Cognitive shadow: A policy capturing tool to support naturalistic decision making
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Policy capturing is an approach to decision analysis using statistical models such as multiple linear regression or machine learning algorithms to approximate the mental models of decision makers. The present work seeks to apply a robust policy capturing technique to functionally mirror expert mental models and create individually-tailored cognitive assistants. The “cognitive shadow” method aims to improve decision quality by recognizing probable errors in cases where the decision maker is diverging from his usual judgmental patterns. The tool actually shadows the decision maker by un-intrusively monitoring the situation and comparing its own decisions to those of the human decision maker, and then provides advisory warnings in case of a mismatch. The support methodology is designed to be minimally intrusive to avoid an increase in cognitive load, either in real-time or off-line dynamic decision making situations. Importantly, user trust is likely to be a key asset since the cognitive shadow is derived from one's own judgments. A use case of the cognitive shadow is described within the context of a maritime threat classification task, using the classic CART decision tree induction algorithm for policy capturing. This approach is deemed applicable to a large variety of domains such as supervisory control, intelligence analysis and surveillance in defence and security, and of particular relevance in high-reliability organizations with low tolerance for error.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.005 |
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 it