STUDYING CONTROLLED WORD ASSOCIATIONS 389 



because they are more suitable for educational and pathological 

 tests. It is to be expected that the average reaction time on 

 such a list of words given to normal subjects would be consider- 

 ably lower than when the list of stimulus words is maximally 

 inclusive as is the case here. The aim here was to retain as 

 many words as possible both difficult and easy, rather than to 

 choose a select, homogeneous list. But aside from these two 

 reasons which might justify unduly long reaction times, it must 

 still be concluded here that these reaction times are not long at 

 all. A careful gleaning of the literature shows that where any 

 considerable number of stimulus words has been used, reaction 

 times running up to four and five seconds are not uncommon 

 even for normal subjects. To what extent these high reaction 

 times have been found is difficult to determine because in most 

 cases detailed results are not given, and mean variations are 

 invariably omitted which might afford a clue to this point. 

 The chief interest has been concerning the nature of the 

 associations themselves and this has helped to minimize a dis- 

 cussion of the length of the reaction time. For this reason 

 detailed results are given here showing the number of cases, 

 averages, mean variations, and distributions. All computations 

 were made on the Burroughs electric adding machine. Mean 

 variations were obtained with the Dunlap formula, which is 

 especially adapted for use on a calculating machine. 



Table 4- This table shows the variation in reaction tune 

 according to the position of the accent in three syllable adjectives 

 and nouns. It was not considered worth while or valid to make 

 comparisons on accent for two syllable words inasmuch as the 

 percentage of these with the accent on the second syllable is 

 extremely small in comparison with those having the accent on 

 the first syllable. And likewise for three syllable adjectives and 

 nouns, comparison is really valid only between words having 

 the accent on the first and on the second syllable, inasmuch as 

 the number of words with the accent on the third syllable is 

 very small. No conclusion can be drawn at all in regard to the 

 effect of accent on reaction time. The variations in reaction 

 time with change of accent for either adjective or nouns follow 



