Psychology common mistakes
Psychology · Understand the traps and how to avoid them.
Reading a study scenario and answering from the overall 'gist' of the design (e.g., 'it's an experiment, so it shows causation') without checking whether participants were randomly assigned and whether the independent variable was truly manipulated.
Exam stems often describe quasi-experimental or correlational designs using causal-sounding verbs ('researchers tested whether X improved Y'). Under time pressure, candidates pattern-match on the topic rather than parsing the design features that determine the valid inference.
What to do instead
Before choosing, underline three features in the stem: (1) was the IV manipulated or merely measured, (2) random assignment or intact groups, (3) was there a control/comparison condition. If the IV was measured, eliminate every option containing causal language ('causes', 'leads to', 'produces') and keep only associational phrasing.
Selecting the option that describes the most famous or most-studied phenomenon in a category rather than the one that matches the specific defining features in the stem (e.g., choosing 'classical conditioning' for any learning scenario, or 'bystander effect' for any group scenario).
High-frequency concepts are over-rehearsed and become attractors. The stem's discriminating details (presence of reinforcement, number of witnesses, timing of the outcome) get skimmed because the topic label feels familiar.
What to do instead
Cover the answer options, restate the stem in one sentence using only its concrete details, then uncover options. Explicitly name the one feature that separates the correct concept from its nearest competitor (e.g., reinforcement vs. punishment; availability vs. representativeness heuristic) and verify that feature is present in the stem.
Confusing diagnostic criteria with associated features or risk factors, and applying a diagnosis to a case that does not meet duration, impairment, or exclusion requirements (e.g., labeling a two-week low mood as major depressive disorder, or diagnosing a grief reaction as depression without checking the bereavement/exclusion rules).
Textbook descriptions blend criteria, prevalence, and correlates into a single narrative, so candidates remember the whole cluster as 'the disorder'. Duration thresholds and functional-impairment requirements are easy to blur under recall pressure.
What to do instead
For each major disorder, memorize a compact checklist: core symptoms, minimum duration, required impairment, and key exclusions. In case vignettes, write the duration and impairment status in the margin first; if either is missing or insufficient, the diagnosis is premature and the correct answer is usually 'no diagnosis' or a milder/other specified category.
Mixing up neurotransmitters, their primary functions, and the drugs that act on them (e.g., attributing schizophrenia primarily to GABA, or confusing SSRIs with dopamine antagonists), and confusing brain structures with their functions.
Biological content is memorization-heavy with many similar-sounding names, and candidates study it as lists rather than as systems. Exam items often test one-step inferences ('a drug that blocks reuptake of X will...') that require functional understanding, not just name recognition.
What to do instead
Build a two-column system map: transmitter/structure to core function to clinical correlate to drug class that alters it. Rehearse by generating the inference chain aloud ('blocking dopamine D2 in mesolimbic pathway reduces positive symptoms but increases prolactin because tuberoinfundibular...'). Test yourself with 'what happens if' questions, not definition matching.
Treating psychometric concepts as interchangeable: reporting reliability as validity, assuming a test is valid because it is reliable, confusing standard error of measurement with standard deviation, or misreading a correlation coefficient's direction or magnitude.
Reliability and validity are taught together and share vocabulary ('consistency', 'accuracy'), and the formulas for SEM and confidence intervals are memorized without the underlying logic. Candidates also over-interpret small correlations as strong effects.
What to do instead
Anchor the distinction in one sentence: reliability bounds validity but never guarantees it. Practice computing SEM from SD and reliability and interpreting a 95% confidence band around an observed score. For any correlation, state direction, magnitude (r-squared as variance accounted for), and whether the design permits causal inference, before looking at options.
Answering treatment questions by matching the disorder to a therapy name without considering evidence level, indication, sequencing, or contraindication (e.g., choosing a specific phobia treatment of psychodynamic therapy, or recommending benzodiazepines as first-line for long-term generalized anxiety).
Therapy orientations are learned as schools of thought rather than as indicated treatments, and candidates default to the approach they know best. Guideline-level distinctions (first-line vs. adjunct, acute vs. maintenance) are under-studied.
What to do instead
Create a disorder-to-first-line-treatment table with columns for psychotherapy, pharmacotherapy, combined, and contraindicated/avoid. Include exposure-based treatment for anxiety disorders, behavioral activation and CBT for depression, and antipsychotics with monitoring requirements. For each, note one reason the alternative is wrong.
Applying a stage theory as a rigid timetable rather than as a sequence with variability: assuming a child who fails a conservation task is 'behind', or attributing a behavior to the wrong Piagetian/Eriksonian stage because of the child's age alone.
Stage labels are memorable and ages are easy to memorize, so candidates answer from the age cue and ignore the described behavior. Modern findings (earlier competence, domain-specific development, cultural variation) are also under-weighted.
What to do instead
Classify by the described behavior first (e.g., conservation, object permanence, imaginary audience, generativity), then check whether the age is consistent. Memorize the defining cognitive accomplishment of each stage and one classic task. Treat ages as modal ranges, and prefer options that acknowledge individual and cultural variation over options that assert fixed timing.
Confusing related social-psychological phenomena that share a setting: conformity vs. obedience vs. compliance, prejudice vs. discrimination vs. stereotype, fundamental attribution error vs. actor-observer bias vs. self-serving bias.
These clusters are taught in the same chapter with overlapping examples, and exam items deliberately use scenarios that could fit two labels. Candidates choose the more familiar term rather than the one matching the causal structure (peer pressure vs. authority command vs. direct request).
What to do instead
For each cluster, write a one-line discriminator keyed to the mechanism: who is the source of influence, is the target's behavior public or private, is the judgment about a group or an individual. Then sort practice vignettes into the cluster by mechanism before naming the term.
Ignoring base rates and misapplying heuristics or Bayesian reasoning in statistics and judgment items, including assuming a highly accurate test means a positive result is probably a true positive.
The intuitive answer conflicts with the normatively correct one, and candidates answer from intuition under time pressure. Base-rate neglect is itself a documented bias, so the error feels correct.
What to do instead
For any screening or diagnostic accuracy item, write out the four-cell table (true positives, false positives, false negatives, true negatives) with a concrete population of 1,000 before answering. For heuristic items, name the heuristic and the specific cue driving the judgment (similarity for representativeness, ease of recall for availability, initial value for anchoring).
Answering memory and cognition items by assuming encoding and retrieval are interchangeable, or that confidence predicts accuracy, leading to wrong choices about encoding specificity, levels of processing, working-memory components, and eyewitness identification.
Everyday experience conflates storing and recalling information, and vivid, confident recall feels trustworthy. The experimental distinctions (shallow vs. deep processing, context-dependent retrieval, phonological loop vs. visuospatial sketchpad) are learned as labels without contrasting examples.
What to do instead
For each memory item, identify the stage being tested (encoding, storage, retrieval) and the manipulation used to test it. Memorize one canonical contrast per concept (deep vs. shallow processing; same vs. different context at recall; verbal vs. spatial dual-task interference). For eyewitness and confidence items, recall that confidence and accuracy dissociate and that lineup instructions affect false identifications.