Spearman's hypothesis

The English psychologist Charles Spearman, in his 1904 book, General Intelligence - Objectively Determined and Measured, described his two-factor theory of intelligence, using his strong background in statistics.

Additional recommended knowledge

Guide to balance cleaning: 8 simple steps

Daily Visual Balance Check

How to ensure accurate weighing results every day?

According to his theory, the performance of any intellectual act requires a general intellectual ability ("g") as well as some specific factors ("s"). While the g factor is available to the same individual to the same degree for all intellectual acts, the "s" factors are specific to an act and vary in strength from one act to another. He showed that while one's performance at tasks requiring strong general intellectual ability will be highly predictable by establishing his "g" beforehand, tasks requiring strong specific abilities will result less predictable performance.

His thesis was that in general - at a wide variety of tasks - a person's general intellectual ability (his "g") is a statistically significant predictor of performance - that is, his performance can be predicted significantly better by his known "g" than by chance.

Controversy

In the United States during the 20th century, a great number of studies have shown the existence of significant differences in the mean test scores between Blacks and Whites.

Arthur Jensen's legacy of the method of correlated vectors (a "vector" of scores is a set or collections of variables that possesses both direction and quantity) shows that the vector of any test's "g" loading is the best predictor not just of that test's correlation with scholastic and work-place performance, but also with physiological factors, like average evoked potential, reaction time, brain size, brain pH, and brain glucose metabolic rate.

With this, Jensen established the biological significance of "g" beyond that of Spearman's statistics. In the 1969 publication of his genetic research results in a 123-page article in the Harvard Educational Review, Jensen noted:

"The well-known differences in performance on intelligence tests of American blacks and white, with whites as a group regularly scoring higher than blacks as a group at all social-class levels, were due to inherent and essentially unchangeable intellectual differences between the two races, rather than to the effects of poverty, discrimination, and similar remediable factors". (Moritz, p. 211).

Many perceived these findings of Jensen as racist; he set off an intense controversy, which is yet to be settled. For a further discussion of this, see race and intelligence.

Dolan and Hamaker in 2001 reanalyzed the data from several previous studies (Jensen and Reynolds 1982; Naglieri and Jensen 1987) that used the statistical method invented by Jensen (the method of correlated vectors) with a more recent and improved method (multigroup confirmatory factor analysis). Dolan and Hamaker argue:

"On the basis of the present, as well as other results (Dolan, 2000), we are convinced that the Spearman correlation cannot be used to demonstrate the importance of g in b-w differences with any confidence." and "It is possible that the analysis of all available data sets (perhaps using an appropriate meta-analytic procedure) will demonstrate that a model incorporating the weak version of Spearman's hypothesis provides the best description of the data. However, until this work is undertaken, we cannot accept Spearman's hypothesis as an "empirically established fact" This leaves the validity of Spearman's hypothesis, considered a central justification for the genetic explanation, an unresolved question."

However in a group paper Famous artefacts: Spearman's hypothesis. Author's reply researchers found that positive correlations predicted by Spearman's hypothesis are likely only psychometric artefacts which also arise with measures which have nothing to do with 'general ability', for example, the number of toys and books a child has. In fact, these positive correlations will arise with any set of moderately correlated random data, once the sample is split into high and low groups.^[1]

References

1. ^ Famous artefacts : Spearman's hypothesis. Author's reply SCHÖNEMANN P. H., GARRIGA-TRILLO A., HAY D. A., HIRSCH J., HORN J., HUMPHREYS L. G., KADLEC H., MARAUN M. D., MILLSAP R. E., NAGOSHI C. T., STEIGER J. H., STERNBERG R. J., GRIGORENKO E. L., TURKHEIMER E. and WAHLSTEN D.