Skip to main content

Verified by Psychology Today

Intelligence

James Flynn and the Flynn Effect: Death of an IQ Philosopher

He revolutionized the making and forensic utilization of US intelligence tests.

Key points

  • The death of James Flynn is an opportunity to discuss the "Flynn Effect." His studies of IQ norms revolutionized the field.
  • One area where the effect was often mischaracterized was in death penalty trials involving cognitively impaired defendants.
  • A common mischaracterization, involving IQ changes in Scandinavia is examined and shown to be irrelevant.

The recent death of James Flynn in New Zealand touched me in two ways:

  • I bonded with him over BBQ and music when we hung out together a few years ago while waiting to testify in a death penalty trial in Memphis Tennessee
  • The subject of the “Flynn Effect” named after him, often comes up in the course of my forensic testimony as a psychological expert in so-called “Atkins vs Virginia proceedings” where a diagnosis of Intellectual Disability (ID, formerly known as “Mental Retardation”) can cause the death penalty to be taken off the table.

The reason for that is because when applied correctly it can cause someone’s full-scale IQ score to be lowered by a few points, and (given that IQ still plays a big role in determining ID) a lowered score can save the life of a defendant in a capital trial.

Bryce Edwards from Dunedin, New Zealand on Wikimedia Commons
James Flynn
Source: Bryce Edwards from Dunedin, New Zealand on Wikimedia Commons

There is much confusion over the Flynn Effect (FE) so here is a boiled-down explanation. Flynn was a Professor of Political Philosophy (an Aristotle specialist) at the University of Otago who was interested in the misuse of intelligence research (such as by Jensen) to justify racist attitudes or policies. Before emigrating to New Zealand in the 1960s, Flynn was a civil rights organizer in eastern Kentucky and left the U.S. due to disillusionment paired with credible threats against his life. In the course of his research, he discovered that publishers of the three most widely used US tests—WISC, WAIS, Stanford-Binet—toughen the norms every time they come out with a new edition. The reason: The norming samples (constituted to reflect the population at the most recent census) obtained higher raw scores on items on the new version (determined by having a sub-sample of the norming subjects take for example both the WAIS-3 and the WAIS-4 with order varied), with the improved performance being especially notable for non-verbal tasks.

The reason for toughening the norms was so that the mean IQ score for the U.S. population continued to be 100. The consistent finding, from studies over many years, is that mean full-scale IQ rises by 3 points per decade. When looking at someone’s obtained IQ score, the Flynn-relevant practice is to subtract 0.3 full-scale points per year of norm obsolescence, determined by subtracting the year of norm-collection (which on average is two years prior to publication) from the year of test administration.

If someone obtained a full-scale score of 77 on a test with 10-year-old norms, a Flynn correction would adjust the score to 74 (10 times 0.3 is 3, subtracted from 77 is 74). As the cut-score typically used in ID definitions is 75 (formerly it was 70), this adjustment would move someone from the range of ”Borderline Intelligence” to the range of “Mild ID.” (The obscenity of using IQ cut-scores to determine disability status is another matter, which I shall save for a later entry). In the course of observing or participating in many criminal trials over the past 18 years, I have seen psychologists and lawyers make a multitude of errors when talking about the FE. Here is just one of those errors.

A common mistake is to point to research in Scandinavia involving testing of military recruits, suggesting that the trend line has reversed and the population IQ is now declining, rather than increasing. By this reasoning, the FE should no longer be considered operative. There are three reasons for rejecting this argument:

  • The FE has been established, to my knowledge, only on editions of the WISC, WAIS, and Stanford-Binet, published, and normed, in the US. It has not been meaningfully studied for tests published in other countries, and for tests other than the three named above. Thus, what happens in Scandinavia or other countries outside the US (virtually none of which have capital punishment) is irrelevant. When applying the FE to individuals tested in the US, what matters is the tests given to them, not a test normed and administered in say Denmark
  • A second reason is that the studies in Scandinavia do not really use IQ tests; rather they are jury-rigged cognitive batteries including the Raven’s Progressive Matrices, put together for incoming military recruits. Thus, this research cannot be used to characterize the IQ scores of these subjects let alone a nation’s population.
  • The third reason is that the subjects were tested only once with the same battery, while in the Flynn studies they were tested twice, with different test editions.

Such a cross-sectional method is very subject to confounding by extraneous population changes such as the education level of the recruits. In the time when the Scandinavian military scores were going up, there were increasing numbers of soldiers coming in with college degrees, while in recent years there has been an influx of people emigrating from Middle Eastern countries, with far fewer of them having college degrees. Such a methodological confound makes any argument about an end to the FE to be nonsensical.

All of the studies conducted by Flynn and others (including the test publishers themselves) have shown the FE to still be alive and well. As long as this continues to be the case, then it should be applied to adjust the scores on any test for which norm obsolescence data has been collected. The reason, as stated by James Flynn himself, is so that determination of ID, and imposition of the death penalty, should not be a lottery affected by the accident of which test edition was administered and on which date. It is very possible, as Flynn himself acknowledged that U.S. population performance on IQ items will tail off or even reverse direction at some point in the future. That day has not yet arrived, however, from what I can tell.

Copyright Stephen Greenspan

advertisement
More from Stephen Greenspan Ph.D.
More from Psychology Today