COGNITIVE TRAPS: RECOGNIZE THE FAILURE MODE
ALL YEARS | PAGE 78B - ANCHORING -> CLOSURE -> FRAMING -> PROBABILITY ERROR -> SEARCH FAILURE ->
B. 10 HIGH-FREQUENCY COGNITIVE TRAPS
| Trap | What it looks like | Countermove |
|---|---|---|
| Anchoring | Initial impression remains dominant despite later discordant data. | Restate the case using only current evidence; ask what finding the anchor cannot explain. |
| Premature closure Confirmation bias | The search stops once one plausible diagnosis is found. You seek/support data for the favored diagnosis and discount contradiction. | Require a dangerous alternative and one competing explanation before closure. Actively search for disconfirming evidence and the strongest rival hypothesis. |
| Availability Representativeness | A recent or memorable case makes a diagnosis feel more likely than it is. Atypical features make a real diagnosis feel impossible because it does not match the prototype. | Reset to prevalence + patient-specific discriminators. Use criteria, mechanism and probability; do not demand textbook appearance. |
| Search satisficing | Finding one abnormality ends the search even when the whole presentation is unexplained. | Ask whether the finding explains every major feature; continue if residual discordance remains. |
| Framing effect | Prior labels, demographics, handoff wording or emotionally charged information shape the conclusion. | Rebuild the problem representation from raw data before accepting the frame. |
| Base-rate neglect Overconfidence | Rare diagnoses are overweighted because their pattern seems exciting or elegant. Confidence exceeds evidence; uncertainty and alternatives are not made explicit. | Start with prior probability, then update with likelihood-changing evidence. State confidence level, unresolved evidence and what would trigger reconsideration. |
| Diagnostic momentum / blind obedience | A diagnosis or test interpretation is inherited because others already accepted it. | Independently reconstruct the case; verify whether the evidence still supports the label. |
TOP 10 DIAGNOSTIC-REASONING CONCEPT - 4: ANCHORING BECOMES DANGEROUS WHEN NEW DATA NO LONGER FIT
| DEEP EXPLANATION | WHY STUDENTS MISS IT | EASIEST CLUE |
|---|---|---|
| Anchoring is efficient when the initial pattern is strong, but it becomes unsafe when contradictory findings, trajectory or | Changing a diagnosis feels like admitting the first judgment was wrong, so learners reinterpret new data to preserve the anchor. | Ask: What is the strongest finding my current diagnosis does not explain? |
| treatment response accumulate and the model is not reopened. | ||
| TOP 10 DIAGNOSTIC-REASONING CONCEPT - 5: CONFIRMATION AND SEARCH SATISFICING CHANGE WHAT YOU LOOK FOR | ||
| DEEP EXPLANATION | WHY STUDENTS MISS IT | EASIEST CLUE |
| Reasoning can fail before interpretation: you may stop collecting useful evidence or preferentially collect evidence that | Because the chosen diagnosis feels coherent, additional search feels unnecessary or inefficient. | Before stopping, identify one observation that would make you abandon the leading diagnosis. |
| favors the current hypothesis. | ||
| TOP 10 DIAGNOSTIC-REASONING CONCEPT - 6: DEBIASING REQUIRES A PROCESS, NOT A VOCABULARY LIST | ||
| DEEP EXPLANATION | WHY STUDENTS MISS IT | EASIEST CLUE |
| Biases are difficult to eliminate by awareness alone. More reliable approaches change the workflow: diagnostic timeout, | Learners memorize bias names for exams but do not build a repeatable interrupt into real reasoning. | Use the same forcing question every time uncertainty is high: “What else could this be, and what would prove me wrong?” |
| rival hypothesis, probability check, second perspective, patient input, explicit follow-up and feedback. 2026/27 diagnostic-reasoning expanded edition - educational use; follow school requirements, supervising clinicians and current clinical guidance for patient care |