60 HAZEL M. STANTON 



talent of the 54 students was higher than the talent of the 103 students as 

 shown by the solid talent profiles for each group. The variation in the 

 average centile ranks for the 54 group extends from .8 unit in pitch to 9.2 

 units in intensity, with an average variation for the five tests of 5.1 units. 

 For the 103 group the average centile rank variation extends from 1.1 units 

 in pitch to 12.3 units in intensity, with an average variation for the five 

 tests of 5.4 units. For the combined groups of 157 students the average 

 centile ranks in each test vary from .5 in pitch to 11.2 in intensity, with an 

 average variation of 5.6 units for the whole profile. 



The variations of T 2 centile rank averages over Ti are very similar in both 

 groups and cannot be said to be a significant increase. These average 

 variations in centile ranks of T2 over Ti . should be interpreted in the light of 

 the fact that the same centile rank norm was used for both test series. Had 

 there been a separate norm for re tests, the centile ranks in T 2 would have 

 been lower and the variation span in the two average profiles would prob- 

 ably tend to disappear. 



Talent profiles are classified empirically into a six letter group of A, B, 

 C+, C— , D, E (E the lowest). The individual talent profile classifications 

 for the 157 students fell in the first four letter groups of A, B, C-J-, C— . 

 These classifications were used for a further correlation of Ti and T 2 by 

 means of the coefficient of mean square contingency (C) which was found 

 to be .695. When the number of classes is 4, which it is in this case, the 

 C cannot exceed .866, hence the correlation of talent profile classifications 

 is significantly high. Therefore, those students who have an A talent 

 profile, a B, a C+, or C— talent profile at entrance to the school can be 

 expected to have a similar talent profile classification in a retest. 



Subsequent study of these and similar data have not changed the con- 

 clusion cited in the following quotation from page 18 of the study noted: 

 "Interpret as the reader may, the fact remains that did these tests not come 

 somewhere near measuring native capacities there should be much greater 

 gains after three years of intensive musical study. Not having data on 

 retests of groups that had no musical education in the interim between 

 tests, we are not justified in ascribing the slight gains shown for these groups 

 to the effects of music education .... the gains as well as the losses can 

 easily be due to cognitive factors as well as other attendant circumstances." 



The last part of this paper deals briefly with the retests of 645 children 

 over a period of ten years. These children have been divided into three 

 groups for study according to their day school grades at the time of the first 

 test; the pre-adolescent group, grades 4, 5, 6, includes 285 pupils, the 

 adolescent group, grades, 7, 8, 9, includes 208 pupils, the post-adolescent 



