I just arrived back home in Switzerland after a two week trip to Vancouver, Canada. Together with a work colleague from our s.e.a.l. research group, we conducted two studies (one each) on analyzing and improving developer’s effectiveness. Both studies were exploratory studies where we wanted to follow our ideas and explore their applicability and if developers would be willing to use our prototypes over two weeks or more. This is how the two studies looked:

1st Study: Retrospection

In previous studies, we learnt how software developers perceive and measure their own productivity during their work. By using a monitoring tool, we can now track the measures that are of interest to developers and we want to evaluate how we can visualize the tracked data in a way that offers developers a meaningful retrospection of their workday. This could help developers to improve the awareness and self-analysis of their own work to better understand why they make progress on tasks or not, where they could improve themselves and what aspects of a workday have an influence on their productivity.

During the study we run a monitoring tool during the developer’s normal workdays for two weeks. A short time before the end of each developer’s workday, we showed them a retrospection of their workday, visualizing the tasks they worked on, the activities they had on their computer, how productive they perceived themselves during work and their user input. Subsequently, we interviewed them about what they liked or not, and how we could improve the visualizations and add value to the retrospection. Over the weekend, I created an update to the monitoring tool and visualizations, which we evaluated again in the second week of the study. Overall, we got very promising results and already have some developers that will continue using the tool in the future. I am happy to share more about the monitoring tool, visualizations and results later.

 

2nd Study: Biometric Sensor

Biometric sensors like heart rate measurement devices have become more and more suitable for daily use. In the biometrics study, we investigate how we can harness them to support developers during work. For instance, we investigate whether the interruptibility, perceived difficulty or emotions can be determined with data from such sensors. This could help to avoid interruptions at inopportune moments or suggest code artifacts that should be refactored.