DISCRIMINATIVE RESPONSES TO VISUAL STIMULI 



487 



The results show in general that the condition which yielded 

 the shortest average reaction-time for all the subjects, gave also 

 the most narrow distribution of the individual reaction-times 

 about the mean. This fact is inferable from a comparison of 

 the probable errors of the compared averages. 



DISTRIBUTION OF ERRORS 



In the course of accumulating the averaged reaction-times, 

 a number of wrong or anticipatory reactions were also obtained 

 on each subject. The distribution of these errors according to 

 the lighting condition under which they were given, is of consider- 

 able interest. The data are summarized in table 8. 



TABLE 8 



The results follow the same general tendencies as the reaction- 

 times. Those of subjects M and A suggest that discrimination 

 was more difficult under conditions DI and B 2 than under B 1 

 and that the subjects sometimes attempted to compensate by 

 an excessive strain of "attention;" i.e., that they were sometimes 

 set for a "motor" type of reaction, in the expectation that a 

 given side would be darkened; and that this expectancy was so 

 strong that when the stimulus finally appeared they reacted 

 without regard to its direction. The same tendency is very 

 clear in the results of subject E, but errors appeared much less 

 frequently than in the results of subjects M and A. I do not 

 regard the differences in the distribution of the errors of subject 

 C as significant. 



