Identifying Developer Engagement in Open-Source Software Blockchain Projects through Factor Analysis

Abstract

The ubiquity of GitHub for software developers to coordinate software development in a community platform has resulted in a rich source of public data. Blockchain teams put open-source code as a founding principle since the release of Bitcoin and nearly all blockchain-based projects have code visible on GitHub. Developer engagement is known to be important to the health and viability of open-source software, yet has varying definitions and no standard method of measuring what constitutes engagement. This work uses exploratory factor analysis to identify dimensions that represent engagement in a community of open-source developers. We find that a latent factor composed of pull-requests, commits, comments, and authors based on a monthly average of the previous three months is a representation of Developer Engagement. A secondary factor consists of stars, forks, and total authors. Cross validation of the dataset is carried out with good support for the model.