3fJ0 AXNALt^ NEW YORK ACAn[■:^rY OF XVIE.WEH 



from (J to 1() 3ears of ^ge, the majority of cases (23G) being under 13 

 years. Experimental conditions were adhered to as strictly as possible. 

 The three investigators were always in ignorance of the ])hysical age of 

 the child being examined. 



A normal distribution of cases about the "at age" position was found, 

 83 per cent, of the cases uiider 12 testing, "at age," 3 per cent, "above 

 age," and 14 per cent, "below age." 



The scale was not uniform for all ages, as showii by the average age 

 difference of each physical age group, given in the following array : 



I'hysical jise 7 8 9 10 11 12 



Averai?e (liffen-nce — 0.5 — 0.7 — 1.4 



The lack of tests above twelve years, and the difficulty of the "twelve- 

 year" tests cause the deviations from the norm at 10. 11 and 12 years. 



The teachers and the principal graded the t-iiiJilren into five groups 

 according to mental capacity. The average age difference of the five 

 groups correlated with the teachers' judgments were as folloAvs: "Very 

 briglit" + 0.9, "Bright" 0, "Average" —0.5, "Dull" —0.9, "Very dull" 

 —1.8. In -4 per cent, of the cases there was a disagreement between the 

 judgments of the school authorities antl the results of the tests. 



From the results of the investigation, it was found possible to conclude 

 that the scale, as now standardized, measured the development of intelli- 

 gence of the children examined with at least 96 per cent, efficiency, and 

 served as an adequate measure of comparatively slight individual differ- 

 ences in groups of the same physical age. The "twelve-year" tests were 

 found to be unsatisfactory. Sex differences were slight, girls possibly 

 tending to vary more than boys. The influence of the personal equation 

 of the experimenters upon the results of the tests was found to be neg- 

 ligible. 



Professor Woodworth said: Though tbo question of mental fatigue 

 has been most examined in prolonged work, it is possible that a clwrac- 

 te'ristic work curve should be obtained from short periods. In collabora- 

 tion with Drs. AVells and Pedrick, the author has studied periods of 5-40 

 seconds in controlled association tests (logical relations, color naming, 

 simple directions), series of 10 or 20 stimuli being visually presented all 

 at once, and the subject' being required to react to the stinmli one after 

 another without intermission. The time of each single reaction was re- 

 corded in order to see whether the speed of reaction changed in the course 

 of the series. Tlie work curve so obtained varies from trial to trial, but 

 on the average runs a definite course. The initial reaction is the slowest, 

 the next few the quickest of all, then comes a gradual decline of speed 



