In Reply to Croskerry and to Patel and Bergl
Bibliographic record
Abstract
Croskerry claims that we “deny [the] existence” of cognitive biases. Not so. Our concern is not whether cognitive biases exist but, rather, to understand “the relative contribution of heuristics and cognitive biases versus that of knowledge deficits in clinical reasoning errors.”1 While Croskerry concludes that “several studies have found cognitive failings more likely due to how physicians think rather than what they know, and most clinicians confirm this is what happens in practice,” we are more circumspect. As we described, inducing bias in experimental settings does not prove that errors in practice are caused by bias, or that it would be possible to overcome these biases. And retrospective reviews linking bias to errors are subject to a bias of their own—hindsight.2 Croskerry is concerned that “the potential of cognitive bias mitigation … strategies is … minimized,” and cites a recent systematic review.3 The review—published after our article was accepted—listed six educational interventions directed at learning cognitive biases. Four looked at impact on diagnostic accuracy and found no effect; two others used a self-report thinking inventory. Three studies were classified as “cognitive forcing strategies” but never mentioned cognitive biases. Consistent with our findings, this review found no evidence that focusing on identifying and overcoming cognitive bias reduces diagnostic errors. Instead, the evidence suggests that physicians with more knowledge commit fewer errors, and thus that our curricular efforts should be directed toward strategies that improve knowledge acquisition and application. Patel and Bergl elaborate on this point, particularly with respect to Type 2, analytical thinking. We agree; in fact, we believe that the traditional approach to learning diagnostic reasoning in clinical settings is, at best, inefficient. Woods et al4 have shown the benefit of understanding basic science. We could also use contemporary education strategies such as interleaved practice to greatly enhance the efficiency of Type 1 learning. Such strategies are likely to have far greater yield than a misguided attempt to teach students cognitive biases. Geoffrey Norman, PhDProfessor emeritus, Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada; [email protected] Jonathan Sherbino, MDAssociate professor, Department of Medicine, McMaster University, Hamilton, Ontario, Canada. Jonathan S. Ilgen, MDAssociate professor, Department of Medicine, University of Washington School of Medicine, Seattle, Washington. Sandra D. Monteiro, PhDAssistant professor, Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada.
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 | 0.016 | 0.149 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.004 | 0.003 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.003 | 0.005 |
| Scholarly communication | 0.007 | 0.014 |
| Open science | 0.005 | 0.006 |
| Research integrity | 0.042 | 0.066 |
| Insufficient payload (model declined to judge) | 0.013 | 0.010 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".