Why did they do THAT? The hidden impact of programmers' psychological differences on software development

Conference: BCS 8th London Hopper Colloquium

Abstract:

Researchers into the merits of software development processes and tools often report the finding that individual differences predict success independently of the process or tool under investigation. Since their objective is to evaluate a method rather than its users, the phenomenon is reported as an interesting confound and is not usually explored in depth.

The research presented in this poster seeks to better understand the role of individual differences. It explores psychological characteristics behind programmer behaviours which affect the productivity of others. Regardless of whether the application is one which needs to be user-friendly, the source code needs to be ‘team-friendly’ to facilitate the work of peers who will integrate or maintain it.

The first phase of the research will investigate the experiences of such people as ‘users’ of the source code in order to understand which behaviours facilitate or hinder smooth and productive software development – what makes code ‘team-friendly’? The second phase will use experiments to focus on programming tasks relevant to common themes that emerge from the qualitative phase. It will explore the approaches that different programmers take in making detailed design decisions as they write their software. The definition of ‘good’ source code in this context is that it must be not just accurate but ‘team-friendly’. The aim is to identify psychometric measures or patterns of decision making that are characteristic of adept programmers.

The research has only recently begun, so the poster will set out the author’s current summary of the topic to invite feedback and discussion.

Source: Manual

Preferred by: Gail Ollis