Subjects · Psychology & Behavioral Sciences
Psychology & Behavioral Sciences: the methods that matter most here
Know which of the many methods in this guide actually earn their place in psychology, and which popular ones quietly make the subject's own problems worse.
What you'll be able to do: Know which of the many methods in this guide actually earn their place in psychology, and which popular ones quietly make the subject's own problems worse.
What makes psychology hard, specifically
The methods below are chosen against these difficulties, not against a general idea of studying. Psychology is hard in ways that are particular to it:
Effect size and statistical significance get taught as one idea and examined as two. "Significant" answers "is this probably not noise." It does not answer "does this matter," and a huge fraction of both textbook and popular coverage states the first while implying the second.
The replication crisis overturned some famous findings and left others standing, and the two get mixed up. Ego depletion and power posing failed badly in large preregistered replications. Milgram's obedience effect and loss aversion have held up comparatively well. Growth mindset and stereotype threat sit in contested middle ground, real, but smaller and more conditional than the popular version. Treating the crisis as "psychology is unreliable" is as wrong as treating it as "nothing changed."
The field's foundational data is disproportionately WEIRD: Western, educated, industrialised, rich, democratic, because it came overwhelmingly from the undergraduates its researchers had cheapest access to. That's a biographical fact about who runs psychology studies and where, not a design decision anyone chose, and its consequences show up constantly: a finding about "people" is often a finding about a specific, narrow, unrepresentative slice of them.
The leap from a lab study to advice happens in one sentence and almost never gets flagged. A significant result in an artificial task with a small, unrepresentative sample becomes "so do this in your relationship / career / parenting," and the join is invisible unless you're looking for it.
Students learn findings and are examined on methods and evaluation. Being able to state what Loftus and Palmer found is worth little on its own; the marks are in naming the confound, the ecological validity problem, the alternative explanation. This is the opposite emphasis to how most people revise the subject.
Psychology's claims are about the reader, which makes motivated reasoning unusually easy. A finding about willpower, attachment or personality is something you have a stake in believing or disbelieving, in a way a fact about photosynthesis isn't, and an AI trying to be agreeable will hand you exactly the reading you already wanted.
The ten
| # | Method | Why it earns its place here |
|---|---|---|
| 1 | K1 Replication verification drill | The subject's defining danger; checking whether a finding held up is the single most productive habit available |
| 2 | L13 Textbook-versus-literature gap | Textbook lag is worse here than almost anywhere else in this corpus; this targets it directly |
| 3 | H2 Misconception elicitation | A densely documented, stable catalogue of pop-psychology myths exists and is carried by teachers as much as students |
| 4 | M9 Figure reading walkthrough | "Significant" and "big enough to matter" are different questions, and only the figure (the effect size, the confidence interval) answers the second |
| 5 | C22 Forced operationalisation | Nearly every construct in the subject is a word before it's a measurement, and the measurement is where an argument holds or doesn't |
| 6 | D13 Cause-consequence mapping | Correlation dressed as causation, with a plausible story attached, is the field's single most common public-facing error |
| 7 | C7 Assumption surfacing | Every study smuggles in assumptions (that self-report is accurate, that the lab task generalises) and evaluation marks target exactly these |
| 8 | F8 Methodology critique | Coursework asks you to design and defend a study; having your own design attacked before you run it is direct practice for the actual skill examined |
| 9 | M4 Plot my own data | Your own sleep, mood or reaction-time data makes effect size and noise concrete instead of abstract |
| 10 | K5 Bias probing | The subject's claims are about you, which makes framing effects unusually easy to fall for; asking the same question several ways is the direct countermeasure |
Two popular methods that work badly here, and why
Analogy generation (A5), used without stress-testing it (A6). In most subjects an analogy is a scaffold you later discard. In psychology, the popular analogy and the debunked mechanism are frequently the same object. Ask for an analogy for self-control and you will very likely be handed "it's like a muscle that gets tired", which is not a helpful simplification of ego depletion theory, it is ego depletion theory, the specific claim that failed to replicate in the largest preregistered multi-lab test run on it. Ask for an analogy for the mind and you'll get "it's like a computer," a framing the field itself has spent decades complicating. Requesting an analogy without immediately asking where it breaks doesn't simplify the material here, it hands you last century's theory relabelled as an aid to understanding.
Flashcard-style drilling of findings (B13/B14) as the primary study method. This is efficient at what it drills: "Milgram, 1963, roughly 65% went to the maximum shock." It is close to useless against what the subject actually examines, which is your ability to evaluate that finding, the sample, the ecological validity, the ethical critique, the partial replications under tighter constraints. A student who can recite twenty findings fluently and cannot critique one of them has practised the wrong half of the subject, and flashcards make that half feel like progress because recall feels like mastery. Use retrieval practice on methods and evaluation points, not on findings alone.
What this subject's toolkit looks like in practice
A well-run psychology session has a recognisable shape:
- Before opening a topic, ask for its documented misconceptions and its headline studies' replication status.
- For each headline study, get the actual design, sample, IV, DV, operationalisation, not just the conclusion.
- For any causal-sounding claim, force the correlation-to-causation check: what would have to be true, and does the design support it.
- Where the material concerns you personally, ask for the less comfortable reading first.
- Practise the evaluation points, not the findings, against past-paper wording.
That's five methods composed into about forty minutes, and it addresses the subject's two defining risks, inherited overconfidence about famous findings, and under-practice of the evaluation skill that's actually marked, in one pass. The ten articles that follow take one method each and go deep.
The ten articles in this subject
01 The replication verification drill · 02 The textbook-versus-literature gap · 03 Misconception elicitation · 04 Figure reading walkthrough · 05 Forced operationalisation · 06 Correlation-causation mapping · 07 Assumption surfacing · 08 Plot my own data · 09 Methodology critique · 10 Bias probing