2.B.5
Base rates, development, culture and the traps of clinical judgment
ASPPB's competency statement 2.B.5: Integrate relevant factors (e.g., base rates, group differences, cultural biases and differences, heuristics, typical growth and development, risk and protective factors) when interpreting assessment findings and making clinical decisions
Assessment findings never speak for themselves. Their meaning depends on how common a condition is in the setting, what is typical for the person's age and cultural context, and which risk and protective factors are present. Clinicians also rely on mental shortcuts that save time but produce predictable errors, so sound interpretation brings these factors in deliberately and checks the shortcuts.
Key ideas
- Base rates
- The base rate is how often a condition occurs in the population being assessed. Meehl and Rosen showed that when a condition is rare, decisions from a test sign or cutting score can produce more errors than simply predicting from the base rate.
- Heuristics
- Tversky and Kahneman described representativeness (judging by resemblance to a typical case), availability (judging frequency by how easily examples come to mind) and anchoring (adjusting too little from a starting value): economical shortcuts that lead to systematic, predictable errors.
- Debiasing
- Strategies against cognitive errors include considering alternative explanations, seeking evidence that could disconfirm a first impression and metacognition, stepping back to examine one's own reasoning, which Croskerry names as a principal strategy.
- Development and culture
- Behavior is judged against what is typical for the person's age and cultural context. DSM-5-TR provides the Cultural Formulation Interview and describes cultural concepts of distress, so that culturally expected experiences are not mistaken for disorder.
- Risk and protective factors
- Risk factors raise and protective factors lower the probability of an outcome, but for rare outcomes even established risk factors predict poorly: a meta-analysis of 365 studies found that risk factors predicted suicidal thoughts and behaviors only slightly better than chance.
Worked example
Mr. Quinlan, 47, is in treatment for depression after a divorce and scores in the high range on a suicide risk checklist. He lives with his teenage daughter, whom he cares for, and he has attended every session.
- Suicide is rare even among clinical patients, and risk factors predict it only slightly better than chance, so a high checklist score alone mostly produces false positives if read as a prediction.
- Instead of predicting, the psychologist formulates: what drives his distress now, what warning signs would signal rising danger, and which protective factors, such as his daughter and his engagement in care, can be strengthened.
- To avoid anchoring on the checklist total, the psychologist gathers current information directly from him and, with his consent, from his sister.
- The plan targets factors that can change and sets a time to review, since risk changes over time.
Use base rates and risk factors to guide attention and planning, not to issue predictions about one person.
Check yourself
In one week a psychologist evaluates two clients with a rare dissociative disorder. Over the next month she suspects the disorder in many new clients and overestimates how common it is. Which heuristic does this illustrate?
Sources
- Meehl and Rosen (1955), Antecedent probability and the efficiency of psychometric signs, patterns, or cutting scores, Psychological Bulletin (PMID 14371890), Psychological Bulletin 52, 194-216
- Tversky and Kahneman (1974), Judgment under uncertainty: Heuristics and biases, Science (PMID 17835457), Science 185, 1124-1131 (representativeness, availability, anchoring)
- Croskerry (2003), The importance of cognitive errors in diagnosis and strategies to minimize them, Academic Medicine (PMID 12915363), Academic Medicine 78, 775-780 (cognitive debiasing; metacognition)
- Franklin, Ribeiro, Fox, Bentley, Kleiman, Huang, Musacchio, Jaroszewski, Chang and Nock (2017), Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research, Psychological Bulletin (PMID 27841450), Psychological Bulletin 143, 187-232 (365 studies; prediction only slightly better than chance)
- Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR; American Psychiatric Association, 2022), Section III, Culture and Psychiatric Diagnosis (Cultural Formulation Interview; cultural concepts of distress)