Myths and famous claims
Are IQ tests biased?
What is documented
It depends on what “biased” means. Researchers separate differences in average scores from predictive bias, which asks whether a test predicts later school or job performance less accurately for some test takers. The 1996 APA task force concluded that, as predictors of future performance, the tests it reviewed do not seem to be biased.
Want your own number? The test is 25 visual puzzles, about 12 minutes, no sign-up.
The myth
The claim is that IQ tests are simply biased: that they reward a particular background so heavily that a score says nothing fair about the person who earned it. Like many myths, it mixes a real concern with an overstatement, so the first step is to be precise about the word.
What "bias" means to test researchers
The 1996 American Psychological Association task force report "Intelligence: Knowns and Unknowns" separated two ideas that everyday use runs together (Neisser et al. 1996):
- Outcome bias refers to differences in average scores. A difference in averages, on its own, does not tell you whether the test is measuring unfairly.
- Predictive bias asks whether a test predicts later school and college performance differently for different test takers. If the same score forecast different outcomes depending on who took the test, the test would be biased in this technical sense.
The short verdict
On predictive bias, the task force concluded that, as predictors of future performance, the tests it reviewed do not seem to be biased (Neisser et al. 1996). That is a specific, tested claim, and it is the one most researchers mean when they discuss bias. It does not mean a test is perfect or measures everything that matters.
Try one · Pattern Recognition
Which tile comes next?
Tap the option you think is right, or show the answer.
Show the answer
The missing tile shows a very large hatched arrow pointing right (option 1).
How to solve it
- Compare neighboring tiles and name what changes each time.
- From tile to tile, the arrow turns a quarter turn counterclockwise.
- The fill cycles hatched, solid and outline, then starts again.
- Everything else stays the same from tile to tile.
- The last tile shows a very large outlined arrow pointing down. Apply each change once more.
- So the missing tile shows a very large hatched arrow pointing right.
That was 1 puzzle. The full test has 25.
The real limits
Language
A test given in a language you do not know well partly measures that language. The 1994 mainstream statement noted that people who do not understand English well can be given a nonverbal test or one in their native language (Gottfredson, 1997).
Familiarity and schooling
Experience with tests changes scores. Retaking cognitive ability tests raises scores by about 4 IQ points on average (Hausknecht et al., 2007), and each additional year of education raises IQ by about 1 to 5 points (Ritchie & Tucker-Drob, 2018). Scores therefore reflect opportunity to learn, not only some fixed potential.
Old norms
An IQ compares you with a norm group of your own age (Wikipedia: Intelligence quotient). Because of the Flynn effect, people who score 100 on a newly normed test would probably average about 106 on a version normed 20 years earlier (Neisser et al., 1996). Out-of-date norms make scores look higher than they are.
Culture-reduced, not culture-free
Test makers have tried to reduce reliance on words and schooling. Raven's Progressive Matrices are non-verbal and widely used as a measure of fluid, culture-reduced reasoning (Wikipedia: Raven's Progressive Matrices). The Cattell Culture Fair test was designed to measure intelligence with minimal influence from verbal fluency and educational level (Buros: CFIT (store)). Reduced is the honest word: familiarity with puzzles and testing still counts.
What is actually true
In the technical sense researchers test, the task force found that the tests it reviewed predicted future performance without apparent bias. In the everyday sense, any single score is shaped by language, schooling, practice and norms, and should be read with those in mind. IQ Octopus uses visual puzzles generated fresh for each attempt, which reduces reliance on vocabulary, but it is not a clinical instrument and familiarity still matters.
Read more in the guide to culture-fair IQ tests, the page on Raven's Progressive Matrices and why IQ scores differ between tests. Try a culture-fair practice test, or see the related myth that IQ tests measure nothing.
Sources
- Neisser, U., Boodoo, G., Bouchard, T. J., Jr., Boykin, A. W., Brody, N., Ceci, S. J., Halpern, D. F., Loehlin, J. C., Perloff, R., Sternberg, R. J., & Urbina, S. (1996). Intelligence: Knowns and Unknowns. American Psychologist, 51(2), 77-101 (read via Wayback Machine copy of UConn PDF) www.gifted.uconn.edu/siegle/research/Correlation/Intelligence.pdf
- Gottfredson, L. S. (1997). Mainstream science on intelligence: An editorial with 52 signatories, history, and bibliography. Intelligence, 24(1), 13-23. doi.org/10.1016/S0160-2896(97)90011-8
- Raven's Progressive Matrices. Wikipedia. en.wikipedia.org/wiki/Raven%27s_Progressive_Matrices
- Buros Center for Testing. Culture Fair Intelligence Test, marketplace listing. marketplace.unl.edu/buros/culture-fair-intelligence-test.html
- Hausknecht, J. P., Halpert, J. A., Di Paolo, N. T., & Moriarty Gerrard, M. O. (2007). Journal of Applied Psychology, 92(2), 373-385. doi.org/10.1037/0021-9010.92.2.373
- Neisser, U., Boodoo, G., Bouchard, T. J., Jr., Boykin, A. W., Brody, N., Ceci, S. J., Halpern, D. F., Loehlin, J. C., Perloff, R., Sternberg, R. J., & Urbina, S. (1996). Intelligence: Knowns and unknowns. American Psychologist, 51(2), 77-101. doi.org/10.1037/0003-066X.51.2.77
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358-1369. doi.org/10.1177/0956797618774253
- Intelligence quotient. Wikipedia. en.wikipedia.org/wiki/Intelligence_quotient