Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Psychological Safety Framework in Pull-based Open Source Projects

T0 review · 3 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Pull request interactions can reveal a repository's psychological safety, and repositories that score higher keep contributors active longer.

desk verdict The central result is a tautology: the psychological safety index is built from the very retention outcome it claims to predict, so Table V's positive coefficients are guaranteed by construction. read the letter →

arxiv 2504.17510 v4 pith:GEUQBN3E submitted 2025-04-24 cs.SE

Emeralda Sesari , Federica Sarro , Ayushi Rastogi This is my paper · ORCID
classification cs.SE
keywords psychologicalsafetypull-baseddevelopmentopensourcesoftwaresustainedparticipationcodereviewlogisticregressioncontributorretention
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that team-level psychological safety in open-source projects can be read from pull request interactions, and that repositories with higher levels of this measured safety retain contributors longer. The authors build a ten-cue psychological safety index from observable review behaviours, including merges, comments, exchanges between contributors and integrators, participants, and mentions, using 60,684 pull requests from 26 popular repositories. They report that contributors in higher-scoring repositories are more likely to remain active both within 12 months and over 4 to 5 years. A third model shows that a contributor's prior participation is an even stronger predictor of future activity, and the psychological safety effect, while still statistically significant, becomes much smaller once past activity is taken into account.

What carries the argument

The central object is the psychological safety index, a 0 to 10 score defined at the pull request level and averaged over contributors and then repositories. Each of ten cues, including merge outcome, comment counts, contributor-integrator exchange, contributor, integrator, reviewer and other comments, number of participants, and mentions, contributes one point to a pull request's score. In the construction used here, a cue counts only when the contributor remained active after that pull request; the repository-level mean of this index serves as the predictor in three logistic regression models of sustained participation, with controls for contributor centrality, visibility, language breadth, team familiarity, and repository size.

What would settle it

Re-run Models 1-3 with the psychological safety index recomputed without conditioning on the contributor's later activity, scoring each cue by whether the interaction occurred rather than by the contributor's survival; if the positive coefficients disappear or reverse, the result is an artifact of outcome-conditioned construction. A complementary check is to compare the index against survey-based psychological safety scores from the same contributors.

Watch

Extended reading notes

Core claim

On its own terms, the paper demonstrates that psychological safety, defined as the shared belief that a team is safe for interpersonal risk-taking, can be operationalized at scale from code review activity rather than surveys. The central result is that the repository-level psychological safety index is positively and significantly associated with sustained participation in all three logistic regression models, with coefficients of 1.03 in the short-term model, 0.32 in the long-term model, and 0.05 once prior participation is included. The authors interpret this as evidence that a psychologically safe review environment supports retention, especially in the near term, while established participation patterns dominate long-term engagement.

Load-bearing premise

The psychological safety index counts a cue only when the contributor remained active after the pull request, so the predictor contains the outcome it is later used to predict; if that scoring rule is invalid, the positive associations in all three models are partly a measurement artifact.

Editorial extensions

If this is right

  • If the index is a valid proxy, maintainers can track and improve psychological safety without relying on surveys, using pull request data they already have.
  • Short-term retention is more strongly tied to the psychological safety index than long-term retention, suggesting safety matters most early in a contributor's involvement.
  • Prior participation is a stronger predictor than the safety index, so retention efforts may need to target newcomers before participation patterns harden.
  • Repository scale is a major confound: contributors in large repositories were far more likely to stay active, so safety effects should be interpreted relative to project size.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because each cue is counted only when the contributor stayed active after the pull request, the index partially encodes the outcome it predicts; recomputing the index from interaction features alone would test whether the reported associations are an artifact of that construction.
  • The shrinking effect size once prior participation enters the model hints at self-selection: contributors who feel safe stay, accumulate activity, and then persist for habit or investment reasons; future work could model psychological safety as a mediator of the participation-activity loop.
  • A natural extension is to score negative cues, such as hostile comments, rejected pull requests without feedback, or ignored questions, so the index captures psychological unsafety rather than only engagement-positive signals.
  • Maintainer-facing interventions, such as standardized welcoming replies on a contributor's first pull request, could be randomized across repositories to test whether the modifiable cues causally affect retention.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes a framework for measuring team-level psychological safety in open-source repositories from observable pull-request (PR) interaction cues. The authors construct a psychological safety index from ten cues, aggregate it to contributor and repository levels, and use logistic regression on 60,684 PRs from 26 repositories to test whether higher repository-level psychological safety is associated with contributors' sustained participation. Three models are reported: short-term sustained participation, long-term sustained participation, and long-term participation controlling for prior participation. The central claim is that contributors are more likely to remain active in repositories with higher psychological safety.

