THE NEED FOR ANALYTICAL STUDY OF THE MAZE 



PROBLEM 



J. F. DASHIELL 



The University of North Carolina 



The maze method of studying the process of learning, as it 

 is so commonly used in animal psychology and to some extent 

 in human, is still employed almost entirely in an unanalyzed 

 manner so far as concerns the apparatus. The particular pattern 

 of labyrinth to be used by an experimenter seems to have been 

 principally determined by the patterns used by earlier workers. 

 The Hampton Court maze is the classic instance. A variation 

 from earlier forms may be forced by such considerations as: 

 the limitations of the learning capacity of the animal to be used; 

 the limitations of available space; the nature of the problem, 

 e.g., whether motor or sensory habits ; the limitations of the mate- 

 rials to be used in the construction ; etc. Watson and his students 

 have introduced the use of a circular maze, but its particular 

 advantages have not been definitely set forth, so far as the writer 

 is aware. A thorough-going and detailed planning of what may 

 be called the architecture of the mazes that are to be used in a 

 given research is at present impossible on account of our igno- 

 rance of the relative values of different elements within a maze 

 problem. 



Many analyses have been made of sensory cues as they control 

 the maze running, and some analyses have been made of the 

 data available in scoring runs, but no definite analytical study 

 has been given the maze pattern itself. This would have some 

 importance in the study of the building up of single habits. 

 For instance, there is at present disagreement as to whether an 

 animal tends to eliminate the earlier or the later blind alleys 

 first. This is not to be settled by having two different investi- 

 gators at different laboratories use two different kinds of mazes. 



181 



PBTCHOBIOLOGT, VOL. II, NO. 3 



