CRITICISMS OF RESULTS 411

But a fuller and more precise correspondence can be estab-
lished, if we turn to the results of those investigators who
have used a statistical procedure. Here the work of Spear-
man and his school, together with Spearman's critical
resume of earlier researches, and the conclusions of Stephen-
son and others who have adopted Spearman's methods,
afford the best basis for comparison.

I have endeavoured to meet Stephenson's criticisms
of my methods. May I now briefly answer his criticisms
of my results ? As before, I believe that a little examination
will show that the difference between us is far less than he
imagines. The earlier conclusions of Stephenson and those
who collaborated with him were based chiefly on tests
applied to adults ; but, as they point out, the factors
discovered very closely resembled those that had been
obtained in my own work with normal and abnormal
children. The final summary of their results is given in
the issue of Character and Personality, which also contains
Stephenson's first experiments on correlating persons. For
testing personality, however, and for studying clinical and
temperamental types, they relied at this stage exclusively
on R-technique—i.e. on correlating tests or traits. Since
they were themselves the observers and the testers of every
person in their group, this procedure was always open to
them. I, on the other hand, had usually to collect the data
for my clinical cases from different observers ; and so had
often been forced to start by correlating persons. Neverthe-
less, wherever it was practicable, I should still consider that
6 R-technique3 was the safer procedure, though Dr.
Stephenson now seems to repudiate it.

shrinks from self-display; E's laughter is frequent and free ; I is timid and
shy ... I is curious about scientific and metaphysical problems, gloomy,
sensitive and even sardonic: his emotional expressions are repressed and
restrained," etc. (Outline of Abnormal Psychology, pp. 436 et seq., abridged).
Holt epitomizes what is common to the two types of response by summarizing
the one asf aggressive or adient * and the other asc avoidant> (Animal Drive
and the Learning Process, 1931); and Katz regards ' these two fundamental
reactions as the clue to extraversion-introversion' (in his chapter on
' Personality,' a$. Boring, Langfeld & Weld, Psychology, p. 510). For a
general review of the whole subject, see Guilford and Brady, Psycbol. Bull'.,
XXVII, 1930, pp. 96-107.