PSYCHOLOGICAL MEASUREMENTS. 



631 



from October 6 to November 3, and there is a marked decline on the 

 latter date. Inspection of table 188 will show that 7 of the men 

 increased the time for performing the task. They seem to have 

 approached the problem in an overconfident attitude, and not to have 

 paid careful attention to the directions. Five of these men who re- 

 quired a longer time at the second trial became confused ; these subjects 

 had to make three or four beginnings before they completed the task.^ 

 Another cause for the depression in the curve of Squad B on November 

 3 is that on this date How was included, but he was not included in the 

 average for October 6 as he failed to complete the task. His score for 

 November 3 is largest of all, being 770 seconds. Beyond this point, 



Fig. 122. — Average time required to complete the maze-tracing task. 

 Solid lines represent Squad A, broken lines Squad B. 



Squad B continued to improve and reached a level on January 5, 

 from which they improved very slowly on the three food-reduction 

 dates. 



The curve for Squad A is depressed on October 27. This is not shown 

 by the individual performance of all the group, but appears to be due 

 largely to the poor and irregular performance of Gar and Vea; the 

 former became confused twice and was obhged to begin again, with 

 the result that the time required (1,110 seconds) is the longest in table 

 187 for the 9 men whose records are averaged. There is also a depres- 

 sion in the curve on December 8. This was also due to the irregular work 

 of Vea and Gar, and to some extent of Tom. The averages for the last 



*No record was kept of the number of beginnings required on the first evening. This record 

 for later experiments is included in the table, as it illuminates the time record, indicating in the 

 case of a long time whether they moved very slowly and carefully or carelessly. 



