370 MILDRED WEST LORING 



of age, sex, environment and the like on the character of the 

 associations, and lengthy efforts on the part of many authors to 

 find satisfying classifications of the associations. But it will be 

 noted though that comparatively little investigation has been 

 done where the chief end in view was an examination into the 

 actual technic of the method. The work of some of the earlier 

 German investigators, it is true, does show considerable effort to 

 regulate the length and type of stimulus words, but many have 

 neglected these factors entirely. Many, too, have been content 

 to use the stop watch to measure reaction time, and it may be for 

 this reason that important variations in reaction time with differ- 

 ent types of stimulus words and controls have been overlooked, 

 inasmuch as the watch method introduces into the total reac- 

 tion time three reaction times instead of one. The unrelia- 

 bility of the Hipp chronoscope without considerable modification 

 and standardization, and its operating difficulties, doubtless 

 caused many to abandon a careful determination of the reaction 

 time. 



A second fact which the history of the word association method 

 discloses is that most of the conclusions have been drawn from 

 free associations. In some of the early work continuous asso- 

 ciations were studied but the obvious disadvantages of this 

 method prevented its extended use. The controlled word asso- 

 ciation method, on the other hand, has had little investigation, 

 except by some of the German investigators, probably because 

 of the prevalent opinion that free associations represent more 

 truly the natural course of ideas in the individual. And yet the 

 controlled word association method has an advantage worthy 

 of consideration. Because the free association method leads 

 necessarily to a heterogeneous mixture of types of response, 

 nearly every investigator has been forced to expend a great deal 

 of effort in classifying the responses. Nevertheless they have 

 reached no uniformity in their classifications. This difficulty is 

 avoided in the controlled word association method, where the 

 stimulus words are uniform, and the response words by virtue of 

 the instruction to the subject are likewise uniform, so that the 

 problem of classification is obviated. 