Significance. If the measurement and regression were valid, the paper would offer a scalable, data-driven proxy for a construct that is usually studied with surveys and interviews, and the replication materials (scripts, data, and instructions) are a genuine strength for the mining-software-repositories community. The theory-informed cue selection and the explicit treatment of control variables also show care. However, the central empirical result is undermined by the way the predictor is constructed: the psychological safety index in Section III-E is defined conditional on the outcome (sustained participation), which makes the reported associations in Table V partly tautological. This is not a disagreement with the literature's consensus on psychological safety; it is an internal measurement problem in the study's own design.

major comments (3)
  1. [Section III-E (Calculating Psychological Safety Index as Predictor Variable)] The PS index is defined so that a PR receives credit for a cue only if 'the contributor remained active after the PR (i.e., sustained participant)' and an interaction condition is met. Since the regression outcome in Models 1-3 is exactly this sustained-participation status, the predictor is a function of the outcome. Aggregation to the contributor or repository level does not remove this leakage: repositories with more contributors who stayed active will mechanically have higher PS indices. The positive coefficients for PS_index_repository in Table V (beta = 1.03, 0.32, 0.05) are therefore expected by construction and do not provide empirical support for the hypothesis that psychological safety sustains participation.
  2. [Section III-E, Table II (merged_or_not cue)] The first cue condition is written as 'PR was merged or not merged (merged_or_not = 1 or = 0)'. Because either value of merged_or_not satisfies the condition, this cue contributes 1 for every PR of a contributor who remained active, regardless of the actual outcome. Thus even the one seemingly technical cue in the index reduces to an indicator of sustained activity, reinforcing the circularity in the predictor construction.
  3. [Section VI (Limitations)] The limitations section states that psychological safety is measured indirectly from PR interactions and suggests future validation with surveys and qualitative methods, but it does not acknowledge that the index construction in Section III-E already incorporates the outcome variable. This omission matters because the reported association is not a falsifiable empirical test as designed; the manuscript needs a predictor definition not conditioned on the outcome, or the central claim cannot be evaluated.
minor comments (4)
  1. [Section III-C] The definition of sustainedp_or_not_12 is phrased ambiguously: a contributor is 'marked as a sustained contributor if they remained active during this period, and as non-sustained if they became inactive before June 30, 2019.' It should clarify that eligibility depends on being active within the 12-month window after the 2019 snapshot, and that contributors whose last commit falls near the end of the study period are handled as described.
  2. [Section III-E, Table IV] The text says 'Table IV presents PS index for each of the ten selected repositories,' but Table IV lists 26 repositories and the study uses 26 repositories. This appears to be an inconsistency between the narrative and the table.
  3. [Section III-D, Table II and Table III] comment_conflict is excluded from the final set of cues because of imbalance (1.8% positive), yet it still appears in Table II as a cue and in Table III with descriptive statistics. The manuscript should explicitly mark excluded variables in Table II or state that Table II lists candidate cues that were later filtered.
  4. [Section III-E, aggregation] The repository-level PS index is computed by averaging contributor-level means, without any reported weighting by number of PRs per contributor. The manuscript should state whether this unweighted aggregation is intentional and discuss its effect on repositories with many one-time contributors.

Circularity Check

1 steps flagged · score 9.0 of 10

Psychological safety index is defined using the sustained-participation outcome, so the positive associations in Table V are contaminated by construction.

