230 president's address — section j. 



appropriate vocations; these and numerous other problems would 

 be largely solved by a scientific census of educational abilities 

 carried out and periodically repeated." 



Of the methods of conducting such a survey of the general and 

 scholastic ability of a school population it is impossible here to 

 say much. The distribution ol the pupils throughout the school 

 classes in relation to their ages gives for large numbers a fairly 

 accurate picture of the distribution of natural ability. An age 

 class table will show that for each age group the members are 

 distributed throughout a series of classes, the range of distribu- 

 tion increasing with age up to about thirteen. Beyond this age 

 the pupils in the public primary schools are hardly a fair sample, 

 since many havei left, and some have been promoted to higher 

 schools. Two things are, however, clear — first, that the varia- 

 bility of the group increases as the age rises, and that the distri- 

 bution is roughly symmetrical and in accord with the normal 

 curve. For London children, Burt found the standard deviation 

 to be about one-tenth of the age of the group, but for New South 

 Wales the variability appears to be less. Burt believes that 

 scholastic ability, if we could measure ii apart from the complica- 

 tions introduced by the influence of circumstances, would be found, 

 likei physical traits, toi be distributed normally. If v/e may accept 

 this assumption, we may use it to criticise the actual age class 

 distribution which we find in our schools. This actual distribution 

 gives too large a proportion of older pupils in each class group, 

 and too small a proportion of younger. It is probable, and experi- 

 encei seems to confirm the suggestion, that the brighter pupils tend 

 unduly to be retarded, and fail to move forward as rapidly as 

 their natural capacity would justify. If this is so, it involves a 

 loss, not only to the individual, iDut to the community. The 

 lesser variability of New South Wales children niay, perhaps, also 

 be due to faulty classification, for it is doubtful if a range of six 

 classes is sufficient for pupils from six to fourteen years. If the 

 range is adequate for the majority, it is certainly too small for 

 the abler pupils. We have, then, in the hypothesis of the normal 

 distribution of scholastic ability throvighout each age group a 

 useful method of determining the probable correctness of our 

 classification of a school population. It must, of course, be remem- 

 bered that if the population considered is small it may not be a 

 fair sample, and soi may not show the typical normal distribution. 

 But any marked departure, even in a small group, such as the 

 pupils of a, single school, should arouse inquiry as tO' whether it is 

 justified by the special circumstances of the school. 



Our present methods of classification of children, for instance, 

 that of one grouping for all school subjects, and that of yearly 

 or even of half-yearly promotions, is likely to distort the natural 

 distribution of scholastic ability. 



