A GRAPHIC METHOD OF RECORDING 

 MAZE-REACTIONS 



ROBERT M. YERKES AND CHESTER E. KELLOGG 



From the Harvard Psychological Laboratory 



One figure 



For nearly twenty years the maze or labyrinth has been em- 

 ployed by students of animal behavior. Both apparatus and 

 procedure have been improved steadily during the last decade, 

 but even to-day we lack an intelligently standardized form of 

 maze and a reliable method of recording the several important 

 aspects of the subject's reaction. 



We propose, in this paper, to describe a method of recording 

 maze-reactions which should greatly increase the value of the 

 results obtained in experiments with the maze. We shall not 

 attempt to describe a type of maze which promises to be worthy 

 of standardization, but instead shall limit ourselves to a brief 

 discussion of methods of observation. 



Experiments with the maze offer opportunities for the measur- 

 ing of several aspects of reaction. Especially important among the 

 data obtainable are (1) time of reaction; (2) distance; (3) number 

 of errors; and (4) nature and distribution of errors. Prior to 

 the devising of the method herein described, it has been prac- 

 tically impossible for even the highly practiced observer to obtain 

 accurate measurements of all of these features of reaction. 

 Indeed, with a rapidly moving subject like a rat or a mouse, it 

 has been impossible, during the first few trials, to obtain with 

 accuracy any other measurement than that of time. This is 

 obviously quite as unnecessary as it is unfortunate, for we have 

 good reason to believe that distance and error data are in many 

 experiments more important than time data. Because of our 

 conviction that a variety of data should be obtained in every 

 maze experiment and that all measurements should be made 

 with a reasonable degree of facility and accuracy, we have made 

 it our business to attempt to devise a method which shall enable 

 an experimenter to record the various aspects of reaction directly 

 and graphically. 



50 



