Student perceptions of screen-viewing distributed pair programming : An explorative study of benefits and shortcomings with screen-viewing distributed pair programming

Sammanfattning: In this thesis we research distributed pair programming by making an explorative study of a distributed pair programming method, screen-viewing distributed pair programming. This method has a user share the visual contents of his screen with other users in order to enable collaboration. This study is conducted by distributing a questionnaire to undergraduate students and analyzing the results. The students were required to have previous experience with screen-viewing distributed programming. This is a practice where two pieces of software are utilized, one for sharing the computer display with the programming partner and another for sharing the code written together. Our goal is to contribute to the knowledge of pair programming and how it can be applied in these kinds of environments while giving further insight into benefits and drawbacks about the usage of the screen-viewing distributed pair programming. This thesis is of an explorative nature. Surveys from students constitute the bulk of the data set which was analysed using pie charts and tables. The survey questions were based on a previous article about distributed pair programming utilizing a synchronized development environment, written by Stelios Xinogalos, Maya Satratzemi, Alexander Chatzigeorgiou and Despina Tsom-panoudi. The results of the study indicated that the screen-viewing distributed pair programming approach experienced many benefits but also some drawbacks. These results are similar to what distributed pair programming with a synchronized development environment experiences. The students also experience a high satisfaction in partner choice regardless of being free to select their own partner or if they are allocated one. A majority of the respondents would prefer to work collaboratively in the future with regards to their screen-viewing distributed pair programming experience. Because the perceived benefits incurred from the screen-viewing distributed pair programming is similar to the benefits of a synchronized development environment, we conclude that the perceived benefits are not necessarily bound to the choice of method. Screen-viewing distributed pair programming is split into two main aspects, screen-viewing and code sharing. The screen-viewing part represented a smaller amount of perceived issues while the code sharing part represented a larger part. We conclude that there are difficulties with this approach, however that a substantial part are confined to the code-sharing aspect.