282 



ANNALS NEW YORK ACADEMY OF SCIENCES 



of all lengths, we have an average number of mistakes for letters of 5.72, 

 with an A.D. of 2.24; while for figures it is only 3.93, but with a greater 

 A.D. of 2.39. Wlien we here consider the sexes separately, we have the 

 same greater number of mistakes for letters, and decidedly greater varia- 

 bility for figures. For both letters and figures, with series of all lengths, 

 there is a grand average number of mistakes of 4.83, with an A.D. of 

 2.32. From the point of view of the sexes, the same relation holds. 

 However, with all the series, in both letters and figures, the women make 

 more mistakes than the men (av., women, 4.96, men, 4.50). But the 

 men are more variable than the w^omen (A.D., men, 2.41, women, 2.22). 

 The following table gives the facts for the Immediate Memory Span: 



Immediate Memory Span — 100 Subjects 





Letters 



Figures 



Grand Av. 





Mode 



6.0 



6.0 



6.0 





Av 



5.69 



6.47 



6.08 





A. D 



0.59 



0.92 



0.76 





S. D 



0.73 



0.09 



0.91 





Men 



Women 



Men 



Women 



Men 



Women 



On per cent basis. 



Mode 



5.0 



6.0 



7.0 



6.0 



6.0 6.0 





Av 



5.58 



5.74 



6.76 



6.36 



6.17 1 6.05 





A.D .... 



0.60 



0.58 



0.89 



0.87 



0.75 0.73 



Using the Pearson coefficient, the writer correlated the memory spans 

 of the 100 subjects and for figures. /•= + .26 (only), with a P.E. of 

 .06. For the 31 men alone, r^ + .31, with a P.E. of .10. For the 69 

 women, r= + .27, with a P.E. of .07. 



Miss Barrett said : In this experiment, the order of merit method and 

 the method of paired comparisons were applied to three series of ma- 

 terials involving judgments of varying subjectivity. The three series 

 consisted of (1) weights, to be judged with respect to their heaviness, 

 (2) specimens of handwriting, to be judged with respect to their excel- 

 lence, and (3) propositions of varying validity, to be judged with respect 

 to the subject's degree of belief in the fact stated. 



The results were used as data by which to compare the relative effi- 

 ciency of the two methods with regard to statistical investigation of judg- 

 ment. Seven main problems are suggested, each of which involves a 

 basis of comparison between the two methods. 



I. The variability of each specimen in the series from the average 

 position accorded to that specimen, and the consequent average varia- 

 bility of the series. In the case of weights, this average variability is, by 



