What does the Flynn effect describe?
The Flynn effect describes rising average scores on IQ tests over the 20th century. Estimates differ by analysis, but the increase was roughly 2 to 3 points per decade.
This is a pattern in test results. It does not mean every person’s reasoning ability rose by the same amount, or that a score alone explains what changed.
Why does the average IQ stay at 100?
The average stays at 100 because IQ tests are normed against comparison groups. Wechsler IQ scores use a mean of 100 and a standard deviation of 15. Test publishers update norms so a score reflects performance relative to people of a similar age taking the current version.
As a result, rising raw performance does not make the average score on a newly normed test rise above 100. A score shows a position within its reference group, rather than a fixed count of correct answers comparable across versions. See the IQ scale for more about how scores are arranged.
Why can an old test overstate a score today?
An old test can overstate a person’s standing today when its norms come from an earlier group. If average test performance has shifted since then, the same answers may compare more favorably with that older group than with people tested under current norms.
The issue is the comparison group used to set the scale. Read a historical score with its test version and norms in mind. For a closer look at what tests measure, see how IQ tests work.
What causes the rise in test scores?
The causes of rising scores are still debated. Researchers have proposed possible explanations, but the score trend itself does not establish any of them as fact.
The effect concerns scores on tests. It does not by itself show that every kind of reasoning or ability changed, or that one explanation accounts for all changes across tests, places, and periods.
How should you use the Flynn effect?
Use the Flynn effect as a reason for care when comparing IQ scores from different eras. Tests may use different questions, norms, and reference groups, so the figures may not support a direct comparison without that context.
The trend cannot tell an individual what they would score on a current test. A current score also needs to be read against the norms for its test version. For context on why the normed average is 100, see average IQ. IQClock’s free timed reasoning test gives a reasoning result, not an IQ score.