  1. self definitional [Section III-E (Calculating Psychological Safety Index as Predictor Variable) and Section IV, Table V (Models 1-3)]
    "We considered a PR as psychologically safe (i.e., assigned a value of 1 for a specific cue) if the contributor remained active after the PR (i.e., sustained participant) and one of the following interaction conditions was met:"

    The PS index is constructed by setting each cue to 1 only when the contributor remained active after the PR, i.e., when the contributor was a sustained participant. That is exactly the outcome variable used in Models 1-3 (sustainedp_or_not_12 and recent_sustainedp_or_not). A PR by a contributor who disengages contributes 0 to every cue, while a PR by a contributor who stays active contributes the number of satisfied interaction conditions. Aggregating to the contributor and repository level does not remove this leakage: repositories with more sustained contributors mechanically receive higher PS_index_repository values.

full rationale

The central empirical claim is that repository-level psychological safety is positively associated with sustained participation. The paper's own definition of the PS index, however, makes 'psychologically safe' conditional on the contributor remaining active after the PR. Since the regression outcomes measure exactly that kind of sustained participation, the predictor and outcome share the same underlying event by construction. Repository-level averaging over contributors does not break the dependence: contributors who leave contribute zero to every PS cue, while contributors who stay contribute a positive cue count, so repositories with more retained contributors will tend to have higher PS indices. The strength of the coefficients is therefore inflated by the measurement procedure, making the results in Table V partly tautological. The paper also frames the Model 3 shrinkage as a substantive finding about prior participation outweighing PS, but prior participation is already baked into the PS index, so the comparison is not clean. No external, independently validated measure of psychological safety is introduced, and no machine-checked or code-reproduced validation is cited to break the circularity. Because the interaction-condition component provides some independent variation, the circularity is not absolute, but the headline claim reduces in large part to the paper's own definition, warranting a score of 9 rather than 10.

Assumptions & free parameters 4 free parameters · 3 assumptions · 1 invented entities

The central measurement instrument, the psychological safety index, is defined using the outcome variable. Continuous cue thresholds are unspecified and likely data-driven. The statistical models rely on unclustered standard errors despite repeated observations and a repository-level predictor.

free parameters (4)
  • High thresholds for continuous psychological safety cues = Not reported
    The psychological safety index binarizes pr_comment_num, num_comments_con, and num_participant as 'high' (Section III-E) without stating the cutoff. The threshold is likely data-driven and directly determines cue values, so it is a free parameter affecting the central result.
  • Skewness exclusion cutoff = Not reported
    Variables were removed when skewness remained high after log transformation, with reopen_num skewness 7.31, emoji_count 12.33, and watchers -3.39. The cutoff for 'high skewness' is not pre-specified, and these exclusions affect the psychological safety index composition and the control set.
  • Imbalance threshold for comment_conflict = 1.8% positive
    comment_conflict was excluded because only 1.8% of instances had the value 1. The threshold for problematic imbalance is not stated, and this is a post hoc modeling choice.
  • Repository size category boundaries = Large >1000 PRs, medium 101-1000, small <100
    The bins for the repo_size control are arbitrary and influence the very large odds ratios reported, such as OR 46.5 for large repositories in Model 1.
assumptions (3)
  • domain assumption Psychological safety in open-source teams can be inferred from observable pull request interaction cues.
    Invoked throughout Section III-D; the paper itself flags in Limitations that the cues do not directly capture contributors' perceptions, so this is an unvalidated assumption.
  • domain assumption A contributor is disengaged after 12 consecutive months without commits, and the GitHub commit history up to June 2020 and December 2024 is sufficient to classify retention.
    Borrowed from Qiu et al. [24] in Section III-C; this definition ignores non-commit contributions and assumes the snapshot windows capture disengagement.
  • standard math Logistic regression with 60,684 pull request rows provides valid inference, despite multiple observations per contributor and repository-level predictors.
    The models treat each pull request as independent in Section III-G, but contributors and repositories are repeated; no clustered or mixed-effects models are reported, so standard errors are likely underestimated.
invented entities (1)
  • Psychological Safety Index
    purpose: A 0-10 score summarizing team-level psychological safety from pull request cues, used as the main predictor.
    The index is defined conditional on the outcome of sustained participation, so it is not an independent measure and provides no falsifiable handle outside the paper's own construction.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Psychological Safety Framework in Pull-based Open Source Projects." pith.science (2026). https://pith.science/paper/GEUQBN3E

@misc{pith2026250417510,
  author       = {Pith},
  title        = {Pith review of: Psychological Safety Framework in Pull-based Open Source Projects},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GEUQBN3E}},
  note         = {Machine review of arXiv:2504.17510}
}
read the original abstract

