Guilt-free data reuse
Bibliographic record
Abstract
Existing approaches to ensuring the validity of inferences drawn from data assume a fixed procedure to be performed, selected before the data are examined. Yet the practice of data analysis is an intrinsically interactive and adaptive process: new analyses and hypotheses are proposed after seeing the results of previous ones, parameters are tuned on the basis of obtained results, and datasets are shared and reused. In this work, we initiate a principled study of how to guarantee the validity of statistical inference in adaptive data analysis. We demonstrate new approaches for addressing the challenges of adaptivity that are based on techniques developed in privacy-preserving data analysis.<!-- END_PAGE_1 --> As an application of our techniques we give a simple and practical method for reusing a holdout (or testing) set to validate the accuracy of hypotheses produced adaptively by a learning algorithm operating on a training set.
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
Direct model labels (unvalidated)
Per-model category and study-design labels from the labeling rounds. They are machine output, unvalidated, and the disagreement between models ships as data. No study design here is MEDLINE-validated yet.
| Model arm | Categories | Study design | Confidence |
|---|---|---|---|
| gemma | MetaresearchOpen science Domain: Reproducibility · Genre: Methods About the Canadian research system: no · About a Canadian topic: no | Theoretical or conceptual | low |
| gpt | Open science Domain: not available · Genre: Commentary About the Canadian research system: no · About a Canadian topic: no | Not applicable | low |
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 | 0.137 | 0.454 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.005 | 0.006 |
| Bibliometrics | 0.004 | 0.005 |
| Science and technology studies | 0.005 | 0.022 |
| Scholarly communication | 0.010 | 0.030 |
| Open science | 0.017 | 0.033 |
| Research integrity | 0.007 | 0.014 |
| Insufficient payload (model declined to judge) | 0.007 | 0.004 |
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, unvalidatedLabeled directly by 2 models reading the full record.
The models disagree on parts of this classification; every voice is preserved in the section at the end of the page.
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".