Evaluating Large Language Models for Assisting in Meta-Analysis
Bibliographic record
Abstract
Large language models (LLMs) are receiving increased attention in academia as aids for scientific research due to their superior performance in tasks related to natural language processing and understanding. Meta-analysis, a research method involving extensive text processing to extract and code qualitative and quantitative information from empirical studies, is particularly well-suited to the application of LLMs. In this study, we empirically evaluated the ability of LLMs to perform automatic coding tasks within meta-analytic contexts, using Bing Chat (based on GPT-4.0) and ChatPDF (based on GPT-3.5) as examples. Our findings indicate that Bing Chat outperformed ChatPDF in accurately extracting and coding qualitative information such as publication type, country, and survey methods. However, its performance decreased when handling quantitative data, such as correlation coefficients. We also noted an upward trend in Bing Chat's performance over time. The potential and utility of LLMs in facilitating meta-analysis from a researcher’s perspective are further discussed.
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 | 0.389 | 0.659 |
| Meta-epidemiology (narrow) | 0.007 | 0.004 |
| Meta-epidemiology (broad) | 0.006 | 0.023 |
| Bibliometrics | 0.020 | 0.016 |
| Science and technology studies | 0.003 | 0.002 |
| Scholarly communication | 0.013 | 0.011 |
| Open science | 0.005 | 0.008 |
| Research integrity | 0.004 | 0.006 |
| Insufficient payload (model declined to judge) | 0.005 | 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".