Changing intelligence test performance: A century of IQ increases

Changing intelligence test performance
A team led by Martin Voracek of the University of Vienna has analyzed intelligence tests over a period of more than one hundred years. Credit: University of Vienna
In the general population, IQ increases have been observed over the past 100 years. But are we really becoming more intelligent and if so, how much more intelligent are we becoming? University of Vienna psychologists Jakob Pietschnig and Martin Voracek now provide insights into how strong IQ gains are and perhaps why they occur. In their recent publication in the renowned journal Perspectives on Psychological Science, they show global increases of three IQ points per decade since the early 20th century.

In 1984, the first systematic account of IQ changes in the was published indicating increasing scores in the US general population. This phenomenon which has become widely known as the "Flynn effect" has been puzzling to the scientific community. Specifically, trajectories of IQ changes over time and possible reasons for gains have been heavily debated.

Global IQ gains

Now, researchers of the University of Vienna provide the largest account of IQ gains so far: Based on data of almost four million participants from 31 different countries, they were able to show that IQ increases amounted to about three IQ points per decade over a period of more than 100 years (1909-2013). Interestingly, global increases could be observed for reasoning as well as knowledge, although reasoning increases were substantially larger.

Based on this large data set and differences in strengths of gains over time and in different IQ domains, it was possible to evaluate potential causes of the Flynn effect: Improved nutrition, hygiene and availability of medical services which benefit foremost child development, as well as better education emerged as prime candidates for gains.

IQ test standardization

IQ tests are standardized to an average IQ of 100 and an average deviation of 15 and test results follow a bell-shaped curve. Many naturally measurable variables follow such a bell-shaped curve. This means that two out of three people score between 85 and 115 and 96 out of 100 between 70 and 130 IQ points. To account for changes of population ability, IQ tests are routinely reset to a quota of 100.

As mentioned above, the study results showed an average increase of about three IQ points every ten years. But do these results mean that an average IQ test result of 100 points in the present day would translate into an IQ of 130 a century ago? Although gains of about 30 points over a hundred years might suggest so, such an interpretation seems unlikely. Rather than increases in general cognitive ability, these gains are more likely to reflect improvements in specific abilities. "A person with an average IQ score of 100 in the early 20st century might have had quite different capabilities than a person with a seemingly equivalent IQ score of 70 in the present day", explain Jakob Pietschnig and Martin Voracek of the University of Vienna. IQ gains thus appear to be hollow in terms of global cognitive ability changes. Higher IQ test scores are more likely reflective of increasing specialization and better test taking strategies of participants.

Periodic changes

Interestingly, the strength of gains appeared to be non-linear. This means that periods of rather strong gains alternated with periods of smaller gains. For instance, periodic changes revealed smaller gains during the time of World War II in Europe, a time of considerable environmental stress.

Although IQ gains appear to be still ongoing, study findings suggest that the strength of gains has been substantially decreasing in the recent decades. This may indicate that beneficial effects of factors improving our test scores have peaked and IQ increases might cease in the coming decades. Future research will show whether these observed decelerations of the IQ gains will lead to an end or ultimately a reversal of the Flynn effect.

Explore further

Cuts in epilepsy drugs boost children's post-op IQ

More information: Pietschnig, J., & Voracek, M. (2015). One century of global IQ gains: A formal meta-analysis of the Flynn effect. Perspectives on Psychological Science, 10, 282-306.
Citation: Changing intelligence test performance: A century of IQ increases (2015, June 1) retrieved 24 August 2019 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments

Jun 01, 2015
Intelligence tests are partially a socially mandated con game. Does it matter if we are more intelligent but less aware, often to the point of complete oblivion?

Jun 01, 2015
"Average" IQ is a moving target. Individuals' performances on IQ tests, whatever they *really* measure, depend heavily on the individuals' development. This is not a universal variable, but has arguably been increasing worldwide over the past century as thje opportunities to develop intellectually have increased, a direct result of not having to struggle for day-to-day survival as much as in the past. Thank the improved availabilty of food, power, transportation, education, etc. over that of 1909.

Jun 01, 2015
You are taught everything on an I.Q. test by around the fifth grade, so it comes as no surprise to me that I scored above 130 every time I've ever taken a real I.Q. test, and have scored as high as 155 on some portions of the test.

It doesn't require much thought to score that high. Like I said, you are taught everything on the test by the time you are a 5th grader Especially for the math and spacial reasoning/pattern finding etc. For vocabulary I'll admit a few of the harder words are probably something mentioned to you a time or two in high school and you just forget them because nobody actually uses the words in day to day communication and you never see them in books either. I mean I've read hundreds of books on everything from Astronomy to zoology, and everything between, and never seen those words used.

They asked me to define some words and I knew more definitions of some of the words than the test is designed to catch. Like "acute" seriously who doesn't know two or more?

Jun 02, 2015
To account for changes of population ability, IQ tests are routinely reset to a quota of 100.

This is actually wrong. IQ tests are not "standardized" to 100 points, but the results are -normalized- to that.

The test may be the same, and the raw scores come out as they do, but in interpreting the results the mean of the raw scores is interpreted to be 100 and the scores corresponding to the standard deviations are assingned the +-15 points and so fort - the whole raw score distribution is squeezed and stretched to fit the normalized curve, so that you can say a raw score of say 725 corresponds to IQ 104.

The Flynn effect comes apparent when one compares the raw result to an earlier sample - the mean of the new raw score distribution would have gotten about 103 points in the previous normalized result a decade ago.

I.e. the raw scores on a particular test go up over time. Either because people become more intelligent, or become more familiar with the test.

Jun 02, 2015
Comparing different tests is of course meaningless, because they have different raw scoring.

If one test mean is 346 points and another test ten year later gets a mean of 923 points, it's impossible to tell whether there's been any improvement between the tests because they have different questions and answers and different points given per answer. Changing the questions changes the weights of your scoring system.

It's possible to "normalize" the two tests by having people take them side by side, but then again you get the problem of people simply learning the earlier test.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more