Psychological safety refers to the belief that team members can speak up, ask questions, and make mistakes without fear of negative consequences. Although psychological safety has been studied in traditional software teams, less is known about how it may appear in pull-based open-source software development, where contributors are self-directed and often collaborate voluntarily. This paper introduces a theory-informed framework for understanding how psychological safety may be reflected in pull request interactions. Drawing on psychological safety theory and prior work on software teams and open-source collaboration, the framework identifies observable interaction patterns related to feedback exchange, active participation, asking for input, and visible engagement from relevant project actors. To examine the framework empirically, we operationalize these patterns using nine observable variables from 60,684 pull requests across 26 popular GitHub repositories. The empirical results refine the framework by showing that visible engagement from contributors, reviewers, integrators, and other project members is positively associated with sustained participation, while interaction appears most useful when there is enough discussion without becoming excessive.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. What Motivates Whom? A Survey of Newcomers to OSS and Experienced OSS Practitioners

    cs.SE 2026-07 conditional novelty 5.0 of 10

    Demographics and motivations correlate with OSS project-selection preferences, with distinct patterns for newcomers versus experienced practitioners in a 208-person survey.

Reference graph

Works this paper leans on

61 extracted references · 1 canonical work pages · cited by 1 Pith paper

  1. [1]

    Software e ngineering team diversity and performance,

    V . Pieterse, D. G. Kourie, and I. P . Sonnekus, “Software e ngineering team diversity and performance,” in Proceedings of the 2006 annual research conference of the South African institute of compu ter scientists and information technologists on IT research in developing countries, 2006, pp. 180–186

    2006

  2. [2]

    Bringing the human factor to softw are eng

    L. Fernando Capretz, “Bringing the human factor to softw are eng.” IEEE Softw., vol. 31, no. 2, pp. 104–104, 2014

    2014

  3. [3]

    Psychological safety and learning behav ior in work teams,

    A. Edmondson, “Psychological safety and learning behav ior in work teams,” Administrative science quarterly , vol. 44, no. 2, pp. 350–383, 1999

    1999

  4. [4]

    What Google Learned From Its Quest to Build the Perfect T eam,

    “What Google Learned From Its Quest to Build the Perfect T eam,” https://www.nytimes.com/2016/02/28/magazine/what-google-learned-from-its-quest-to-build-the-perfect-te am.html, 2016, [Accessed 01-02-2025]

    2016

  5. [5]

    Exploring psychological safety in softwar e engineer- ing: Insights from stack exchange,

    B. S. D. Santana, S. Freire, L. Cruz, L. Monte, M. Mendonca , and J. A. M. Santos, “Exploring psychological safety in softwar e engineer- ing: Insights from stack exchange,” in Proceedings of the XXXVII Brazil- ian Symposium on Software Engineering , ser. SBES ’23. New Y ork, NY , USA: Association for Computing Machinery, 2023, p. 503– 513

    2023

  6. [6]

    Psychological safety in the software work environment,

    B. Santana, S. Freire, J. A. M. Santos, and M. Mendonc ¸a, “ Psychological safety in the software work environment,” IEEE Software, vol. 41, no. 4, pp. 86–94, 2024

    2024

  7. [7]

    Psychological safety and norm c larity in software engineering teams,

    P . Lenberg and R. Feldt, “Psychological safety and norm c larity in software engineering teams,” in Proceedings of the 11th International W orkshop on Cooperative and Human Aspects of Software Engin eering, ser. CHASE ’18. New Y ork, NY , USA: Association for Computing Machinery, 2018, p. 79–86

    2018

  8. [8]

    It is giving major sa tisfaction: Why fairness matters for developers,

    E. Sesari, F. Sarro, and A. Rastogi, “It is giving major sa tisfaction: Why fairness matters for developers,” arXiv preprint arXiv:2410.02482, 2024

    arXiv 2024

Show all 61 references
  1. [9]

    ”was my contribution fairly reviewed?

    D. M. German, G. Robles, G. Poo-Caama˜ no, X. Y ang, H. Iida , and K. Inoue, “”was my contribution fairly reviewed?” a framewo rk to study the perception of fairness in modern code reviews,” in Proceedings of the 40th International Conference on Software Engineering , 2018, pp...

