Selection and Mental Factors


METHODS of 'factorizing' the correlation-matrix of a team of intelligence tests have been described in the last few years by Spearman, Kelley, Thurstone, Hotelling and others, and it is of interest to observe what the influence of selection (including natural selection) in the population tested is on the number and nature of the mental factors thus arrived at. I have discovered that if a team of n = p + q tests has been resolved into r general and n specific factors which are normalized and mutually orthogonal, and if the variances and covariances of p of these tests are changed by selection, then the team can still be analysed into r generals and n specifics, but, in addition, a certain number l of new group factors will appear which, however, run through the p directly selected tests only. Further, these r + n + l factors will again be uncorrelated and normalized, but they are in general different from the original factors. Selection can thus be seen creating, destroying and changing factors.

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

THOMSON, G. Selection and Mental Factors. Nature 140, 934 (1937).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.