30 Effect of Alcohol on Psycho-Physiological Functions. 



pages, the average (arithmetical mean) is used for the statement of 

 central tendency and the mean variation (average deviation) as a 

 measure of variability. 1 It is conceivable that alcohol may affect not 

 only the average, which represents speed, amount, or sensitivity accord- 

 ing to the process measured, but the mean variation, also, since in 

 psychological data this stands for uniformity of conditions, consistency 

 of judgment, and controlled attention of the subject. Wherever possi- 

 ble, differences between normal and alcohol data will be given for the 

 mean variation as well as for the averages. The " probable correct- 

 ness" will be computed only for differences between averages. 



Dodge and Benedict assumed that the values obtained in measuring 

 any process were functionally related to the particular day on which 

 the subject and process were tested. 2 Hence, in addition to providing 

 for an equal number of normal and alcohol days, it will be seen that the 

 first period of each alcohol day was also normal, since the dose was not 

 given until the beginning of the second period. These first-period 

 values serve to indicate the neural condition, disposition, or level of 

 excitability on the different days for the different processes measured. 

 In the presentation of results emphasis is placed upon the differences, 

 plus or minus, between the first-period values of a particular measure 

 and those obtained for the succeeding periods of the same day. A 

 comparison of the differences obtained for normal days with those 

 obtained for alcohol days provides true differences between the two 

 groups of data and supplies a numerical statement of any effect of 

 alcohol. This is also the method used in this monograph; it may be 

 illustrated simply and concretely as follows: 



A series of eye-reaction time averages with their mean variations, 

 given in thousandths of a second, is shown in table 2. The average 

 reaction time (R. T.) for the first period of the normal day is 212 c. 

 (See section i.) At the close of this period and just previous to the 

 beginning of the reactions of the second period, the control mixture was 

 taken. Following this we have the average values for succeeding 



1 See Whipple, Manual of mental and physical tests; simpler processes, Baltimore, 1914, Chap. 

 Ill, for a very readable account of the usual statistical methods. 



2 The statement is made by Dodge and Benedict (p. 28 of their report) that their tables "are 

 commonly accompanied by a statement of average measurements, but the latter are regarded as 

 of relatively little importance. They are only given as details that may be of interest to some 

 future investigators who may be measuring similar processes." 



The actual average measurement values were not included in the case of the patellar reflex, but 

 only the differences were given. While the average values would have considerably enlarged the 

 table, their omission is already a source of regret, as it greatly hampers the comparison of other 

 data with the material collected by Dodge and Benedict, which is the most extensive concerning 

 this human reflex that is readily available. 



Dodge and Benedict found it necessary in some cases, notably in the memory (see Dodge and 

 Benedict's report, p. 130) and pulse sections, to make the averages the basis for comparison. 

 Also, their tables 27 on eye movements and 30 and 32 on finger movements are entirely devoted 

 to them. 



That wide differences in nervous excitability will be found is illustrated in Dodge and Benedict's 

 table 1 , page 46 of their report, and can be seen in many places throughout their tables where the 

 measured values are given. The differences between days are probably larger when the experi- 

 ments occur once a week and are scattered over a month or possibly several months than when 

 consecutive days are employed. 



