A Hidden Markov Model of Developer Learning Dynamics in Open Source Software Projects

Abstract

This study develops a stochastic model to capture developer learning dynamics in open source software projects (OSS). A hidden Markov model (HMM) is proposed that allows us to investigate (1) the extent to which individuals learn from their own experience and from interactions with peers, (2) whether an individual's ability to learn from these activities varies as she evolves/learns over time, and (3) to what extent individual learning persists over time. We calibrate the model based on six years of detailed data collected from 251 developers working on 25 OSS projects hosted at Sourceforge. Using the HMM, three latent learning states (high, medium, and low) are identified, and the marginal impact of learning activities on moving the developer between these states is estimated. Our findings reveal different patterns of learning in different learning states. Learning from peers appears to be the most important source of learning for developers across the three states. Developers in the medium learning state benefit the most through discussions that they initiate. On the other hand, developers in the low and the high states benefit the most by participating in discussions started by others. While in the low state, developers depend entirely upon their peers to learn, whereas in the medium or high state, they can also draw upon their own experiences. Explanations for these varying impacts of learning activities on the transitions of developers between the three learning states are provided. The HMM is shown to outperform the classical learning curve model. The HMM modeling of this study contributes to the development of a theoretically grounded understanding of learning behavior of individuals. Such a theory and associated findings have important managerial and operational implications for devising interventions to promote learning in a variety of settings.