    2018

  2. [10]

    Psychological Safety Short Course,

    “Psychological Safety Short Course,” https://handbook.gitlab.com/handbook/leadership/emotional-intelligence/psychological-safety-short-cour se/#what-impacts-psychological-safety, 2021, [Accessed 01-02-2025]

    2021

  3. [11]

    The role of psycho logical safety in promoting software quality in agile teams,

    A. Alami, M. Zahedi, and O. Krancher, “The role of psycho logical safety in promoting software quality in agile teams,” Empirical Softw. Engg. , vol. 29, no. 5, Jul. 2024

    2024

  4. [12]

    How do team cohesion and psychological saf ety impact knowledge sharing in software development projects?

    A. K. Kakar, “How do team cohesion and psychological saf ety impact knowledge sharing in software development projects?” Knowledge and Process Management, vol. 25, no. 4, p. 258–267, 2018

    2018

  5. [13]

    Perceptio ns of diversity on git hub: A user survey,

    B. V asilescu, V . Filkov, and A. Serebrenik, “Perceptio ns of diversity on git hub: A user survey,” in 2015 IEEE/ACM 8th International W orkshop on Cooperative and Human Aspects of Software Engineering . IEEE, 2015, pp. 50–56

    2015

  6. [14]

    Understanding sustained parti cipation in open source software projects,

    Y . Fang and D. Neufeld, “Understanding sustained parti cipation in open source software projects,” Journal of Management Information Systems , vol. 25, no. 4, pp. 9–50, 2009

    2009

  7. [15]

    Underst anding the motivations, participation, and performance of open sourc e software developers: A longitudinal study of the apache projects,

    J. A. Roberts, I.-H. Hann, and S. A. Slaughter, “Underst anding the motivations, participation, and performance of open sourc e software developers: A longitudinal study of the apache projects,” Management science, vol. 52, no. 7, pp. 984–999, 2006

    2006

  8. [16]

    Motivation of s oftware devel- opers in open source projects: an internet-based survey of c ontributors to the linux kernel,

    G. Hertel, S. Niedner, and S. Herrmann, “Motivation of s oftware devel- opers in open source projects: an internet-based survey of c ontributors to the linux kernel,” Research policy , vol. 32, no. 7, pp. 1159–1177, 2003

    2003

  9. [17]

    Developer turnov er in global, industrial open source projects: Insights from applying su rvival analy- sis,

    B. Lin, G. Robles, and A. Serebrenik, “Developer turnov er in global, industrial open source projects: Insights from applying su rvival analy- sis,” in 2017 IEEE 12th International Conference on Global Software Engineering (ICGSE) . IEEE, 2017, pp. 66–75

    2017

  10. [18]

    Who will remai n? an eval- uation of actual person-job and person-team fit to predict de veloper retention in floss projects,

    A. Schilling, S. Laumer, and T. Weitzel, “Who will remai n? an eval- uation of actual person-job and person-team fit to predict de veloper retention in floss projects,” in 2012 45th Hawaii international conference on system sciences . IEEE, 2012, pp. 3446–3455

    2012

  11. [19]

    Will my patch make i t? and how fast? case study on the linux kernel,

    Y . Jiang, B. Adams, and D. M. German, “Will my patch make i t? and how fast? case study on the linux kernel,” in 2013 10th W orking conference on mining software repositories (MSR) . IEEE, 2013, pp. 101–110

    2013

  12. [20]

    Wil l they like this? evaluating code contributions with language models,

    V . J. Hellendoorn, P . T. Devanbu, and A. Bacchelli, “Wil l they like this? evaluating code contributions with language models, ” in 2015 IEEE/ACM 12th W orking Conference on Mining Software Reposi tories. IEEE, 2015, pp. 157–167

    2015

  13. [21]

    A study of external c ommunity contribution to open-source projects on github,

    R. Padhye, S. Mani, and V . S. Sinha, “A study of external c ommunity contribution to open-source projects on github,” in Proceedings of the 11th working conference on mining software repositories, 2014, pp. 332– 335

    2014

  14. [22]

    Writing acceptable patches: An empirical study of open source project patches,

    Y . Tao, D. Han, and S. Kim, “Writing acceptable patches: An empirical study of open source project patches,” in 2014 IEEE International Conference on Software Maintenance and Evolution . IEEE, 2014, pp. 271–280

    2014

  15. [23]

