What do intelligence test scores measure?

For: Readers seeking measurement foundations before debating intelligence claims.

A test score is the outcome of particular tasks, scoring rules and comparison procedures. A theory of intelligence is an explanation of why performances relate as they do. A causal account of differences between people is a further claim. This route keeps those three levels separate so that neither confidence in a score nor scepticism about it becomes a shortcut around measurement.

The MIT intelligence session explicitly discusses the strengths and weaknesses of IQ testing, models of intelligence and biological and environmental interpretations. Its 2011 examples offer a starting curriculum, not a current appraisal of every intervention or group comparison. The Yale overview gives a broader comparison with personality, whose measures ask a different kind of question.

The inspected measurement section of Noba’s intelligence chapter describes the historical development of standardised testing and instruments that sample several performances. Use that history to ask what was measured, how scores were compared and why a particular collection of tasks was chosen. Do not let a familiar abbreviation stand in for those decisions.

The NPTEL selections retain a theory-to-testing sequence and then offer a separate two-part introduction. As you compare them, make a table of the construct, task and proposed interpretation. Ask which inference needs reliability evidence, which needs validity evidence and which would require a design capable of addressing causation.

The exercise uses invented tasks rather than an online score. It helps expose what would be required before calling a small collection of items a meaningful scale. Consistent scoring is useful, but does not by itself show that a measure supports every intended use. A relation between two scores is also not a complete explanation of either.

This guide does not assess an individual, settle genetic explanations or rank groups. Quantitative policy and group-difference claims in older materials were not adopted into its argument. The next useful comparison is personality measurement, where ratings of characteristic behaviour introduce different questions about context and interpretation.

Next step

Compare intelligence testing with the personality-dimensions guide.

Limits of this route

The guide uses introductory measurement explanations and older teaching, not a current systematic review. Quantitative group-difference, intervention and genetic claims were not adopted. A fictional task set is not an intelligence test. Source checks establish scope and selection usefulness; they do not claim full-video viewing, human review or scientific validation.