A SIMPLE MAZE : WITH DATA ON THE RELATION 



OF THE DISTRIBUTION OF PRACTICE TO 



THE RATE OF LEARNING 



K. S. LASHLEY 

 The Johns Hopkins University and the University of Minnesota 



I. A TEST OF THE VALIDITY OF RESULTS OBTAINED WITH THE 



SIMPLE MAZE 



The use of complicated mazes for the study of learning in ani- 

 mals originated in an effort to determine the limits of educability 

 of the lower mammals. The problem of habit-formation in 

 animals soon expanded, however, to include the more general 

 problem of the mechanism of learning and this led to a search for 

 a method of comparing individual differences in learning ability 

 and average differences for groups trained under diverse con- 

 ditions. The application of the complicated maze to the deter- 

 mination of individual differences followed as a matter of 

 course, since it provided a practicable technique for the train- 

 ing of animals. But in none of the pioneer work, nor, indeed, in 

 any of the studies thus far recorded, has there been a thorough 

 test of the applicability of the maze technique to the particular 

 problems studied by its aid. This is true also of the various 

 problem boxes and other apparatus used for training animals in 

 complex habits. 



The study of habit formation in animals has now advanced to a 

 stage where accurate work is possible and where statistical 

 methods can and should be employed for the evaluation of data. 

 Many problems require the comparison of the rates of learning of 

 numerous animals trained under diverse conditions, as in studies 

 of the action of drugs. If the animals are trained in complex 

 habits the experiment frequently requires an expenditure of time 

 out of proportion to the results obtained, and the necessary 



353 



