802 



SCIENCE 



[N. S. Vol. XXXIV. No. 884 



reference to the length of his residence? Is 

 the summer student at Columbia or Chicago, 

 who may be in attendance for six weeks to be 

 reckoned in the total on the same basis with 

 students who are in residence for nine months ? 

 Until these things can be determined our sta- 

 tistics will have little value. 



George E. Vincent 



To THE Editor of Science : In your issue of 

 October 27 Mr. Handschin gives statistics 

 from 81 state-supported schools, regarding the 

 number of students to a teacher. These are 

 based on the report of the U. S. Commissioner 

 of Education. I fear that their main use is 

 to illustrate the fact that such statistics often 

 have very little value for purposes of com- 

 paring the several institutions. 



The University of Wisconsin, for instance, 

 is said to have 7.9 students per teacher, while 

 the University of Michigan is given 15. No 

 one can believe that if these institutions were 

 compared on equal terms one would be found 

 to have nearly twice as many students per 

 teacher as the other. I do not know whether 

 the statistics regarding Michigan are correct, 

 or not, but the figure for Wisconsin conveys a 

 wholly wrong impression. In the number of 

 the Wisconsin faculty, as stated by the U. S. 

 Commissioner of Education, are included the 

 staii of the Agricultural Experiment Station, 

 who do no teaching, and that of the Extension 

 Division, whose students are not included in 

 the catalogue. Thus the real number of stu- 

 dents per teacher is greater than that indi- 

 cated by the report. 



The number of " student hours " in the Col- 

 lege of Letters and Science has been recently 

 computed. These are ascertained by multi- 

 plying the number of students reported in each 

 course of study by the number of credits given 

 for the course. If the sum of these products 

 is divided by 15 — the standard number of 

 credits per semester — the result for 1910-11 

 shows a little over 3,000 " full time students " 

 taught by the faculty of that college. If this 

 number is divided by that of the faculty, the 

 quotient shows 11.1 students per teacher. If 

 the number of assistants in laboratories and 



elsewhere, who are employed for only partial 

 service, is reduced to that of the full time 

 instructors who would render the same amount 

 of teaching, the number of students for each 

 instructor is 13.3. In the college faculty are 

 included the teachers and executive officers of 

 the college, but not the executive officers of 

 the university nor the staff of the gymnasium, 

 library, etc. 



An investigation made by President Van 

 ITise before these statistics were compiled, and 

 covering the whole university, shows about 

 one teacher for 12 students — a result which is 

 not widely diilerent from that stated above for 

 the college. 



It is quite probable that if the number of 

 students per teacher were determined in state 

 universities of similar size on a basis which 

 would yield comparable results, the figures 

 would show no such wide differences as are 

 found in Mr. Handschin's list. 



E. A. BiRGE, 



Dean 



Madison, Wisconsin, 

 November 1, 1911 



Since the tabulations on " The Number of 

 Students to a Teacher in State Colleges and 

 Universities " were published in Science, Oc- 

 tober 27, some communications have come to 

 the writer complaining of the inaccuracy of 

 the figures. The writer explained that inas- 

 much as he based his figures on data sup- 

 plied to the U. S. Bureau of Education by the 

 authorities of the schools themselves, he feels 

 no responsibility in the matter. Evidently 

 some schools have not been over-accurate in 

 reporting their data. 



However, this brings up the question of 

 establishing common criteria as to who shall 

 be counted a teacher. This matter has been 

 brought to the attention of the U. S. Commis- 

 sioner of Education with a petition that he 

 establish some standard which might be in- 

 corporated in the blanks which are sent out to 

 receive the data. 



Who shall be counted a member -of the in- 

 structional force? The following suggestions 

 are offered: Eirst, the tabulations should be 



