II. TRANSFER OF TRAINING BETWEEN PAIRS OF MAZES 



The Positive Nature of the Transfer 



An experiment to discover whether there was transfer 

 of training when white rats which had learned one maze 

 were carried over to another, and whether this transfer 

 if present was positive or negative, was performed by 

 Webb.i He effected transfers between ten pairs of mazes, 

 in five of which the first maze remained the same, while 

 the second maze in each pair was different from all the 

 rest; and in five of which the first maze in each pair was 

 different from all the rest and the second maze remained 

 the same; so that the order of the transfers was from 

 A to B, from A to C, from A to D, from A to E, from A 

 to F; and from B to A, from C to A, from D to A, from 

 E to A, from F to A. He found in every case that the 

 number of trials required, the total number of errors made 

 per rat and the total time consumed per rat were less for 

 a group learning a given maze as its second to be learned 

 than for the group that learned the given maze as its first. 

 He is of the opinion that the evidence furnished by his 

 experiments both with rats and with humans is strongly 

 in favor of the presence of positive transfer. There was 

 no accounting for these savings on the ground of chance 

 or of group-differences; and, moreover, Webb constructed 

 his mazes with the express purpose of rendering the condi- 

 tions as favorable as possible for negative transfer. 



The present experiment was designed to further test 

 this matter of transfer as between pairs of mazes, using 

 mazes of another pattern and in a different arrangement 

 from those of Webb's. The results of this experiment 

 corroborate the results of Webb's with regard to the presence 

 and the nature of the transfer. The percentages of transfer 

 were computed on the same basis as Webb's — that of the 



*Webb: Transfer of Training and Retroaction. Psych. Rev. Mon. Sup. Vol. 24 

 No. 3. 



