140 



SCIENCE 



[N. S. Vol. XXXVII. No. 943 



completely in Fig. 5, where the area over 

 each successive fraction of the four years 

 represents the percentage of pupils who 

 drop out from high school during that frac- 



tion of the four years. Thus the chance 

 that the "no" pupil will drop out before 

 the second half year is 58 out of 100, or al- 

 most five times the chance that a "yes" 

 pupil will. The chance that a "no" pupil 

 will stay four years is 7.3 out of 100 or less 

 than one fifth the chance that a "yes" 

 pupil will. 



An important negative fact in diagnos- 

 ing the probable length of a pupil's stay in 

 the New York City public high schools is 

 shown in Fig. 6. Economic conditions, as 

 measured by the family expense for rental, 

 are seen to be insignificant. 



In from one to three months after these 

 pupils entered school they were rated by 

 their teachers for ability, defined as fol- 

 lows : ' ' Native ability apart from success 

 or failure in any particular subject of 



study. Natural brightness." The median 

 expectation of the top tenth and bottom 

 tenth by the rating are as shown in Fig. 7, 

 the detailed facts being given in Fig. 8. 



For those of the pupils who stayed 

 through the first term or nearly to its close 

 so as to receive formal school marks in 

 their studies, the relation of school achieve- 

 ment in the first term to length of stay was 

 calculated. It is shown in part in Fig. 7 

 and Table II. Ten times as many of those 

 marked below 50 leave in the first year as 

 of those marked 90 or above. Of 115 pupils 

 marked below 50 not one remained to grad- 

 uate in four years. As the marks rise the 

 percentage leaving in the early years stead- 

 ily falls, and the percentage graduating, 

 rises. 



