80 Estimation of the General Intelligence of School Children 
tions ; the former, however, largely depends on powers of application, perseverance 
and general care, and is thus not so reliable a test of mental capacity as the latter. 
(ii) Material. For the purpose of the present paper particulars have been 
taken from two Term Report-Sheets of a London Secondary School for Boys. 
These include (1) Age, (2) Form, (3) General Intelligence (as judged by the 
masters), (4) Place in Form, based on the term's marks, and, in one series also, 
(5) Examination Results. Tables have been compiled and correlations obtained 
(a) between General Intelligence and Examination Place from the one Report, 
and (b) between General Intelligence and Place in Form from both. Corrections 
have been made in each case for constant Age and constant Form. 
The ages of the boys range from 10 to 18 ; the numbers investigated are 182 
and 238 respectively, and comprise the whole of the pupils in attendance for the 
periods under review, with the exception of some three or four who were absent 
during the examination or for most of the term through illness. 
The method adopted for estimating the General Intelligence was as follows : — 
The masters, most of whom are specialists and only teach certain subjects, were 
required to give their opinions of the mental capacity of each boy whom they had 
taught during the term, using numbers up to a maximum of 5. The numbers 
5, 4, 3, 2, 1 correspond approximately to the terms Very Able, Distinctly Capable, 
Fairly Intelligent, Slow Intelligent, and Slow, respectively, in Professor Karl 
Pearson's Scale of Ability*. The entrance examination which each boy must 
pass before admission to the school makes it unnecessary to provide for the 
remaining two classes of that Scale, viz. : — Slow Dull and Mentally Defective. 
Every form is taught by four, five or six different masters whose judgments are 
roughly averaged by totalling the marks given for each boy and reducing to a 
possible maximum of 20. We thus have a numerical scale of intelligence, which 
however, must be considered chiefly as qualitative in character, ranging from about 
4 to 20, and a measure of capacity which is in effect the mean of several judgments ; 
moreover, the risk of a judgment formed by a single master on account of a boy's 
special aptitude or exceptional dullness in an)' particular subject and also the 
personal equation are largely eliminated. 
At the time of the first estimation of intelligence the whole of the 182 boys had 
been in the school at least six months and many of them much longer but the 
second estimation was made when about 33 per cent, of the 238 had been less than 
three months in the school. To this shorter period of observation is probably due 
the lower correlation found as the result of the second investigation. 
The examination was in all subjects of the curriculum, and was conducted in 
part by the teacher who made the estimate and in part by the Head Master and 
his colleagues. Place in Form is fixed by marks obtained during a complete term 
for work done in school and under the teachers' supervision. Marks gained for 
homework done under conditions which are not entirely controlled by the teachers 
are not included. 
* See Appendix to tbis paper. 
