president's ADDRESS^SECTION J. 233 



the task is a very large one and requires the co-operation of 

 many workers. During the past two years a considerable amount 

 ot work has been done under the direction of Dr. Phillips. The 

 following tests have been applied to a large number of children, 

 and froin the data so secured the reliability has been calculated. 



The following table shows the reliabilitv of the tests studied: — 



Binet ... ... ... ... ... .89 



Opposites ... .. ... ... .71 



Part whole ... ... ... ... .73 



Coniplelicn ... . ... ... .72 



Porteus ... ... ... ... ... .29 



Healy Form Board . . ... .. .26 



Gcddard's Form Board ... ... ... .70 



Unless a reliability of about .60 is secured no test is worth 

 using in its existing form. Hence the Porteus Test and the Healy 

 Test are not sufficiently reliable. It is interesting to note that 

 Whipple also has found the Porteius Test unreliable. The results 

 cf the work done in Sydney will be published shortly in Schooling. 



From the evidence so far available the following conclusions 

 may be suggested : First, the distribution of mental ability both 

 general and specific, scholastic and non-scholastic, accords with 

 the normal curve and thus resembles the distribution of a physical 

 trait such as height or weight. Second, the rate of mental growth 

 i.? largely innate. So far, however, we know very little as to the 

 rate of mental growth. It is certainly not uniform, even in the 

 case of one individual. Burt remarks of backward children that 

 their " common characteristic is that their rate of educational 

 progress is much slower than that of the mass of children in 

 ordinary schools — about three-quarters of the normal rate " 

 (p. 38). Third, mental growth ceases at cliiferent levels which 

 may be represented in mental years. The highest mental level 

 is placed by some at 16 and by others at 19 years. Only the 

 very gifted reach this highest leivel ; according to American 

 statistics, 4 per cent, of the population. For the others the innate 

 growth cf intelligence ceases at various lower mental levels. 

 Children who are mentally defective, for instance, do not seem 

 able to advance beyond the mental level reached by the normal 

 child at 12 years of age. 



There is need for a scientific study of the efEect of teaching on 

 the different types of innate intelligence. Good teaching would 

 seem to increase the variability of a group heterogeneous in innate 

 ability. " The main effect," says Burt, "of teaching upon educa- 

 tional ability is, as a rule, to increase the individual differences 

 already present from birth" (p. 45). We already, in some 

 measure, recognise the need for adapting our teaching to the 

 differences in intelligence corresponding with age, and we are just 

 beginning to see that teaching procedure may require to be varied 



