176 Irving Lorge 



The appraisal of Intellectual behaviour, however, is not 

 easy. Existing measures of intellectual ability suffer in con- 

 tent, arbitrariness of units, and ambiguity in significance. 

 In content, the tasks are varied — words, numbers, space 

 relations, information, pictures, and the like. Certainly, the 

 sampling of these and of other tasks must be representative 

 of the array of tasks to which a person must respond in the 

 course of a lifetime. Binet and Simon (Binet, 1908) empha- 

 sized, in their selection of tasks, the kinds of problems and the 

 variety of adequate performances that are related to a child's 

 life and experience and validated the goodness of response 

 against a criterion of school success. To a degree, then, these 

 tasks represented the range of cultural acquisitions included 

 in school curricula. For Binet and Simon, such a sampling 

 of intellectual tasks was, and is, reasonable. In appraising 

 adult intelligence, however, it may be more reasonable to 

 sample tasks from the range of experience that adult life 

 affords. Unfortunately, this has not been done. Most of the 

 tests used to estimate intelligence of adults are variations of 

 the same kind of tasks used to predict scholastic success of 

 youths and children. While it is necessary to obtain informa- 

 tion about the success of adults in such tasks, it is more 

 necessary to obtain a representative sampling of tasks that 

 will appraise adults in terms of a complex criterion of success 

 in the intellectual requirements of vocations, leisure-time 

 activities, care of the young, economic adjustment and the 

 like. The criterion of ability to learn may have to be enlarged 

 to cover areas other than success with school work. 



The significance of measures of intelligence depends upon 

 the criterion. Even in the child and youth range, the criterion 

 for the intellectuality of performance is far from perfect. In 

 the adult range, there must be an attempt to obtain an ade- 

 quate criterion for intellectual behaviours, or conversely, to 

 find out what the tests do measure. If the criterion for adult 

 intelligence were known, then the tasks could be selected and 

 weighted to maximize the relationship between the tasks and 

 the criterion. Since it is not known, the best that can be done 



