was 1.75. The average ranking for the outcome of the "Quality" project for the selected 7 risk factors was 1.68.
Survey Part II Data Analysis
Lack of trust between the client and the offshore team (it is difficult to meet customer expectations in short iterations). This led to a list of 7 risk factors which were included in the questionnaire, part-II. All the risk factors that affect 'Time'. and 'Quality' have been marked in bold in Table 1 and Table 2 respectively.
Ranking of risk factors based on perceived impact on project time Average ranking R1 Inadequate prioritization of requirements. We had a list of risk factors (8) that were perceived to affect 'Time' the most and another list of 7 risk factors that affect 'Quality' the most in DAD projects. Based on the extent to which the risk factors affect the project's outcome, respondents were asked to rank the risk factors.
The Kendall test was then applied to the ranking data of risk factors for both project deliverables, time and quality, separately. The detailed results of applying the Kendall test to risk factors affecting project duration are given in Table 3 and the results of applying the same test to risk factors affecting project quality are given in Table 4. The average ranking obtained for project outcome risk factors, 'Time' and 'Quality' were statistically significant (p-value <= 0.000), indicating that the findings can be generalized.
The risk factors in Table 3 and Table 4 are listed according to the rankings given for the risks. Risk factors at the top of the list are those that have a relatively higher impact on the project's outcome compared to those lower on the list.
6 Research Findings and Discussion
- Impact of Risk Factors on Project Time (schedule) of the project
- Impact of Risk Factors on Project Cost
- Impact of Risk Factors on Project Quality
- Review of the impact on Project Goals
Further, Agile teams welcome changes in requirements, which causes changes in system architecture (Dingsøyr, Moe, Faegri & Seim, 2018), which further causes project delays (Begel & Nagappan, 2007). It is also evident in our research findings that "Poor coordination between different sites" causes difficulties in integration activities due to interdependencies leading to schedule delays. A high level of coordination between the solution development team and vendors is required, which if not considered, leads to delays in the project outcome ('Third Party Dependence on Solution Development leading to coordination problems between them').
If there is a delay in the project schedule or it needs to be reworked to improve quality, it will lead to increased project cost (Reichelt & Lyneis, 1999). The 'constraints' layer has been linked to the 'Cost' project goal placed in the goal layer in the GSRM model. Here we mean the focus on the project goal 'Quality' in the Objective layer, as in the GSRM framework and especially the risk factors that act as an obstacle to the achievement of the project goal 'Quality'.
As presented in Table 1, the risk factors numbered as, R25-R42, R43, R44, R45, had a major impact on the project objective 'Quality' and were presented in the 'Obstacle Layer' in the GSRM framework. The research findings of this study show that 'Lack of communication between the Team and the Client' and poor cooperation between the team members significantly affect the project quality. There were other risks such as lack of documentation, technical faults in the code and difficulties in performing pair programming due to team dispersion affecting the project quality.
Poor requirements management and design/architecture changes are primarily responsible for DAD project delays. Research findings show that this communication between the team and with the client has a significant impact on the quality of results in DAD projects.
7 Limitations and threats to validity
This risk can be overcome by good coordination among developers, good modular system design along with good documentation (Herbsleb & Grinter, 1999). One of the cultural differences, such as "strange language" among team members, hinders effective communication, which reduces the quality of the result. Other risk factors to consider are: "No common definition of Done among distributed teams", "Inadequate communication of end-user requirements, "Lack of uniformity in team capabilities" and "Poor trust between team members".
The research results show that different types of risks affect different project goals in DAD projects and this understanding of risks would help implement effective risk management. We found that risks in DAD projects have a major impact on project planning and quality, which are the primary goals considered by Agile teams. Surprisingly, the risks revealed by our findings impact these two most critical project objectives of DAD projects.
