248 THE FACTORS OF THE MIND

Another field in which, I venture to suggest, a factorial
technique would be most fruitful is that of medicine. In
a recent publication [114] I have endeavoured to point out
how such a procedure may elucidate the classification and
diagnosis of nervous disorders, and of mental subnormalities
generally among school children; and, by means of data
collected with the help of school medical officers and others,
we have been hoping to show how it may prove equally
fertile in the study of physical disease. Other research-
students, specially qualified in the relevant field of work,
have been experimenting, so far with decidedly promising
results, to demonstrate its adaptability to problems of
industry, economics, and plant-fertility.1 Indeed, the
type of problem for which factor-analysis would seem the
most appropriate method is one which, so far from being
confined to psychology, is common to all complex sciences
where work is in a preliminary stage. In biology, in medi-
cine, in agriculture, in meteorology, in almost every sphere
of research where we are dealing no longer with a few
simple conditions operating on a few large bodies in the
cosmic void, but with clusters of interacting causes, affect-
ing highly composite reagents, there, as I have more than
once ventured to contend,2 innumerable questions are
waiting to be solved, or at least unravelled, by the factorial
methods that the psychologist has evolved.

can best be established by the c analysis of variance' (which I regard as
essentially a factorial method); but so far the method has rarely been tested
in actual practice.

For an account of the aims and methods of social surveys, cf. A. F. Wells,
The Local Social Survey in Great Britain, 1935, and id. ap.y F. C. Bartlett et al.,
The Study of Society, 1939. The surveys carried out by economists, though far
more adequately planned than those of sociologists, have been chiefly
limited to economic conditions; but their statistical methods, supplemented
by analysis on factorial lines, might serve as models for similar surveys on
social and psychological problems,

1 In these fields the most interesting outcome would seem to be a demon-
stration of the value of applying factor-analysis to problems hitherto treated
by the analysis of variance, just as it appears to be equally valuable in the
psychological field to apply analysis of variance to problems hitherto treated
by factorial methods.

2 [931 p- 313 ; Nature, cxHv, 1939, p. 533.