Summary of Graduate Student Enrollment survey results. G. Vali  May 31, 2000 

Thirtysix responses have been received. These were tabulated by S. Warner and sent to me without identification of the institutions reporting. 1. The response rate of nearly 60% was excellent. We sincerely thank everyone who contributed to this multiinstitutional database. 2.
Responses to the bottomline question regarding the significance of the
problem were as follows: 



Since this is at 65% of the scale between "not significent" and "very significant," it may be regarded as an indication of relatively serious concern. 3. The table below shows the averages of the numbers reported in the questionnaire for each item asked, and the number of responses for each item. The numbers of responses decreased with the increase in years past. In addition, fewer institutions provided responses on GRE results than they did on student numbers. 

Table 1. Mean values and numbers of responses


To reflect the sizes of the entering classes, the total number for the 36 schools, backcalculated from the averages, is 388. 4. In addition to the overall averages, the data have been stratified by the size of the institution. All schools have been sorted in order of the average number of students who entered in the 5 years. This list was then divided into quartiles so that each group contained roughly one quarter of the total number of entering students. This led to a grouping of the 3 largest schools admitting a fourth of the total number, the next group of 6 schools another quarter, 8 schools in the third quartile and 19 schools admitting the last fourth. [One school didn’t report the number of entering students, for the sake of grouping only 0.4*(number admitted) was used. The factor 0.4 is average ratio for all the schools.] For reference, the average number entering in each quartile was 30.7, 18.0, 11.4 and 5.0. This illustrates the large range of school sizes; there is a ratio of factor 6 between the top and bottom quartiles. Needless to say, rather different circumstances govern schools in the different groups. 5. Figures 1a and 1b shows the numbers of applications, admittances and the number of entering students for the overall sample and for the quartiles. In Fig. 1b, the data are shown in terms of the ratio of the number for given years to the average of the 5 years. Generally, the data seem to be fairly stable, i.e. without excessive ‘noise’, except for the numbers of students entering in 199697. That year there was a sudden drop in the number of students entering in the 3 largest schools (all three schools), and this was only partially offset by the number entering in the 8 medium schools. Trends are summarized in the table below: 



As the table shows, most of the trends are negative. The largest decreases are in the numbers of applications and this trend is present for schools of all sizes. Admittances fluctuate the most, probably as a result of attempts to counter the declining numbers of applications. The final outcome in the numbers of students entering also shows large differences among the four groups of schools, with the middle groups loosing significantly and the largest and small schools showing some gains. The combined sample indicates a 3.2 % decrease, which is equivalent to 12 fewer students entering the discipline each year. If only those schools are considered which have reported all of the three categories of student numbers and for all five years, the changes are smaller. The trends are –5.0, +0.8 and –0.9 % per year, with correlation coefficients of 0.97, 0.29 and –0.42, respectively, for the numbers of application, admittances and entries. These trends are free from perturbations introduced by variations in the numbers of schools reporting, but reduce the sample size to 22 schools. Figure 2 shows the ratios of admittedtoapplied, enteringtoapplied and enteringtoadmitted for the combined data set, and for individual schools within each quartile. The first two ratios show increases, reflecting decreases in numbers of applications, while the enteringtoadmitted ratio decreased from 49% to 47%. There was sharp drop in this ratio in 1998/1999 as a result of a simultaneous drop in the average number of entering students and an increase in the numbers admitted. 6. The GRE scores are shown in Figs. 3a, 3b and 3c. No significant time trends are apparent in these data. The verbal scores decreased by 3.2 points per year but the correlation coefficient for this is only 0.34. Perhaps more important than the time trends is that there are impressive differences among the groups of schools. As shown in Figs. 3b and 3c the largest and medium schools have students with higher GRE scores than the other two groups. These differences are quite evident when the 5 years are averaged, as shown in Table 3: 



END OF REPORT 