    An explorato ry study of the pull-based software development model,

    G. Gousios, M. Pinzger, and A. v. Deursen, “An explorato ry study of the pull-based software development model,” in Proceedings of the 36th international conference on software engineering , 2014, pp. 345–355

    2014

  16. [24]

    Going farther together: The impact of social capital on sustained participation in open source,

    H. S. Qiu, A. Nolte, A. Brown, A. Serebrenik, and B. V asil escu, “Going farther together: The impact of social capital on sustained participation in open source,” in 2019 IEEE/ACM 41st International Conference on Software Engineering (ICSE) , 2019, pp. 688–699

    2019

  17. [25]

    Pull req uest gov- ernance in open source communities,

    A. Alami, R. Pardo, M. L. Cohn, and A. Wasowski, “Pull req uest gov- ernance in open source communities,” IEEE Transactions on Software Engineering, vol. 48, no. 12, pp. 4838–4856, 2021

    2021

  18. [26]

    What are they talking about? analyzing code reviews in pull-based de velopment model,

    Z.-X. Li, Y . Y u, G. Yin, T. Wang, and H.-M. Wang, “What are they talking about? analyzing code reviews in pull-based de velopment model,” Journal of Computer Science and Technology, vol. 32, pp. 1060– 1075, 2017

    2017

  19. [27]

    What make long term contributors : Will- ingness and opportunity in oss community,

    M. Zhou and A. Mockus, “What make long term contributors : Will- ingness and opportunity in oss community,” in 2012 34th International Conference on Software Engineering (ICSE) . IEEE, 2012, pp. 518–528

    2012

  20. [28]

    On the shoulders of gian ts: A new dataset for pull-based development research,

    X. Zhang, A. Rastogi, and Y . Y u, “On the shoulders of gian ts: A new dataset for pull-based development research,” in Proceedings of the 17th international conference on mining software repositories, 2020, pp. 543– 547

    2020

  21. [29]

    A dataset for pull-based dev elopment research,

    G. Gousios and A. Zaidman, “A dataset for pull-based dev elopment research,” in Proceedings of the 11th W orking Conference on Mining Software Repositories, 2014, pp. 368–371

    2014

  22. [30]

    Social networkin g meets software development: Perspectives from github, msdn, stack exchan ge, and topcoder,

    A. Begel, J. Bosch, and M.-A. Storey, “Social networkin g meets software development: Perspectives from github, msdn, stack exchan ge, and topcoder,” IEEE software , vol. 30, no. 1, pp. 52–66, 2013

    2013

  23. [31]

    Understanding th e factors that impact the popularity of github repositories,

    H. Borges, A. Hora, and M. T. V alente, “Understanding th e factors that impact the popularity of github repositories,” in 2016 IEEE international conference on software maintenance and evolution (ICSME) . IEEE, 2016, pp. 334–344

    2016

  24. [32]

    What’s in a github star? und erstanding repository starring practices in a social coding platform,

    H. Borges and M. T. V alente, “What’s in a github star? und erstanding repository starring practices in a social coding platform, ” Journal of Systems and Software , vol. 146, pp. 112–129, 2018

    2018

  25. [33]

    4.5 million (suspected) fake stars in github: A growing spiral of popularity contests, scams, and malware,

    H. He, H. Y ang, P . Burckhardt, A. Kapravelos, B. V asiles cu, and C. K¨ astner, “4.5 million (suspected) fake stars in github: A growing spiral of popularity contests, scams, and malware,” 2024. [ Online]. Available: https://arxiv.org/abs/2412.13459

    2024 arXiv

  26. [34]

    No ri sk, no reward: Towards an automated measure of psychological safe ty from online communication,

    S. A. Ferguson, G. V an de Zande, and A. Olechowski, “No ri sk, no reward: Towards an automated measure of psychological safe ty from online communication,” in Extended Abstracts of the CHI Conference on Human Factors in Computing Systems , 2024, pp. 1–7

    2024

  27. [35]

    The effects of m ild task conflict and relationship conflict on psychological saf ety and team effectiveness: Team-level analysis,

    O. Seon Y ong, K. Y eong Sik, and K. In Hye, “The effects of m ild task conflict and relationship conflict on psychological saf ety and team effectiveness: Team-level analysis,” Korean Jounal of Industrial Organizational Psychology, vol. 32, no. 1, pp. 83–106, 2019

