Attempts to resolve the problems in software development have concentrated on the tools and methodologies used, despite an acceptance by many that it is a sociological problem. An example of this is the procedures and processes surrounding evaluations within projects, yet ultimately it depends on individuals more than process. This paper examines one of the sociological factors inherent in a software development team to determine its impact on evaluation within a project. Social loafing occurs where individual members of a team demonstrate a tendency not to work as hard as they could or should. This "slacking off" occurs because the team provides a degree of anonymity – the individual feels their lack of work will be hidden from evaluation within the overall output of the team. Some authors purport that Agile Software development teams have low incidences of social loafing (though these are opinions rather than research findings); the contrary can also be argued. An examination of the philosophy behind Agile Software Development, demonstrated by the Agile Manifesto, highlighted the possibility of occurrences of social loafing brought about by the Agile values. Agile espouses the importance of cohesive teams, the empowerment of these teams, and the collective ownership and self‑ evaluation of work by the team. These values map onto factors which are described as affecting social loafing. An investigation of two teams over an eight month period examined if the Agile values could lead to incidences of social loafing, specifically when their work is being evaluated The investigation determined that the opposite was actually the case. This paper then goes on to determine why the findings go against the initial hypothesis and to show the impact this can have on those evaluating software development projects.

Abstract: As researchers evaluate organisations, projects, and teams, there is a desire for a consensus from those within the organisations who are participating in the research. A common consensual perspective from a team appears to reflect an optimal st ate where those being evaluated have a common understanding of the current state of events within the context of their environment. The question arises, though, whether an evaluation finding consensus reflects the reality: there are a variety of reasons w hy a common understanding may be false consensus. Hidden behind this false consensus may be a variety of unaddressed issues which are actually the core of the problem. This paper proposes an evaluation method incorporating the principles of sensemaking an d devils advocate, where a consensus of perspectives is challenged before they are considered valid. This is achieved in a workshop where participants reflect on their own perception of reality and represent this reality in a matrix of influencing and re levant factors. The individual matrices are then combined and used to highlight disparities in the participants perspectives through a single matrix visualisation. Discussion in the workshop then focusses on the areas, highlighted by the matrix, where di fferences of perspectives are identified. In effect, the consensus presented by those being evaluated will be challenged, and a new common understanding will have to be created. Problems such as groupthink can create a false consensus, and it is proposed herein that the workshop provides a mechanism for challenging this. The objective of the research herein was to determine the feasibility and potential benefits of the proposed workshop. The workshop itself is evaluated in this paper, to determine if it has value. The benefits of such a workshop are described, showing how an organisation went from a false consensus concerning problems within the organisation, to the start of a process to address the real underlying issues.

For future projects to improve, it is necessary to evaluate the lessons from previous projects. The majority of software methodologies recommend a review of the project to evaluate what worked and what needs improvement. These reviews are commonly referred to as project post‑mortems. Existing research into post‑mortems has found problems with the actual process itself and the use of the output from the process — the lessons learned. This research examines project post‑mortems before the post‑mortem has occurred — it is an examination of the beliefs and attitudes that project members bring with them into post‑mortems. These attitudes can ultimately cause the failure of a post‑ mortem, even before it has begun. It is somewhat paradoxical that team members initially espoused positive views about post‑mortems in a survey, yet further examination of key informants showed that these espoused views did not translate into reality. It is shown how hierarchical groupthink can help to forge negative beliefs and attitudes about post‑mortems that will have a detrimental affect on the process itself.