An Empirical Analysis of Community and Coding Patterns in OSS4SG vs. Conventional OSS

Open Source Software for Social Good (OSS4SG) projects aim to address critical societal challenges, such as healthcare access and community safety. Understanding the community dynamics and contributor patterns in these projects is essential for ensuring their sustainability and long-term impact. However, while extensive research has focused on conventional Open Source Software (OSS), little is known about how the mission-driven nature of OSS4SG influences its development practices. To address this gap, we conduct a large-scale empirical study of 1,039 GitHub repositories, comprising 422 OSS4SG and 617 conventional OSS projects, to compare community structure, contributor engagement, and coding practices. Our findings reveal that OSS4SG projects foster significantly more stable and “sticky” (63.4%) communities, whereas conventional OSS projects are more “magnetic” (75.4%), attracting a high turnover of contributors. OSS4SG projects also demonstrate consistent engagement throughout the year, while conventional OSS communities exhibit seasonal fluctuations. Additionally, OSS4SG projects rely heavily on core contributors for both code quality and issue resolution, while conventional OSS projects leverage casual contributors for issue resolution, with core contributors focusing primarily on code quality.

💡 Research Summary

The paper presents a large‑scale empirical comparison of Open Source Software for Social Good (OSS4SG) projects and conventional open‑source software (OSS) projects. Using a curated dataset of 1,039 GitHub repositories—422 classified as OSS4SG and 617 as conventional OSS—the authors examine three research questions: (RQ1) differences in project characteristics, stability, and contributor retention; (RQ2) how contributor engagement patterns cluster and evolve over time; and (RQ3) the impact of those patterns on code quality.

Data collection involved strict filtering: each repository must have at least ten contributors, 500 commits, 50 closed pull requests, a lifespan longer than one year, and at least one update in the past year. After filtering, the final sample comprised 198 OSS4SG and 91 conventional OSS projects, encompassing more than three million commits. The authors combined information from the GitHub REST API, GraphQL API, and locally cloned repositories, and they applied three identity‑resolution techniques (email matching, username‑email normalization, and a machine‑learning approach) to de‑duplicate contributors, ultimately adopting the username‑email normalization method, which reduced duplicate records by roughly 10 %.

For RQ1, the study computed 23 quantitative metrics covering size, activity, community composition, and code characteristics. Metrics were normalized by non‑comment source‑code characters to control for project size. Statistical comparisons employed Mann‑Whitney U tests with Bonferroni correction (α = 0.0022) and effect‑size estimation using Cliff’s δ. Results show that OSS4SG projects tend to form “sticky” communities: 63.4 % of OSS4SG projects exhibit high contributor retention (re‑engagement rate ≈ 0.68), whereas 75.4 % of conventional OSS projects are “magnetic,” attracting many contributors but suffering higher turnover (re‑engagement rate ≈ 0.42). OSS4SG projects have a larger proportion of core contributors (average core‑contributor share ≈ 27 % vs. 15 % in OSS) and demonstrate more stable project lifespans, with longer intervals between major releases and shorter issue‑resolution times.

RQ2 investigates temporal engagement patterns. By aggregating monthly commit and pull‑request counts, the authors generated activity heatmaps. OSS4SG projects display a relatively flat activity curve across the twelve months, indicating consistent year‑round development. Conventional OSS projects, in contrast, show pronounced seasonal spikes—particularly in summer and during the holiday season—followed by troughs in winter months. Clustering of contribution trajectories reveals two dominant archetypes: (1) “core‑sustaining” clusters, prevalent in OSS4SG, where core contributors remain active throughout the project lifecycle; and (2) “core‑decaying” clusters, common in conventional OSS, where core activity wanes after an initial burst, leading to reliance on occasional contributors.

RQ3 links engagement patterns to code quality. The authors employed the Qodana static‑analysis suite to quantify critical, high, moderate, and low severity issues, and they measured structural metrics such as cyclomatic complexity, nesting depth, and comment‑to‑code ratio. Findings indicate that OSS4SG projects rely heavily on core contributors for both structural integrity and issue resolution. When core activity declines, the proportion of critical/high severity issues rises sharply, suggesting a vulnerability to core attrition. Conventional OSS projects, however, distribute issue‑resolution work more broadly among casual contributors, which buffers code‑quality fluctuations despite a lower overall core‑contributor share. Moreover, OSS4SG projects exhibit higher average comment‑to‑code ratios, reflecting a possible emphasis on documentation and knowledge transfer aligned with their social‑impact mission.

The paper also examines cross‑project contributor overlap. Within the OSS4SG ecosystem, 22 % of contributors are “boundary‑spanning,” meaning they have contributed to at least two OSS4SG projects, compared with 13 % in the conventional OSS ecosystem. This higher intra‑ecosystem overlap suggests stronger knowledge diffusion and community cohesion among mission‑driven projects. Inter‑ecosystem overlap (contributors active in both OSS4SG and conventional OSS) is modest (≈ 7 %), indicating that most contributors specialize in one ecosystem.

Based on these empirical insights, the authors propose actionable recommendations for OSS4SG maintainers: (1) strengthen onboarding pipelines for core contributors to reduce reliance on a few individuals; (2) organize year‑round community events (hackathons, workshops) to sustain engagement and mitigate seasonal dips; (3) adopt automated code‑review and static‑analysis tooling to maintain code quality even when core activity fluctuates; and (4) incorporate mentorship and recognition schemes proven effective in conventional OSS, adapted to emphasize social impact.

The study contributes the first large‑scale quantitative comparison of OSS4SG and conventional OSS, identifies distinct sustainability models (“sticky” vs. “magnetic” communities), and demonstrates how mission‑driven goals shape contributor behavior and software quality. All data, analysis scripts, and supplemental materials are publicly released (Zenodo DOI: 10.5281/zenodo.16337983), enabling replication and further research.