    2019

  28. [36]

    Optimizing team confli ct dynamics for high performance teamwork,

    T. A. O’Neill and M. J. McLarnon, “Optimizing team confli ct dynamics for high performance teamwork,” Human Resource Management Review, vol. 28, no. 4, pp. 378–394, 2018

    2018

  29. [37]

    Statistics for c linicians: An introduction to logistic regression

    M. M. Wiest, K. J. Lee, and J. B. Carlin, “Statistics for c linicians: An introduction to logistic regression.” Journal of Paediatrics & Child Health, vol. 51, no. 7, 2015

    2015

  30. [38]

    Poster: The impact of group im balance on logistic regression analyses with assessment data,

    A. Alkhalaf and B. Zumbo, “Poster: The impact of group im balance on logistic regression analyses with assessment data,” in ITC 2016 Conference, 2016

    2016

  31. [39]

    Effect of zero imputation methods for log-t ransformation of independent variables in logistic regression,

    S. Y . Park, “Effect of zero imputation methods for log-t ransformation of independent variables in logistic regression,” Communications for Statistical Applications and Methods , vol. 31, no. 4, pp. 409–425, 2024

    2024

  32. [40]

    Application of robust regression in translational neuros cience studies with non-gaussian outcome data,

    M. Malek-Ahmadi, S. D. Ginsberg, M. J. Alldred, S. E. Cou nts, M. D. Ikonomovic, E. E. Abrahamson, S. E. Perez, and E. J. Mufs on, “Application of robust regression in translational neuros cience studies with non-gaussian outcome data,” Frontiers in Aging Neuroscience , vol. 1...

    2024

  33. [41]

    Logistic regression in rare events data,

    G. King and L. Zeng, “Logistic regression in rare events data,” Political analysis, vol. 9, no. 2, pp. 137–163, 2001

    2001

  34. [42]

    Predictive per formance of logistic regression for imbalanced data with categorica l covariate

    H. A. A. Rahman, Y . B. Wah, and O. S. Huat, “Predictive per formance of logistic regression for imbalanced data with categorica l covariate.” Pertanika Journal of Science & Technology , vol. 29, no. 1, 2021

    2021

  35. [43]

    Imbalance effects on clas sification using binary logistic regression,

    H. A. Abd Rahman and B. W. Y ap, “Imbalance effects on clas sification using binary logistic regression,” in Soft Computing in Data Science: Second International Conference, SCDS 2016, Kuala Lumpur , Malaysia, September 21-22, 2016, Proceedings 2 . Springer, 2016, pp. 136–147

    2016

  36. [44]

    Aggregation of binary evaluations: a borda-like approach,

    C. Duddy, A. Piggins, and W. S. Zwicker, “Aggregation of binary evaluations: a borda-like approach,” Social Choice and W elfare, vol. 46, pp. 301–333, 2016

    2016

  37. [45]

    Should we care (more) abou t data aggre- gation?

    K. Gr¨ undler and T. Krieger, “Should we care (more) abou t data aggre- gation?” European Economic Review , vol. 142, p. 104010, 2022

    2022

  38. [46]

    Injury sever ity analysis: comparison of multilevel logistic regression models and ef fects of collision data aggregation,

    T. Usman, L. Fu, and L. F. Miranda-Moreno, “Injury sever ity analysis: comparison of multilevel logistic regression models and ef fects of collision data aggregation,” Journal of Modern Transportation , vol. 24, pp. 73–87, 2016

    2016

  39. [47]

    Understanding

    J. Sheoran, K. Blincoe, E. Kalliamvakou, D. Damian, and J. Ell, “Understanding” watchers” on github,” in Proceedings of the 11th working conference on mining software repositories , 2014, pp. 336–339

    2014

  40. [48]

    Do women suffer from network closure? the mo derating effect of social capital on gender inequality in a project-b ased labor market, 1929 to 2010,

    M. Lutter, “Do women suffer from network closure? the mo derating effect of social capital on gender inequality in a project-b ased labor market, 1929 to 2010,” American Sociological Review , vol. 80, no. 2, pp. 329–358, 2015

    1929

  41. [49]

    Disruptive diver sity and recur- ring cohesion: Assembling creative teams in the video game i ndustry, 1979-2009,