To some extent, cultural differences in DAD projects also affect quality and should be taken into account during project implementation. To ensure construct validity, we provided a brief explanation of each risk factor and project goal in the questionnaires, which helped clarify the meaning of each construct. Because we obtained statistically significant results, this indicates that practitioners' understanding of the survey was correct.
We could not get answers from some respondents who had previously participated in the first part of the survey, as they were not willing to participate in the survey several times. The quantitative methods (Westen & Rosenthal, 2003) to establish construct validity could not be used due to the short time and limited number of respondents likely to participate in the study.
8 Conclusion
We found that as the number of risk factors in Part II of the questionnaire was reduced and the data was collected both via email and through an online platform, more responses could be received. The data was collected primarily in Australia, in the city where the principal investigator is based, as well as in other cities around the country. We attempted to improve external validity by collecting data from other countries, including India, Japan, the UK and the US.
Due to time and budget constraints, we were unable to utilize external research services for further data collection. Both the developers and research communities benefit from the findings of this study.
Exploring very large-scale software development: an exploratory case study and research agenda for adapting the agile method. Agile Software Development with Distributed Teams: Agility, Distribution and Trust in Agile Conference (AGILE), 2013, pp Empirical Studies of Agile Software Development: A Systematic Review Information and Software Technology, p. 2009) Lean and agile manufacturing: external and internal drivers and performance outcome.
Agile practices bridge the gap in global software development, information systems management Hossain, E., Babar, M.A., Paik, H.Y., & Verner, J. Processes for using Scrum in global software development: A conceptual framework. An empirical study on the implementation and evaluation of a goal-driven software development risk management model. Review Article: Challenges and Improvements in Distributed Software Development: A Systematic Review, Hindawi Publishing Corporation, Advances in Software Engineering, Volume 2009, Article ID 710971.
A New Perspective on GDSD Risk Management: Agile Risk Management, 5th IEEE International Conference on Global Software Engineering (ICGSE). 1995). The effect of coordination and uncertainty on software project performance: Residual performance risk as an intervening variable. Factors related to software development flexibility of successful projects. 2002) Factors Affecting Technical Communication in Distributed Product Development: An Empirical Study in the Telecommunications Industry IEEE Transactions on Engineering Management, (1), 45.
Appendix
Ranking of Project Goals for the enlisted Risk Factors
6 Rework caused by architectural changes (frequent changes in requirements) 7 Inconsistency in design standards of distributed teams. 8 Technical debt in the code. deferred work not directly related to functionality, but required for the overall quality of the system) 9 Pair programming issues (distributed developers). 12 Cross-functional teams insufficient for testing in large projects. the 'whole team concept' may not be sufficient for complex tests) 13 Test data management where tests are pulled from multiple databases. small iterations in Agile and lack of test database automation) 14 Losing time for end-to-end at scale Interdependent transactional testing cycle across distributed teams (applicable to transactional software applications).
16 Different software development practices and standards followed by multiple teams. cultural diversity and lack of stakeholder support) 17 No common Definition of Done between the distributed teams. 21 Using Component Teams instead of Feature Teams (e.g. GUI team or component X team) cannot provide end-to-end functionality). 24 Greater interdependence between the teams. system complexity and lack of communication between distributed teams).
25 Lack of uniformity in the capabilities of multisite teams (loss of productivity) 26 Unavailability of Business Analyst (BA). single BA is assigned to . multiple teams). 30 Lack of documentation since Agile downplays documentation (DAD projects require minimum documentation to compensate for informal team interaction). 38 Unavailability of the product owner (non-collocation or lack of time) 39 Poor coordination between multiple suppliers involved in the solution. especially if they compete with each other and depend on their work).
Respondents are asked to share their experiences openly and to answer the questions honestly. Respondents are assured that their responses will be used for academic research purposes only and confidentiality will be maintained with utmost care.
I): Ranking of DAD Risk Factors based on the impact on the Time of the project
II): Ranking of DAD Risk Factors based on their perceived impact on the Quality of the project outcome
