Does the normal bell curve apply to human intelligence or is giftedness far more abundant than we’ve been expected to believe?
Some gifted education researchers have proposed that the proportion of gifted people in the general population is larger than would be predicted by normal distribution, a view called the “overabundance hypothesis.”
While this is an appealing hypothesis, a team of researchers concludes in the Journal of Advanced Academics that a review of the scientific evidence does not support this position.
“As is well-known among psychologists, all measures of intelligence and related constructs (e.g., academic aptitude, cognitive ability) are susceptible to a gradual inflation, a phenomenon called the Flynn effect,” the authors write.
“Because of the Flynn effect, the average intelligence test score in a population increases by about one fifth to one third of a standard deviation per decade, with tests of verbal content having slower increases than tests that consist primarily of nonverbal content.”
Another explanation for the overabundance hypothesis, according to the authors, is that proponents may rely on old tests that use ratio IQs instead of the more modern deviation IQs. One of the reasons that ratio IQs were abandoned is that they often produced standard deviations much larger than the purported 15 or 16 points, especially in younger samples.
For the study, researchers reviewed scientific samples of intelligence distribution from Scotland, the Netherlands and the US. The 10 data sets from 6 sources met the following criteria:
The sample consisted of at least 1,000 people who were administered an intelligence or ability test; the sample had to be representative of the general population. The test also had to be free of a ceiling effect, meaning that the test had to be sufficiently difficult for most of the highest scorers to score below the maximum.
The test also had to have been normed in the previous 15 years and the sample could not be the same sample that the cognitive ability test was normed on.
The data had to be expressed as raw scores o,r if converted into an IQ scale, had to have a deviation IQ and not a traditional IQ score. Ratio IQ scores were not allowed because the standard deviations vary by age group and do not permit comparisons across some age groups, the researchers say.
The researchers compared the number of subjects in the top portion of the scoring distribution of the samples to what would be expected in a normal distribution.
The definition of giftedness in this study was focused on intelligence rather than a broader conception that would include leadership, developed talents or creativity.
“Indeed, some researchers who speak of many types of giftedness may still expect fully to find large proportions of gifted subjects in their methods of identification permit many ways for examinees to be labeled ‘gifted’, “ the researchers write. “Although we recognize that other definitions are valid, we believe that the correlation between intelligence and many educational and life outcomes makes it the most important construct related to giftedness….”
“Are There More Gifted People Than Would Be Expected in a Normal Distribution? An Investigation of the Overabundance Hypothesis,” by Russell Warne et al., Journal of Advanced Academics, Volume 24, Number 4, 2013, pp. 224-241.