Enhancing retrieval of best evidence for health care from bibliographic databases: calibration of the hand search of the literature.
Bibliographic record
Abstract
BACKGROUND: Medical practitioners have unmet information needs. Health care research dissemination suffers from both "supply" and "demand" problems. One possible solution is to develop methodologic search filters ("hedges") to improve the retrieval of clinically relevant and scientifically sound study reports from bibliographic databases. To develop and test such filters a hand search of the literature was required to determine directly which articles should be retrieved, and which not retrieved. OBJECTIVE: To determine the extent to which 6 research associates can agree on the classification of articles according to explicit research criteria when hand searching the literature. DESIGN: Blinded, inter-rater reliability study. SETTING: Health Information Research Unit, McMaster University, Hamilton, Ontario, Canada. PARTICIPANTS: 6 research associates with extensive training and experience in research methods for health care research. MAIN OUTCOME MEASURE: Inter-rater reliability measured using the kappa statistic for multiple raters. RESULTS: After one year of intensive calibration exercises research staff were able to attain a level of agreement at least 80% greater than that expected by chance (kappa statistic) for all classes of articles. CONCLUSION: With extensive training multiple raters are able to attain a high level of agreement when classifying articles in a hand search of the literature.
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 | 0.688 | 0.907 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.007 | 0.004 |
| Bibliometrics | 0.038 | 0.020 |
| Science and technology studies | 0.004 | 0.005 |
| Scholarly communication | 0.008 | 0.014 |
| Open science | 0.005 | 0.013 |
| Research integrity | 0.003 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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; the direct Gemma label and the distilled Codex classifier 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".