196 THE FACTORS OF THE MIND

mutual agreement) of examiners or observers grading or judging a
series of individual items. In evaluating the Binet-Simon tests,
for example, I tried to show that there was " a central factor, pre-
sumably the ideal order," underlying all the individual orders given
by the different examiners ; such a * central factor 9 could not
possibly have been demonstrated, had I started by correlating
tests.1 Again, in investigations on the degree of agreement among
school and university examiners, we endeavoured to show that, as a
rule, there was a " common factor, namely, the ideal order of merit,"
influencing the marking of each one ; this again could never have
been established had we correlated the marks by scripts instead.2
Somewhat closer to Stephenson's ' new method of testing' are the
various ranking tests we have tentatively sought to construct for
moral, aesthetic, humorous, and logical judgment, in which the
examinee is required to arrange series of moral offences, poetic
extracts, jokes, comic pictures, logical absurdities, musical pieces
for a concert programme, etc., etc., in order, and his grading is
correlated with a weighted average.8 Without correlating persons,
for instance, it would have been impossible to discover " an appar-
ently objective element in judgments on artistic taste . . . some one
general underlying factor influencing the judgments of all " ; more-
over, a practicable method of measurement was at the same time
described : "by correlating the order furnished by the individual
tested with the order furnished by art-critics, taken as a standard,
we could," it was stated," obtain a coefficient which might be a mark
or measure " for that individual's artistic capacity or " taste." 4
The same principle was tentatively applied to the familiar absurdities
test to obtain assessments of intelligence for vocational guidance
with supernormal adults : the examinees were given a series of short
reasoned statements of varying validity, and asked, not to point
out the absurdities, but to arrange the arguments in order, putting
the most cogent first and the most absurd last: thee intuitive logical
judgement * of each was then assessed by correlating his order with

1 Mental and Scholastic Tests (1921), p. 136 and Table 137.

2 Marks of Examiners (1936), pp. 275, 292; also [29] and [134].

8 Cf. The Young Delinquent (1925), pp. 404-5, and theses by Moore, Fraser,
Wood, Williams, Dewar, Wing, and others, referred to below.

4 I.e. what is technically called his e saturation coefficient' for the factor.
How the Mind Works (1933), p. 292; cf. also [118], p. 32 and Table I.
As will be gathered from the preliminary account, several hundred adults and
children, of different ages and sexes, were measured by Miss Felling and
myself in this way ; and the correlations between the measurements and the
teachers' judgments (particularly at the L.C.C. Art Schools) proved
encouragingly high ('66 to 79).