Definition
Psychometrics is the branch of psychology devoted to the theory and technique of measuring mental attributes that cannot be observed directly — intelligence, personality traits, aptitudes, attitudes, and clinical symptoms. Where physics measures length with a ruler, psychometrics constructs instruments that infer the magnitude of an unseen psychological quantity from a person's responses to a structured set of items. The discipline is as much about how to build, evaluate, and interpret those instruments as about the constructs themselves.
A psychometric test is judged by two technical standards: reliability, meaning the test produces consistent scores across time and across equivalent items, and validity, meaning the test actually measures what it claims to measure rather than some adjacent quantity. Standardization — administering and scoring the test under uniform conditions, and comparing scores against a representative normative sample — is what allows individual results to be interpreted meaningfully.
Why it matters
How it works
The construction of a psychometric instrument begins with a clear operational definition of the target construct — what, precisely, is meant by "extraversion" or "verbal reasoning." From that definition, item writers generate a large pool of candidate questions and pilot them on a sample. Statistical analyses then trim the pool: items that everyone passes or fails reveal nothing about individual differences, items that correlate poorly with the rest of the scale probably measure something else, and items that show different difficulty patterns across demographic groups raise fairness concerns that must be examined.
The surviving items are administered to a large, representative normative sample to establish the score distribution against which future test-takers will be compared. Reliability is estimated using techniques such as test-retest correlation and internal consistency coefficients. Validity is built up over time through accumulated evidence: scores should correlate with related measures, predict relevant outcomes, and show the right pattern of group differences where theory predicts one. No single study establishes validity; the case is made cumulatively across studies, samples, and contexts.