    M. d. V aan, B. V edres, and D. C. Stark, “Disruptive diver sity and recur- ring cohesion: Assembling creative teams in the video game i ndustry, 1979-2009,” 2011

    1979

  42. [50]

    Scientific collaboration networks. ii. s hortest paths, weighted networks, and centrality,

    M. E. Newman, “Scientific collaboration networks. ii. s hortest paths, weighted networks, and centrality,” Physical review E , vol. 64, no. 1, p. 016132, 2001

    2001

  43. [51]

    Developer onboarding in github: the role of prior social links and lang uage expe- rience,

    C. Casalnuovo, B. V asilescu, P . Devanbu, and V . Filkov, “Developer onboarding in github: the role of prior social links and lang uage expe- rience,” in Proceedings of the 2015 10th joint meeting on foundations of software engineering , 2015, pp. 817–828

    2015

  44. [52]

    How early par - ticipation determines long-term sustained activity in git hub projects?

    W. Xiao, H. He, W. Xu, Y . Zhang, and M. Zhou, “How early par - ticipation determines long-term sustained activity in git hub projects?” in Proceedings of the 31st ACM Joint European Software Enginee ring Conference and Symposium on the F oundations of Software Engineering, 20...

    2023

  45. [53]

    A large scale stud y of long- time contributor prediction for github projects,

    L. Bao, X. Xia, D. Lo, and G. C. Murphy, “A large scale stud y of long- time contributor prediction for github projects,” IEEE Transactions on Software Engineering, vol. 47, no. 6, pp. 1277–1298, 2019

    2019

  46. [54]

    On the effect of disc ussions on pull request decisions

    M. Golzadeh, A. Decan, and T. Mens, “On the effect of disc ussions on pull request decisions.” in BENEVOL, 2019

    2019

  47. [55]

    Women’s participation in open source software: A survey of the lit- erature,

    B. Trinkenreich, I. Wiese, A. Sarma, M. Gerosa, and I. St einmacher, “Women’s participation in open source software: A survey of the lit- erature,” ACM Transactions on Software Engineering and Methodology (TOSEM), vol. 31, no. 4, pp. 1–37, 2022

    2022

  48. [56]

    Climate coach: A dashboard fo r open- source maintainers to overview community dynamics,

    H. S. Qiu, A. Lieb, J. Chou, M. Carneal, J. Mok, E. Amspoke r, B. V asilescu, and L. Dabbish, “Climate coach: A dashboard fo r open- source maintainers to overview community dynamics,” in Proceedings of the 2023 CHI Conference on Human Factors in Computing Syst ems, 2023, pp. 1–18

    2023

  49. [57]

    Almost there: A study on quasi-contributors in open source software proje cts,

    I. Steinmacher, G. Pinto, I. S. Wiese, and M. A. Gerosa, “ Almost there: A study on quasi-contributors in open source software proje cts,” in Pro- ceedings of the 40th international conference on software e ngineering, 2018, pp. 256–266

    2018

  50. [58]

    Work pract ices and challenges in pull-based development: The contributor’s p erspective,

    G. Gousios, M.-A. Storey, and A. Bacchelli, “Work pract ices and challenges in pull-based development: The contributor’s p erspective,” in Proceedings of the 38th international conference on softwa re engi- neering, 2016, pp. 285–296

    2016

  51. [59]

    Influence of socia l and technical factors for evaluating contribution in github,

    J. Tsay, L. Dabbish, and J. Herbsleb, “Influence of socia l and technical factors for evaluating contribution in github,” in Proceedings of the 36th international conference on Software engineering , 2014, pp. 356–366

    2014

  52. [60]

    The promises and perils of mining github,

    E. Kalliamvakou, G. Gousios, K. Blincoe, L. Singer, D. M . German, and D. Damian, “The promises and perils of mining github,” in Proceedings of the 11th working conference on mining software repositor ies, 2014, pp. 92–101

    2014

  53. [61]

    Supplementary materials for the study on psychologic al safety sustains participation in pull-based open source projects,

    “Supplementary materials for the study on psychologic al safety sustains participation in pull-based open source projects,” 2025. [ Online]. Available: https://figshare.com/s/567fcbccc4fd4b75007 d

    2025

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.