The Table 2 Fallacy and Overfitting: A Persistent Problem in Contemporary Research?
Bibliographic record
Abstract
The "Table 2 fallacy" represents a common methodological error in medical research, characterized by indiscriminate statistical adjustment for multiple variables without considering their causal nature. This article examines the theoretical foundations of the problem, distinguishing between studies with descriptive, predictive, and explanatory objectives, and emphasizing how the research purpose should determine the adjustment strategy. We highlight the fundamental role of Directed Acyclic Graphs (DAGs) in correctly identifying confounding, mediating, and colliding variables, thus avoiding overadjustment and resulting biases. To illustrate these considerations, we present two practical examples: the relationship between obesity and colorectal cancer, and between coffee consumption and breast cancer. In the first case, we demonstrate how adjustment for intestinal dysbiosis (a mediator) can attenuate the association between obesity and colorectal cancer, reducing the adjusted relative risk from 1.78 (95% CI: 1.20–2.65) to 1.49 (95% CI: 0.97–2.29) and eliminating statistical significance (p=0.072). In the second example, we show how including insomnia (a collider) in the model can create artificial associations between coffee consumption and breast cancer, dramatically increasing the adjusted relative risk to 1.94 (95% CI: 1.34-2.81) with high statistical significance (p<0.001), when a correctly specified model shows no such association. We conclude that, in explanatory studies, it is essential to develop causal reasoning prior to statistical analysis, using DAGs to guide the selection of adjustment variables. This rigorous methodological approach prevents both the dilution of real causal effects and the generation of spurious associations, increasing the internal validity of epidemiological findings and their utility for clinical decision-making.
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 | 0.458 | 0.733 |
| Meta-epidemiology (narrow) | 0.002 | 0.003 |
| Meta-epidemiology (broad) | 0.007 | 0.005 |
| Bibliometrics | 0.007 | 0.011 |
| Science and technology studies | 0.003 | 0.021 |
| Scholarly communication | 0.009 | 0.013 |
| Open science | 0.008 | 0.006 |
| Research integrity | 0.007 | 0.014 |
| Insufficient payload (model declined to judge) | 0.010 | 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; 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".