COMPLEX NEURAL ARCS. 89 



SUMMARY OF EYE-RE ACTION DATA. 



A summary of the latent time of the eye-reactions, as well as the 

 average differences, is given in table 8. The first and second normal 

 days are shown on the left, the two alcohol days on the right. The 

 other headings are self-explanatory. 



A summary of the effect of alcohol on the eye-reactions is given in 

 table 9, calculated from the differences. On the left the effect is shown 

 in the units of measurement. On the right it is shown in percentiles. 



VARIABILITY OF THE MEASUREMENTS. 



Inspection of the averages and mean variations of table 8 will throw 

 considerable light on the reliability of this group of measurements: 

 (1) In the first place, it will be noticed that the average mean variation 

 of eye-reaction is about 12 per cent of the average of the measurements. 

 In interpreting this variability it should be borne in mind that, with 

 the exception of Subject VI, none of the subjects had ever served in sim- 

 ilar experiments. We regard it as a conspicuous service of the eye- 

 reactions that they furnished us comparable " choice reaction" data 

 with an average mean variation of approximately 12 per cent from a 

 heterogeneous group of subjects without previous training. No other 

 "choice reaction" with which we are acquainted is so uniformly avail- 

 able. (2) As appears from the table of average reactions, there is a 

 slight but regular improvement in the average reaction time of all 

 subjects as the experiments progress. The averages show a total 

 reduction of 23 a for the main group and 13 a for the psychopathic 

 group between the first and last normal days. The only exception is 

 the second normal day of Subject XII, which prevents the average of 

 the psychopathic subjects from showing any advantage of repetition on 

 the second normal day as compared with the alcohol day. Notwith- 

 standing this exception, the facts are unequivocal. The average latent 

 time of the eye-reactions decreases by an average of about 11 per cent 

 from the first to the last experimental day as a result of repetition. 

 A regular practice effect of 1 1 per cent between the first and last quar- 

 ters of 120 measurements clearly shows that the process was not ini- 

 tially as thoroughly practiced as we had expected, i.e., to the degree that 

 the practice effect of the experimental sessions would be insignificant. 

 The question of the origin of the effect of repetition in the case of the 

 supposed thoroughly practiced eye-reaction, and the possibility of 

 adopting suitable experimental measures to reduce it, will be taken up 

 again in the summary, Chapter IX. 



We would point out here that, notwithstanding the obvious effect 

 of repetition, our normal base-line is adequate for any interpretation 

 of the effect of alcohol. The experimental as well as the statistical 

 procedure of these experiments was especially planned for just such 

 exigencies. 



