How legitimate is the Flynn effect for the gifted?
The Flynn Effect in Gifted Samples: Status as of 2007
by John D. Wasserman
George Mason University
click for printer-ready
The rise in cognitive and intellectual test scores for at least three
generations has been termed the Flynn effect (Herrnstein & Murray, 1994).
Based upon findings reported initially in 1984, the Flynn effect describes a
robust finding of massive IQ gains over time and across nations (Flynn,
1984, 1987, 1994, 1999). For IQ tests, the rate of gain is about .3 IQ
points per year (or 3 IQ points per decade), roughly uniform over time and
similar for all ages (Flynn, 1999). Flynn (2006) clearly intends that his
finding of IQ gains over time should be applied to scores in all parts of
the score distribution, from very low to average to very high. He even
advocates that the Flynn effect, which was derived from large group studies,
be used to generate corrected scores for individual test findings, in spite
of the likelihood that such corrections are likely to contain much more
error than accurate prediction.
My January, 2007 examination of psychological research databases suggests
that the Flynn effect has not yet been adequately demonstrated for all
levels of ability; there is some support for its validity with low ability
individuals (e.g., those with intellectual disabilities or learning
disabilities) but there is no substantive evidence for its validity with
high ability individuals (particularly those who are intellectually gifted).
Evidence supporting the Flynn effect has been reported for the mild mentally
retarded range as well as the borderline range, with IQ gains over time
showing similar magnitudes as Flynn found in the middle of the IQ
distribution (Kanaya, Scullin, & Ceci, 2003). The idea that IQ gains are
concentrated in the lower half of the distribution was asserted by the
researchers who originally coined the term (Herrnstein & Murray, 1994). The
impact of the Flynn effect upon classification of individuals with learning
disabilities has been recently addressed by Truscott and Volker (2005; see
also Sanborn, Truscott, & Phelps, 2003), although their research is based on
LD diagnosis using IQ tests (a practice that has been seriously challenged
in recent years). I have yet to see any sound empirical studies of the Flynn
effect in gifted samples.
In his insightful critique of the Flynn effect, Rodgers (1999; see also
Rowe & Rodgers, 2002) notes that changing means in a distribution of IQ
scores does not identify which (sub)groups in the distribution actually
experienced change. For example, IQ gains over time can be produced by
improving harmful environments (e.g., through improved education and better
nutritional practices) in low-IQ individuals at risk. Some evidence supports
the contribution of low-IQ individuals to rising test scores: In large
samples of Danish draft-age males, for example, Teasdale and Owen (1989)
reported that test score gains appear to be concentrated among low ability
Rowe and Rodgers (2002) also note that changes in the variance of a
distribution of scores can mimic changes in means. A gain in IQ scores over
time can be produced by decreasing the variability in scores at the low
ability end of the distribution, or alternatively by increasing the
variability of scores in the high ability end of the distribution.
The Flynn effect was originally derived from research with tests that are
good measures of general intellectual ability (Spearman’s g factor), such as
Raven’s Progressive Matrices. If the effect does indeed reflect gains in g
over time, then Kane and Oakland (2000) may have found an explanation for
its relative absence in the highly gifted. Kane and Oakland reported that on
the Wechsler intelligence scales, g accounts for less variance in high IQ
individuals than low IQ individuals. Based on the argument advanced by Rowe
and Rodgers (2002) that changes in the variance in selected parts of a
distribution can account for the appearance of mean score changes in IQ, the
Kane and Oakland finding may explain why extremes in the intelligence test
score distribution do not all show the same Flynn effect.
The methodology required to demonstrate the Flynn effect for high ability
samples requires tests measuring comparable constructs with adequate
ceilings and samples that are free from selection bias across cohorts—all
requirements that constitute substantial challenges. To further complicate
matters, some recent findings suggest that the Flynn effect has run its
course and that general population IQ gains over time are reaching plateaus
(Teasdale & Owen, 2005; Sundet, Barlaug, & Torjussen, 2004).
Flynn, J.R. (1984).
The mean IQ of Americans: Massive gains 1932 to 1978.
Psychological Bulletin, 95, 29-51.
Flynn, J.R. (1987).
Massive IQ gains in 14 nations: What IQ tests really
measure. Psychological Bulletin, 101, 171-191.
Flynn, J. R. (1994). IQ gains over time. In R. J.Sternberg (Ed.),
encyclopedia of human intelligence (pp. 617-623). New York: Macmillan.
Flynn, J.R. (1999).
Searching for justice: The discovery of IQ gains over time. American
Psychologist, 54, 5-20.
Flynn, J. R. (2006).
Tethering the elephant: Capital cases, IQ, and the Flynn effect.
Psychology, Public Policy, and Law, 12, 170-189.
Herrnstein, R. J., & Murray, C. (1994).
The bell curve: Intelligence and Class Structure in American Life. New
York: Free Press.
Kanaya, T., Scullin, M. H., & Ceci, S. J. (2003).
Flynn effect and U.S. policies: The impact of rising IQ scores on American
society via mental retardation diagnoses. American Psychologist, 58,
Kane, H., & Oakland, T. D. (2000).
Secular declines in Spearman's g: Some evidence from the United States.
Journal of Genetic Psychology, 161, 337-345.
Rodgers, J. L. (1999).
A critique of the Flynn effect: Massive IQ gains, methodological artifacts,
or both? Intelligence, 26, 337–356.
Rowe, D. C., & Rodgers, J. L. (2002).
Expanding variance and the case of historical changes in IQ means: A
critique of Dickens and Flynn (2001).Psychological Review, 109, 759-763.
Sanborn, K. J., Truscott, S. D., & Phelps, L. (2003).
Does the Flynn effect differ by IQ level in samples of students classified
as learning disabled? Journal of Psychoeducational Assessment, 21,
Sundet, J. M., Barlaug, D. G., & Torjussen, T. M. (2004).
The end of the Flynn effect? A study of secular trends in mean intelligence
test scores of Norwegian conscripts during half a century. Intelligence,
Teasdale, T. W., & Owen, D. R. (1989).
Continuing secular increases in intelligence and a stable prevalence of high
intelligence levels. Intelligence, 13, 255–262.
Teasdale, T. W., & Owen, D. R. (2005).
A long-term rise and recent decline in intelligence test performance: The
Flynn Effect in reverse. Personality and Individual Differences, 39,
Truscott, S. D., Volker, M. A. (2005).
The Flynn Effect and LD Classification: Empirical evidence of IQ score
changes that could affect diagnosis. In A. Columbus (Ed.), Advances in
psychology research, Vol. 35 (pp. 173-204). New York: Nova Science.
John D. Wasserman, Ph.D., Associate Professor of Psychology at George
Mason University, is an educator, practitioner, and researcher in
psychology. After earning his Ph.D. in clinical psychology from the
University of Miami, he completed a fellowship in clinical neuropsychology
at Louisiana State University and Tulane Medical Centers in New Orleans. He
developed and directed a pediatric neuropsychology service at Children’s
Hospital in New Orleans, before spending 8 years in the test publishing
industry, directing the development of psychological tests to measure
cognitive, neuropsychological, social and emotional functioning—including
the Stanford-Binet Intelligence Scale (5th edition). From 2001 to 2005, Dr.
Wasserman directed the George Mason University Gifted Assessment Program in