RECORDS OF MEETINGS 373 



found, on the whole, that the results of these tests followed the general 

 order of preference indicated by the results of the questionnaire. Notice- 

 able sex differences obtained in the order of preferences. For example, 

 eggs, which stood first with the men was ranked by the women as sixth ; 

 milk stood fourth with the men and twenty-second with the women ; 

 salads, twenty-second with the men and fifth with the women; cucum- 

 bers, thirty-sixth (fourth from the last) with the men and nineteenth by 

 the women. 



The correlation between the median performance of the first and second 

 tests was .96 (Spearman's footrule). The average P.E. of the first per- 

 formance (average for all items) was 6.50; for the second 7.04. 



The individual correlations or indices of consistency ranged from .96 

 to — .05 with a median at .84 (P.E. 8.00). The second lowest, however, 

 was .39 and third .51. The subjects were most consistent in the arrange- 

 ment of their foremost likes (first five). For the last five the arrange- 

 ment was likewise more consistent than for the average, but not so con- 

 sistent as for the first five. 



The median number of items disliked by 50 men was for the first test 

 7.70 (P.E. 2.79) and for the second test 9.00 (P.E. 4.00). By the 50 

 girls the corresponding figures were 8.50 (P.E. 3.07) and 8.83 (P.E. 

 2.83). 



]\riss Mulhall said: The present investigation was to study the e<iuiva- 

 lence of rej^etitions for recall and recognition for four materials, pictures 

 of objects, geometrical forms, words, and nonsense syllables. Each sub- 

 ject was shown 15 words successively at intervals of two seconds each and 

 then required to reproduce those he remembered in three minutes. The 

 subject was then given a set of 30, containing the original 15 words, 

 from which he was to select 15 which he thought were previously pre- 

 sented. The first set of 15 words was shown again as before, and then the 

 subject requested to recall those he could and select 15 from the 30 set. 

 This Avas continued until he had recognized and recalled all of the 15 

 words correctly. The experiment was repeated for the three other ma- 

 terials (forms, syllables and pictures). 



The results show that the difference between recall and recognition is 

 greatest for pictures, somewhat less for forms and words and least for 

 nonsense syllables. In examining the material one finds the pictures 

 offer the greatest richness of associations. The forms, too, can be visual- 

 ized and in several cases named ; the words, which were all nouns, have 

 some associations, but lack a form or picture element. The syllal)les, as 

 their name implies, were nonsense, most, if not all, of wliich were devoid 

 of any association. 



