96 
Teachers Appreciation of General Intelligence 
(4) Clothing. The children were classified in addition to their general intelli- 
gence by their clothing. This was judged in accordance with the subjoined scale 
provided by Professor Pearson. The grades are distinguished by Roman numerals, 
ItoV: 
I = Very well clad. II. Well clad, stuff suit, good boots ; sufficient, even if 
poor. III. Clothing poor but passable ; an old and, perhaps, ragged suit with 
some attempt at proper underclothing. IV. Clothing insufficient; boots bad and 
leaking. V. Clothing the worst; no boots or makeshift substitutes for them. 
Here again so few children occurred in Class V — five in all — that for statistical 
purposes they were included in Class IV. 
A boy's " order in class " was taken to be his place less "5 divided by the 
number of boys in the class. Thus if a boy was mth out of n boys, his class 
order was (m — 0"5)//i. The advantage of this method is that the mean order 
in each class is J and independent of the number in the class. 
The numbers in each school are given in Table I. It will be seen that the 
boys in Schools Nos. 3 and 7 were very few in number. They belonged to special 
classes whose teachers were interested in the work. In certain cases those teachers 
may take the higher divisions of certain standards, and thus we find a considerable 
relationship between intelligence and school*. When intelligence was correlated 
with age in the eight schools independently the relationship was found to be 
positive in three schools and negative in five schools, but the latter group 
included the two schools, Nos. 3 and 7, with under 50 boys dealt with in each. 
Similarly age and clothing had positive correlation in three and negative corre- 
lation in five schools, the latter including Nos. 3 and 7 again. Intelligence and 
clothing had inappreciable correlation in two schools, and quite sensible corre- 
lation in six. This was still true, if the correlation of clothing and intelligence 
was taken for a constant age. 
The ultimate relation of clothing to intelligence is an extremely interesting 
and important one, for the state of the clothing is often taken as a measure 
of home conditions and the intelligence of the children is thus asserted to 
depend on environment. The average correlation of the present data between 
clothes and intelligence based on the individual values for the eight schools is 21 
when corrected for age. The correlation between clothing and intelligence for 
constant age and constant standardf for the 1725 boys is "22. Either method 
leads us to the same conclusion : the intelligence is related to the condition of 
the clothing, the more intelligent child having the better clothing. Now there 
* When the headmasters take up an investigation of this kind and the whole schools fall into the 
record the contingency between intelligence and school is low. But it is very usual in schools to 
divide the standard into two parts, one the section with progressive children, the other the inert group ; 
naturally the grade of intelligence is quite different in the two sections, and the casual critic talks about 
the personal equation of the teacher within the same school to explain a result which a little knowledge 
of the actual conditions would have rendered quite clear. 
f The partial correlation for constant standard was taken to correct as far as possible any 
personal equation in the clothing estimates of the teachers. 
