"@alex, this fixes #9": Analysis of Referencing Patterns in Pull Request Discussions
Bibliographic record
Abstract
This publication consists of a dataset of 7k references manually identified in 450 Pull request (PR) discussion threads sampled from GitHub in CSV format. In addition to the dataset, it also contains R code files which were written to analyze this dataset statistically. This dataset is released under the research, which is accepted for publication at CSCW 2021 conference, titled "@alex, this fixes #9": Analysis of Referencing Patterns in Pull Request Discussions". <strong>Paper Abstract</strong> Pull Requests (PRs) are a frequently used method for proposing changes to source code repositories. When discussing proposed changes in a PR discussion, stakeholders often reference a wide variety of information objects for establishing shared awareness and common ground. Previous work has not considered how referential behavior impacts collaborative software development via PRs. This knowledge gap is the major barrier in evaluating the current support for referencing in PRs and improving them. We conducted an explorative analysis of ~7K references, collected from 450 public PRs on GitHub, and constructed taxonomies of referent types and expressions. Using our annotated dataset, we identified several patterns in the use of references. Referencing source code elements was prevalent but the authoring interface lacks support for it. Three classes of contextual factors influence referencing behaviors: referent type, discussion thread, and project attributes. Referencing patterns may indicate PR outcomes (e.g., merged PRs frequently reference issues, users, and tests). We conclude with design implications to support more effective referencing in PR discussion interfaces.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.111 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".