TRANSACTIONS OF SECTION L. 805 



6. The Measurement of Intelligence in School Children. 

 By William Brown, M.A. 



Although, from an abstract and formal point of view, it might be expected 

 that a definition of intelligence should precede any investigation into the problem 

 of its measurement, such an arrangement is only partially justifiable. Unless we 

 are prepared to hypostatise intelligence as merely one of a number of ' faculties ' 

 displayed by the mind, we must face the problem of an all-inclusive survey of 

 the mental ability of individuals ; and if we possess the means of measuring the 

 nature and degree of the interrelations existing between the various distinguish- 

 able factors of such mental ability, we may find ourselves justified in regarding 

 the mind as some sort of system to which in its entirety, at least from one 

 point of view, the name intelligence may best be applied. If such a result should 

 emerge the problems of the measurement and the definition of intelligence would 

 be solved simultaneously. 



Since the mind, like the body, is variable, the method most applicable to the 

 problem will be the statistical method of correlation. Taking a sufficient number 

 of cases we may proceed to determine the magnitude of the tendency to con- 

 comitant variation displayed by the various subsidiary mental capacities dis- 

 tinguished by ordinary thought and measured by ordinary standards. To carry 

 out this plan with any attempt at systematic completeness would involve the 

 evaluation of the 'correlation ratio' ( tj) as well as the 'correlation coefficient' 

 (r) for each pair of capacities under consideration, in order to determine the form 

 as well as the degree of the correlation. A further indispensable part of the 

 mathematical technique would be to apply the method of ' multiple correlation,' 

 whereby, on a certain assumption (the assumption of linear regression) the 

 magnitude of the tendency to concomitant variation possessed by any two of 

 the capacities under consideration, independently of the tendencies of each to 

 vary concomitantly with the other capacities, may be determined. 



The author has applied this method to the investigation of the interrelations 

 of part-capacities in elementary mathematical reasoning in eighty-three boys 

 (already published 1 ), and to that of the relations of elementary mental abilities 

 to one another and to ' intelligence ' as ordinarily measured, either by school marks 

 or by 'general impressions,' and again by the specially devised C ombinatione- 

 Methode of Ebbinghaus, in the case of college students and elementary school 

 children — about two hundred individuals in all, carefully segregated according to 

 age, sex, and other ' irrelevant ' conditions. The results show a certain general 

 tendency to agreement among themselves, though indicating a much more com- 

 plicated scheme of interrelation than that inferred — on somewhat inadequate 

 data — by the champions of a ' central factor. ' The correlations are also ' low. ' 



Much of the correlation hitherto appealed to as evidence of the existence of 

 one single central factor 'is undoubtedly spurious' in nature, i.e., arising from 

 irrelevant factors, such as the influence of strange apparatus on the children, 

 personality suggestion, differences in degree of discipline to which the various 

 members of the groups examined had been accustomed, &c. The mathematical 

 formula, again, which have been employed to demonstrate this central factor from 

 the crude correlation results are much too abstract, involve too many improbable 

 presuppositions to be of any practical applicability. The method of 'multiple 

 correlation ' is the only sound and rational one for the investigation of the law 

 of relation of the various correlation coefficients to one another, but it has never 

 been employed in this connection by the investigators referred to. 



7. Perseveration as a Test of the Quality of Intelligence, and Apparatus for 

 its Measurement. By John Gray, B.Sc. 



Perseveration depends on an elemental property of the brain which determines 

 the persistence of mental impressions or the rapidity with which one impression 

 can follow another. It may be measured in various ways, one of the best being 

 by Wiersma's colour disc. On this disc are two colours which can be seen 



1 Biometrika, vol. vii., No. 3, April 1910. 



