Are a Static Analysis Tool Study's Findings Static? A Replication
Bibliographic record
Abstract
In 2017, Edwards et al. studied a large corpus of Java programs collected through an automated submission and assessment system that integrated static analysis feedback. They found that errors reported were most commonly related to formatting, but that the frequency of errors they categorized as "Coding Flaws" correlated with program correctness grades. They argued that static analysis feedback could detect problems relating to code correctness and could therefore be useful beyond evaluating conformance to style rules, but that students may overlook non-cosmetic error messages because of the relative volume of formatting errors. In this paper we perform a conceptual replication of the Edwards et al. study with 1270 CS1 students learning Python. We confirm that almost a decade later and even after being instructed to use the auto-formatting options within their IDE, students still encounter mostly formatting errors when using a static analysis tool. We find that the second- most common category of errors detected are "Coding Flaws", and, like Edwards et al., that the frequency of coding flaws identified by the static analysis tool correlates to program correctness. When we examine trends based on levels of prior programming experience, we find that all students tend to make more formatting errors than other kinds of errors, but that students with no prior programming experience have more errors reported across all error categories.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.165 | 0.542 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.004 |
| Bibliometrics | 0.007 | 0.005 |
| Science and technology studies | 0.003 | 0.006 |
| Scholarly communication | 0.005 | 0.009 |
| Open science | 0.006 | 0.006 |
| Research integrity | 0.003 | 0.003 |
| Insufficient payload (model declined to judge) | 0.007 | 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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".