PAPERS OX PSYCHOLOGY AND EDUCATION 349 



The correlations between learning and retention, based 

 on the absolute numbers of units, i.e., numbers of " facts" 

 in the first selection, and percentile marks in the second 

 selection, are very high. For the first selection the cor- 

 relation is .87—. 02, and for the second selection, .94=±.01. 

 Lyon's correlations are much lower, "seldom going 

 above .4 and averaging only .25" 1 when prose was used 

 as the material, and recall without any re-learning was 

 the method employed. His method of measuring quick- 

 ness of learning was not as exact as the method here em- 

 ployed. He used the amount of time needed to secure 

 a perfect recital, and when a subject failed to give a 

 perfect recital, he was obliged to resume learning, and 

 this time was of course added to the learning time. 



Summarizing our results we may say that differences 

 in the rate of learning between the best and the poorest 

 in a group increase with increase in the difficulty of the 

 subject matter. By difficulty is meant thought difficulty. 

 In this investigation at least it is true that when persons 

 are allowed to determine the length of their learning 

 time, or time spent in study, those who learn the most 

 take the least time. Concerning the relation of retention 

 to learning, there is a decided tendency in this investiga- 

 tion for fast and slow learners to retain about the same 

 proportion of what they learned, and this is true whether 

 they reproduce what they learned in their own language, 

 or answer questions on it, and whether the subject mat- 

 ter is easy or difficult. Lastly, the material is very well 

 suited for laboratory experiments in educational psy- 

 chology. 



Some corollaries for teaching practice are, that teach- 

 ers should ascertain the relative learning abilities of 

 their students early and assist them to become conscious 

 of their rates of learning, and to adapt the length of 

 their study periods to their abilities, or else teachers 

 must expect less of the slower learners. Secondly, in 

 class reviewing fast learners can review their larger 

 learning in about the same time as slower learners re- 

 view their smaller learning, and with equal effectiveness. 



1 Lyon, D. O. Relation of Quickness of Learning to Retentiveness, p. 49. 



