840 



SCIENCE. 



[N. S. Vol. XXIII. No. 596. 



of the arrangements for entrance to Amer- 

 ican colleges, but will give without com- 

 ment or inferences the facts concerning the 

 actual significance of a student's achieve- 

 ment in entrance for his later achievement 

 in college work, and also certain facts of 

 importance concerning the entrance marks 

 themselves. 



I have recorded the entire entrance rec- 

 ord and the entire college record through 

 1905 of every student entering Columbia 

 College in 1901, 1902 and 1903 who took 

 in whole or in part the entrance examina- 

 tions given by the College Entrance Board 

 of the Association of Colleges and Prepara- 

 tory Schools of the Middle States and 

 Maryland. These facts enable one to 

 measure precisely enough for all practical 

 purposes the relationship between success 

 in the entrance examinations (in any one 

 or in any combination of subjects) and 

 success in college (in any one subject or 

 year or combination of subjects or of 

 years). 



I shall, for the sake of brevity, refer 

 b.ere only to the 130 students for whom I 

 have complete records through junior year. 

 For each of these I have such a record as 

 the following: 



College. 



Freshman: English, 



Latin, 



Mathematics, 



German, 



Physics, 

 Sophomore: English, 



Latin, 



Mathematics, 



History, 



Physics, 

 Junior: English, 



German, 



Economics, 



History, 





 D 

 B 



C 



D 



C and D 



B 











C 



D and 



F 



P 



A and B 



In transmuting the A, B, C records into 

 more convenient quantities I have taken 

 A=6, B = 4, C = 3, D = l and F = 0. 

 Perhaps A-=10, B = 7, C = 5, D = 2 

 and F = would be a trifle closer to the 

 real values of these conventional measures 

 of relative position. The entrance scores 

 have been used in their original form. The 

 use of marks given in the conventional 

 way by examiners or instructors as meas- 

 ures of either quantity of ability or rela- 

 tive position in comparison with other 

 students is a matter involving many com- 

 plex and delicate questions of statistical 

 method. A too naive procedure may in- 

 troduce variable and constant errors of 

 considerable moment. It would take too 

 long to defend the author's methods and 

 would interest only the critical student of 

 mental measurements. Suffice it to say 

 that the author is cognizant of the prob- 

 lems and will state no result which would 

 have been essentially changed by the adop- 

 tion of any rational method of treating the 

 marks. 



The most important question is, of 

 course, the general relation between stand- 

 ing in entrance examinations and standing 

 in college work. The facts are given in 

 Tables I., II.,' III. and IV. The relation- 

 ship is only moderate even in the case of 

 the work of freshman year and dwindles 

 steadily, the coefficients of correlation be- 



