That's an interesting hypothesis, but the effect size of environmental changes is MUCH larger than the effect size of changing gene frequencies in increasing the IQ scores of the general population. It has been typical for decades for authors who take a strongly hereditarian (which means "pre-Mendelian," really) view of influences on IQ to suppose that IQ score trends over time would be for the population mean to decline. In fact, population trends over time in countries all over the world have been HUGE increases in IQ.

The Stanford University researchers who developed the first few versions of the Stanford-Binet IQ tests had access to data for decades that could have revealed a surprising trend in raw item content performance on IQ tests. The trend, now known as the Flynn Effect, after Professor James R. Flynn, who had the greatest role in discovering it, is that raw scores on IQ tests, including all kinds of IQ tests loaded on fluid intelligence, have risen over time in national populations all over the world.

James R. Flynn started slowly but worked steadily in finding data sets and eventually found a large body of tests for which raw score data is available over long time series. His first published paper (1984) on the issue of IQ score changes over time was based on the renorming of the most commonly used IQ tests in the United States. He found that as new editions of the major IQ tests (the Wechsler test and the Stanford-Binet) were published, that the newer edition of each test invariably had a norming sample population that did better on the old edition's item content than had the old edition's original norming sample population. In other words, for a given brand of IQ test, as time went on, the average person did better and better on the raw item content of that test, resulting in higher and higher IQ scores over time for people tested on the same test with score calculation based on the earliest norms.

The possibility that raw scores on IQ tests were rising so consistently, and in such magnitude, was so surprising that it prompted many psychologists to cast doubt on the sample populations used for the early studies that showed this phenomenon--even though those sample populations were none other than the norming sample populations for the major brands of IQ tests. Psychologist Arthur Jensen proposed to James R. Flynn in January 1983, while Flynn's first major article on score trends was awaiting publication, that a good data set to show changes in IQ scores over time should

a) be comprehensive, e.g., a test of an entire national population, to eliminate the possibility of sample bias,

b) use the same test from generation to generation, with time trends shown by raw score differences,

c) emphasize "culture fair" tests such as Raven Progressive Matrices, which were presumed then not to include item content that is taught by compulsory schooling,

d) be based on adult populations, who have reached a mature level of intellectual functioning, to minimize the effect of differing rates of intellectual growth in childhood from one generation to the next.

Amazingly, Flynn found data sets with all of those characteristics, especially data sets from compulsory IQ testing of NATO draftees in the Netherlands, Belgium, and Norway. The data based on a subset of items from the Raven Progressive Matrices test showed substantial raw score rises for Dutch draftees: if the 1952 population mean is taken to be IQ 100, by 1982 the Dutch male mean IQ was 121. The IQ raw score rise over time was smooth, and by the end of this period had resulted in ceiling effects on the test for a significant number of draftees (Flynn 1987). Flynn has suggested an experimental design that might help unravel the causes of the increase in raw IQ scores over time, while reviewing newly discovered data sets that extend observation of IQ score rises backward in time to the beginning of IQ testing. The Raven test, early in its use, was given to an age cohort born in the 1870s as part of a study of IQ changes over the course of adult aging. Careful analysis of this cohort allows extending IQ raw score trends back to the beginning of IQ testing, showing that fully 90 percent of Britons born in 1877 scored below the fifth percentile of Britons born in 1967, or in other words that most turn-of-the-last century Britons had an IQ of 75 or lower on a highly g-loaded IQ test, according to current norms (Flynn 1999; Flynn 2000b).

As Mackintosh (1998, p. 104) writes about the data Flynn found: "the data are surprising, demolish some long-cherished beliefs, and raise a number of other interesting issues along the way."

CITATIONS:

Flynn, James R. (1984). The Mean IQ of Americans: Massive Gains 1932 to 1978. Psychological Bulletin vol. 95, pages 29-51.

Flynn, James R. (1987). Massive IQ Gains in 14 Nations: What IQ Tests Really Measure. Psychological Bulletin, vol. 101, no. 2, pages 171-191.

Flynn, James R. (1998). IQ Gains over Time: Toward Finding the Causes. In Neisser, Ulric (Ed.). The Rising Curve: Long-Term Gains in IQ and Related Measures. Washington, DC: American Psychological Association.

Flynn, James R. (1999). Searching for Justice: The Discovery of IQ Gains over Time. American Psychologist, vol. 54, No. 1, pages 5-20.

Flynn, James R. (2000a). IQ Gains, WISC Subtests and Fluid g: g Theory and the Relevance of Spearman's Hypothesis to Race. In Gregory Bock, Jamie Goode & Kate Webb (Eds.), The Nature of Intelligence (Novartis Foundation Symposium 233) (pp. 202-227). Chichester, United Kingdom: Wiley.

Flynn, James R. (2000b). IQ Trends over Time: Intelligence, Race, and Meritocracy. In Kenneth Arrow, Samuel Bowles & Steven Durlauf (Eds.). Meritocracy and Economic Inequality (pp. 35-60). Princeton: Princeton University Press.

Mackintosh, N. J. (1998). IQ and Human Intelligence. Oxford: Oxford University Press.