120 PSYCHOLOGICAL METHODS OF TESTING INTELLIGENCE 



We shall bring together next the chief require- 

 ments that a teacher must keep in mind when he un- 

 dertakes an estimation of intelligence if it is to be 

 usable for scientific purposes. 



He should conceive of intelligence, just as we de- 

 fined it at the beginning of this monograph, as "gen- 

 eral mental adaptability to new problems and condi- 

 tions of life, ' ' should give particular heed to the two 

 attributes "general" and "adaptation to the new," 

 and should guard against identifying with intelli- 

 gence any sort of special ability or the mere posses- 

 sion of information or readiness in speech. Because 

 of the general nature of intelligence it is essential to 

 take into consideration the way in which the child 

 behaves in quite different situations and when con- 

 fronted by problems of varied sorts. 



But, now, it is of the essence of the estimation of in- 

 telligence that not only shall each pupil be judged in- 

 dividually, but he shall also be compared with the 

 other pupils and be placed in a definite relation of 

 equality or inequality with them. For this purpose 

 of comparison there is to be laid down a rule that, de- 

 spite its fundamental character, has by no means al- 

 ways been observed: Only those pupils shold be lo- 

 cated in a given rank-order of intelligence that are 

 sufficiently like one another in other respects. The 

 reason is that the slight differences of intelligence that 

 have to be considered in making the estimation have 

 significance only when on the common basis of an 

 otherwise homogeneous group. For this reason the 

 comparative estimation of intelligence has usually 

 been restricted to the pupils of a single class ; but it is 

 necessary as well to be careful to secure homo- 



