142 



Effect of Alcohol on Psycho-Physiological Functions. 



decided advantage. In the accompanying table, No. 31, in which the aver- 

 ages computed by Dodge and Benedict are given in part I, it will be observed 

 that the computation of the averages for the normal days without including 

 the first-period values (see part n) has decreased the average saving for the 

 effect of alcohol with all of the subjects except Subject VIII, who shows a 

 saving of 1,027 as compared with 1,005. The only case in which the signs for 

 the percentile effects have been changed from minus to plus is in the values for 

 Subject IX, but all of the plus values have been increased in size and, except 

 with Subject VIII, the minus values diminished. For purposes of comparison, 

 average savings have been computed for the alcohol days, by including the first- 

 period values for the respective sessions, the results being given in part in. 

 The values in the normal column of part in correspond exactly to those in the 

 normal column of part I. With this method of computation, the alcohol val- 

 ues for memory are larger in every instance but one than those shown in the 

 comparison column of part n. The final effect of alcohol, when thus more 

 nearly isolated and given in per cent, is +1.4 as compared to +2.5 previ- 

 ously reported. 



Computing the average normal in the way suggested has sometimes pro- 

 duced only minor changes, as in the case of eye reactions and finger move- 

 ments, showing the effect of alcohol as a little larger than stated; at other 

 times, as in the case of word reactions and eye movements, the effect of such 

 recomputations has been to show a smaller effect for the alcohol. With the 

 association experiments there is no apparent difference; the averages are not 

 discussed for the faradic threshold. These points have been raised solely with 

 a view to justifying the method of elaboration used in the second series of experi- 

 ments and of emphasizing that when differences are available they should be 

 given chief consideration. It must again be urged that changes in the averages 

 of the normal days change in no way the comparisons which are based on 

 average differences between the normal of the day and succeeding periods. 



Table 31. — Memory data of Dodge and Benedict as affected by changes in method 



of computation. 



Part I. 



Average saving: 



Normal 



Alcohol 



Effect of alcohol (alcohol- 

 normal) , 



Percentile effect 



Part II. 



Average saving: 



Normal 



Alcohol 



Effect of alcohol (alcohol- 

 normal) 



Percentile effect 



Part III. 



Average saving: 



Normal 



Alcohol 



Effect of alcohol (alcohol- 

 normal) 



Percentile effect 



Subject 

 II. 



932 

 939 



+7 

 +0.6 



858 

 939 



+81 

 + 6.9 



932 

 963 



+31 



+ 2.6 



Subject 

 VI. 



761 

 856 



+95 



+ 8.7 



711 

 856 



+145 

 + 13.3 



761 

 914 



+153 

 + 14.0 



Subject 

 VII. 



1,382 

 1,189 



-193 

 - 13.1 



1,298 

 1,189 



-109 



- 7.4 



1,382 

 1,247 



-135 

 - 9 2 



Subject 

 VIII. 



1,005 

 974 



-31 

 - 4.1 



1,027 

 974 



-53 

 - 6.9 



1,005 

 909 



-96 

 -12.6 



Subject 

 IX. 



921 



881 



-40 

 - 3.9 



862 

 881 



+19 



+ 1.8 



921 

 903 



-18 

 - 17 



Subject 

 X. 



980 

 1,071 



+91 

 + 7.0 



978 

 1,071 



+93 



+ 7.2 



980 

 1,176 



+196 

 + 15 2 



Average. 



-12 



- 0.8 



+29 

 + 2.5 



+22 

 + 1-4 



