270 THE FACTORS OF THE MIND

devices. For our present purpose it will be sufficient to
group the various methods according to procedure rather
than aim. The distinctions I shall observe are set out in
the table below. There will, of course, always be combined
and transitional types; but that does not invalidate the
main dividing lines.

TABLE I.—Classification of Procedures

I. Analysis of Variance
II. Factor-analysis of

A. Covariances 1 , ^

-n ^ 1 . f between

x>. Correlations J

1. Persons} ,

2. Traits Jb^

(a) Group-factor Methods, with

(i) Non-overlapping Group-factors
(ii) Overlapping Group-factors
(£) General-factor Methods, with
(i) Simple Summation
(ii) Weighted Summation