Context: Motivation is known to improve performance. In software development in particular, there has been considerable interest in the motivation of contributors to open-source.
Objective: We identify 11 motivators from the literature (enjoying programming, ownership of code, learning, self-use, etc.), and evaluate their relative effect on motivation. Since motivation is an internal subjective feeling, we also analyze the validity of the answers.
Method: We conducted a survey with 66 questions on motivation which was completed by 521 developers. Most of the questions used an 11-point scale. We evaluated the answers’ validity by comparing related questions, comparing to actual behavior on GitHub, and comparison with the same developer in a follow-up survey.
Results: Validity problems include moderate correlations between answers to related questions, as well as self-promotion and mistakes in the answers. Despite these problems, predictive analysis—investigating how diverse motivators influence the probability of high motivation—provided valuable insights. The correlations between the different motivators are low, implying their independence. High values in all 11 motivators predict increased probability of high motivation. In addition, improvement analysis shows that an increase in most motivators predicts an increase in general motivation.
Conclusions: All 11 motivators indeed support motivation, but only moderately. No single motivator suffices to predict high motivation or motivation improvement, and each motivator sheds light on a different aspect of motivation. Therefore models based on multiple motivators predict motivation improvement with up to 94% accuracy, better than any single motivator.
Keywords: Motivation · Software engineering · Open-source development · Survey validity
This repository contain the data and source code use for this paper. To reproduce see code/python/main.py Note that original data containied personal identifiers (e.g., GitHub profiles), which was anonimized before publishing.y
See also additional analysis, not included in paper due to length. The analysis included demography and it comparision to the Stack Overflow annual survey. It also contains the use of reduction to supervised learning to differ between participants contacted by email and by social media (they are rather similar). Last, it contains an analysis between the different motivators.