From tasks to a standardized score
An assessment contains multiple subtests. A person may identify patterns, manipulate information in memory, solve verbal problems, or work accurately under time limits. Raw totals have little meaning alone, so publishers compare them with a large reference sample.
Performance is compared with people in the same age band. Scaled subtest scores may then be combined into broader index scores and an overall estimate.
Norms, percentiles, and spread
Many current tests use a mean of 100 and a standard deviation of 15. A percentile expresses position: 100 is near the 50th percentile, while 120 is around the 91st percentile under a normal distribution.
Different instruments use different tasks and norms, so two valid tests do not always produce identical scores.
Why interpretation matters
Every score includes measurement uncertainty and should usually be reported with a confidence interval. A trained examiner also considers subtest patterns, language, cultural background, testing conditions, and the purpose of the assessment.
Good measurement depends on standard instructions, suitable norms, test security, reliability, and validity. A number without this context can mislead.
COMMON QUESTIONS
Frequently asked questions
Why are some IQ tasks timed?
Some subtests measure processing efficiency, but many reasoning tasks are untimed or have generous limits. Timing is only one part of an assessment.
Can different IQ tests give different results?
Yes. Content, norms, measurement error, health, and testing conditions can cause reasonable differences.