Flynn Effect Conundrums
December 16, 2004
Risen 3 IQ Points Per Decade or Have They Risen 30 Points per Century?
Percentages are multiplicative, not additive. 10% + 10% + 10% isn't 30%--it's 33.1%! If IQs have risen 10 points per decade, then they should have risen 35 points in a century. Conversely, if IQs have risen 30 points in a century, then they will have risen at the rate of 2.6 points per decade. You can't have it both ways. Which is it?
2. Have IQs Risen 30 Points Retrospectively? Prospectively? Some Combination of the Two?
Have IQs risen 30 points retrospectively (2.6 points per decade)? Would the average student making the average IQ score of 100 in 1904 register 70 on our IQ tests today? If so, then the score that an average student making an IQ of 100 on our present-day tests should garner an IQ score of 100/70 = 141 on a 1904 IQ test... assuming there were any 1904 IQ tests. More realistically, someone scoring 100 on a present-day IQ test should register 100/74 = 135 on the 1916 Edition of the Stanford-Binet IQ test. I have 30-minute Henmon-Nelson pencil-and-paper IQ tests here that were published in 1931. These tests should yield a Flynn Effect of at least 22 points, so that someone scoring 100 on a current IQ test should exhibit a score of 128 on these old tests. Will they? I'm hoping to round up a few guinea pigs to try this to see what happens.
3. What's the Relationship Between Crystallized Intelligence and Fluid Intelligence?
It's most curious that the Flynn Effect shows up almost entirely with those kinds of tests that endeavor to measure fluid g (Gf). Arthur Jensen describes fluid g ("The g Factor", pg. 123) as "the capacity to figure out novel problems. It is indeed the 'eduction of relations and of correlates' as these demonstrated in mental tests (or life situations) in which specific prior learned knowledge, skills, algorithms, or strategies are of relatively little use." Dr. Jensen continues," But in large part, individual differences in Gf determine individual differences in Gc among persons with similar educational and cultural opportunities. Persons high in Gf tend to acquire more Gc (i. e., they reap greater returns on their initial investment) from their opportunities for learning than persons of lower Gf." " ...each person's Gc would closely parallel his or her Gf. in the person's own culture." So how is it that there have been huge boosts in scores on tests of fluid g such as the Raven Progressive Matrices without significant gains in the Gc areas of vocabulary, information, and arithmetic? (General information is a particularly significant issue, since we're all awash in floods of information.)
This is one of the reasons that James Flynn questioned the reality of the Flynn Effect reflecting an actual boost in IQ's.
Arthur Jensen endorses the observation that the Flynn Effect primarily operating upon low IQ's. But this leads to absurd conclusions, including the expectation that IQs among the severely retarded would have been boosted 30 points since 1904. This means that someone with a present-day IQ of 20 and three-year-old mental abilities would have an IQ of 50, measured by 1904 standards, with third-grade scholastic abilities. It follows that today's three-year-olds should have the mental capacities of second- or third- graders, measured by the standards of 1904. The average three-year-old should be reading at second- or third-grade level. People with IQ's of 50 today should have the mental capacities of people with IQ's of 80 in 1904, and should be fully functioning members of society. The percentage of our population that is sufficiently retarded to be institutionalized should have fallen dramatically since 1904. (It hasn't.) And if the Flynn Effect really began in 1880, then IQ's would have advanced by another 6 points since 1880.
Clearly, something doesn't quite jibe here.
Home Intelligence Index
To be continued