P-, Q-, AND R-TECHNIQUES 185

Incorporating the more important points brought
forward in later publications, we may perhaps summarize
these and other distinctive features of Q-technique as
follows. To discuss them in any detail is hardly fair or
possible, until we have seen more plainly how they actually
work out in practice : my comments will therefore be
relegated to the footnotes.

I. Both P-technique and R-technique, c like all develop-
ments of psychometry up to the present day,3 have con-
cerned themselves solely with the study of individuals.
But Q-technique " is in no way concerned with individual
differences " but with the " deeper aspects of personality."
Thus, the old R-technique sufficed as a basis for applied
psychology ; but Q-technique forms an entirely new
branch of psychometry, which will serve as a basis for
general psychology as distinct from applied, i.e. for theor-
etical psychology " as distinct from educational or vocational
work in the field." ([138], p. 273).1

notation, to the more general proofs available with a matrix notation, and
in particular to demonstrating the identity of the factors obtained by the
two alternative approaches. The two-factor method was regarded, not as
wholly inappropriate or invalid, but as a simplified procedure giving a first
approximation only.

3 On this point he is strongly opposed to the views that had previously
been expressed. In my own Memorandum I had stated that " generally, the
correlation of tests or traits leads to an analysis of the human mind in the
abstract; and the correlation of testees leads to an analysis of the concrete
human population," i.e. to the study of individuals and their grouping into
types ([93], p. 253). This allocation of the two fields Stephenson would
reverse. Nevertheless, * Spearman's work with R-technique* (at least as I
understand it, and, I fancy, in the opinion of its author)—his studies ofe the
nature of intelligence,' of * the principles of cognition,* and of * the abilities
of man '—all this was surely primarily a contribution in the first instance to
general psychology rather than applied : indeed, factor-measurements for
individuals were, as a matter of fact, rarely calculated by the theoretical
investigators who followed the so-called R-technique—e.g. Spearman,
Stephenson, himself, and others working in the same laboratory. On the
other hand, for those of us whose daily business in the schools and school
clinics was essentially concerned with c individual differences,* correlating of
persons naturally arose as a necessary device for " educational and vocational
work in the field " as contrasted with the experimental work of the academic
investigator. I do not, of course, deny that Q-technique or P-technique may
be fruitfully applied to the problems of the theoretical psychologist as well: