Measuring personality

Naming a measure is not describing it. Saying personality is measured with a personality test, or a named questionnaire, names something and stops. The mark needs what the measure does: it asks fixed questions about behaviour, scores them by a set system, and places the person on a scale or in a type. Self-test: does your answer contain a verb — asks, scores, categorises?

A scale is self-report, so answers bend to social desirability — people tick what looks good. Many still score zero: AO3 cannot be awarded without AO2, so a weakness that never touches the scenario earns nothing. Apply it — because the questionnaire was part of a job application, Marek had reason to tick the calm answers, so his score reflects the job not the person.

On strength/weakness items candidates describe instead of explain: it is standardised names the strength and stops. Explaining has three moves — the point (same questions and key), the justification (the result does not depend on who marks it), and the consequence (more reliable, and scores compare fairly). Identification without justification sits in the bottom band.

Asked how a measure could be improved, candidates write another criticism. An improvement is a change that fixes the problem: social desirability, so collect answers anonymously; a single score may not be typical, so re-test another day; closed questions, so add an open one. An improvement uses a verb like add, change or collect — not because it is unreliable.

Asked to describe the difference between a scale and a type, candidates write one definition, then the other, and stop. Two definitions side by side score once. The second mark comes from a connective — whereas, however: a scale reports how much of a trait, on a continuum, whereas a type uses a cut-off to place them in one category, a label with no middle.

A score is a result — what the person did once — not a conclusion. A high score shows only that the person answered a particular way once. A conclusion needs two more things: something to compare it with, and a reason to believe the answers reflect behaviour, not intention. Asked what the measure shows, describe the score; asked what you conclude, say what you infer.

Scoring highly for extraversion and being an extravert are not the same statement. A scale reports how much — a continuum, so a middling score is real. A type reports which group — a cut-off, no middle. So a person just past the cut-off counts as identical to one far inside, and different from one just across the line: a real weakness, not just types are too simple.

Two number-beliefs cost marks. It gives a number, so it is objective: the scoring is objective (a fixed key means two markers agree), but the answers are not — the person's own judgement, so two who behave alike can answer differently. So consistency is better established than validity. A high score does not prove the trait — a score is a result, not a conclusion.