Best practices for building and curating databases for comparative analyses
Bibliographic record
Abstract
Comparative analyses have a long history of macro-ecological and -evolutionary approaches to understand structure, function, mechanism and constraint. As the pace of science accelerates, there is ever-increasing access to diverse types of data and open access databases that are enabling and inspiring new research. Whether conducting a species-level trait-based analysis or a formal meta-analysis of study effect sizes, comparative approaches share a common reliance on reliable, carefully curated databases. Unlike many scientific endeavors, building a database is a process that many researchers undertake infrequently and in which we are not formally trained. This Commentary provides an introduction to building databases for comparative analyses and highlights challenges and solutions that the authors of this Commentary have faced in their own experiences. We focus on four major tips: (1) carefully strategizing the literature search; (2) structuring databases for multiple use; (3) establishing version control within (and beyond) your study; and (4) the importance of making databases accessible. We highlight how one's approach to these tasks often depends on the goal of the study and the nature of the data. Finally, we assert that the curation of single-question databases has several disadvantages: it limits the possibility of using databases for multiple purposes and decreases efficiency due to independent researchers repeatedly sifting through large volumes of raw information. We argue that curating databases that are broader than one research question can provide a large return on investment, and that research fields could increase efficiency if community curation of databases was established.
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 | 0.398 | 0.728 |
| Meta-epidemiology (narrow) | 0.003 | 0.006 |
| Meta-epidemiology (broad) | 0.006 | 0.009 |
| Bibliometrics | 0.053 | 0.047 |
| Science and technology studies | 0.010 | 0.015 |
| Scholarly communication | 0.031 | 0.034 |
| Open science | 0.021 | 0.023 |
| Research integrity | 0.013 | 0.023 |
| Insufficient payload (model declined to judge) | 0.021 | 0.023 |
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".