Standardization is the process of defining meaningful scores by comparison with the performance of a pretested group. It also includes the uniform procedures for administering and scoring the test.

When a new intelligence or personality test is developed, it is given to a large, representative sample under carefully controlled conditions. The resulting norms allow any subsequent score to be interpreted relative to that group—whether it is average, above average, or below. Standardization ensures that differences in scores reflect differences in the people tested rather than differences in how the test was given or scored.

Without standardization, a raw score of 42 is uninterpretable. With it, the same score can be located on a percentile rank or standard-score scale. Periodic renorming is required because population performance can shift over time, as seen in the Flynn effect for intelligence tests.

A test becomes interpretable only when its scores are anchored to a defined reference group and its procedures are held constant.