There is a consensus that around 70% of all problems related to system development are consequences of inadequate requirements elicitation. Thus, an improvement on requirements elicitation is mandatory to any software development process improvement initiative.
When we observe the requirements area in a detailed manner, it is possible to notice that it depends directly on group activities such as requirements elicitation and validation, as there are many people from the client and development organisations involved. Any group task relies on several factors. Among these factors there are: interpersonal relationships, intergroup relationships, social and psychological factors, relationships between the group and the tasks to be performed, and finally the group and tasks’ relationships with the environment.
From this, it is possible to observe that many of the factors above are social rather than technical factors. These social factors are conflict sources such as: different perspectives, different backgrounds, issues related to previous personal divergences, among others. MacGrath (1983) presented figure 1 about groups’ tasks where the horizontal axis divides the graph between conflicts or collaboration. This idea shows that group tasks either rely on collaboration or depend on resolving conflicts.
Poundy (1967) studied conflicts in organizations and defined the “Conflict Theory”. According to this theory any conflict episode follows the same the lifecycle (figure 2) and is present in any organization. According to the Conflict Theory there are two types of conflicts: functional and dysfunctional. Functional conflicts are related to divergent opinions about a solution to a problem or about the way to implement a solution. Functional conflicts are considered positive conflicts as many times they raise a third position better than both individual solutions based on the combination of advantages of the initial ideas. Easterbrook (1994), for example, suggested the idea that groups’ different perspectives generate new versions of solutions. On the other hand, dysfunctional conflicts are related to emotions. These conflicts are not related to the subject under discussion and last long than the subject itself. As they are related to emotions, they are harder to perceive and control.
Unfortunately, even with the perception that in the requirements area social issues are as important as technical issues, most of the existing solutions focus on technical factors, dealing with social factors as a secondary problem. Thus, there is a gap waiting to be solved in requirements elicitation activities. Furthermore, even the few proposals that discuss requirements based conflicts focus uniquely on functional conflicts, leaving emotions uncovered. Consequently, several requirements elicitation activities are less efficient due to dysfunctional conflicts.
According to the conflict theory (Poundy, 1967), there are two ways to solve conflicts. Small conflicts can be solved by increasing collaboration. When dealing with intense conflicts, mainly dysfunctional, parties must be separated, as an increase in the collaboration would just exacerbate the conflict. Another issue to be discussed is that the existing proposals deal with conflicts when they are already manifest (figure 3). This can be the right approach for functional conflicts. However, in the case of dysfunctional conflicts it is important to find the sources of conflicts in order to eliminate them before the conflicts become manifest.
Muradas (2012) presented a solution for requirements elicitation where, in the presence of conflicts, the first attempt would be to increase collaboration by using group techniques such as brainstorming and workshops. In the case of intense conflicts the solution suggests the use of techniques that will reduce the direct interaction between individuals, such as observation or ethnography. The solution is presented in figure 4. This approach requires the observation of stakeholders´ personal attitudes and characteristics, which is the first step of the framework presented by Muradas (2012). Only after understanding stakeholders´ personal characteristics the framework can be used.
The framework proposed by Muradas (2012) was successfully used in the military environment where the main sources of conflicts observed were: hierarchy; high staff turnover; and multicultural teams. However, this framework can be applied in any other environment as long as the sources of conflicts are raised so as to be controlled and eliminated when needed.
- Easterbrook, S., (1994) “Resolving requirements conflicts with computer-supported negotiation” Requirements Engineering, Academic Press ltd.
- McGrath, J., (1983) “Groups: Interaction and Performance” Ed. Prentice Hall
- Muradas, F., (2012) “A Novel Framework for Requirements Elicitation in a Military Setting” Doctoral Thesis – Oxford University, Computer Science Department, UK
- Poundy, L. (1967) “Organizational Conflicts: Concepts and Models” Administrative Science Quarterly