Aptitude Test vs IQ Test: A Distinction That Mostly Collapsed

What the distinction was supposed to be

The classical taxonomy is tidy, which is the first sign that something is wrong with it.

The trouble is the second column. For an aptitude test to measure potential rather than learning, it would need items nobody could have learned to answer. No such items exist. Every question presupposes a language, a symbol system, a familiarity with being tested, and some prior encounter with the kind of material it presents.

The College Board conceded the point

The most famous aptitude test in the world stopped calling itself one, and said why.

The SAT was introduced as the Scholastic Aptitude Test. In 1990 a commission concluded that the test could no longer accurately be described as a test of aptitude. In 1993 the College Board changed the name.

…to correct the impression among some people that the SAT measures something that is innate and impervious to change regardless of effort or instruction.— The president of the College Board, on the 1993 renaming

The replacement, Scholastic Assessment Test, did not survive contact with the language. As the historian of education Diane Ravitch observed, calling it the Scholastic Assessment Test is like calling it the Scholastic Test Test. In 1997 the College Board announced that the letters were simply a trademark and no longer stood for anything at all.

What the distinction does still track

The philosophical distinction collapsed. A practical one remains, and it is worth stating precisely because it is the useful part.

The reference group line is the one that trips people up. An aptitude test may tell you that you scored in the 80th percentile — of people applying for that job. An IQ test tells you where you fall among people generally. The two numbers are not comparable, and only one of them requires a standardisation sample.

Most aptitude tests are measuring general ability anyway

Here is the finding that dissolves what remains of the distinction.

Since Spearman's discovery in 1904 that performance on diverse mental tasks correlates positively, it has been repeatedly confirmed that specific-ability tests are heavily saturated with general ability. Give someone a mechanical reasoning test, a numerical reasoning test and a verbal reasoning test, and their scores will correlate substantially, because a large share of what each measures is g.

The Armed Forces Qualification Test, derived from the ASVAB and used to determine military enlistment eligibility, is a clear case: it is composed of specific subtests and functions as a close approximation to a general ability measure.

If you are being tested for a job

Two things are worth knowing, and neither is what the test-preparation industry tells you.

First, the score you obtain is not an IQ, and should not be read as one. Employment aptitude batteries are normed against applicant pools rather than the population, are usually unproctored, and are constructed to rank candidates rather than to locate them on a population distribution.

Second, cognitive ability tests no longer sit at the top of the predictive hierarchy. In the corrected meta-analytic estimates published by Sackett and colleagues in 2022, structured interviews predict job performance at .42 and biodata at .38, both above cognitive ability's .31. The old ranking, in which mental ability dominated every other method, rested on a statistical correction that had been systematically misapplied for decades.

Myths and facts