Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 17
3
PAIRWISE CORRELATIONS
In its statement of task, the panel was asked to examine the correlations among a number of the
variables in the Assessment (see Box 1-1). Several of the correlations are presented in this
chapter, including correlations of student time to degree and completion rates with various
characteristics of doctoral programs, and correlations between the diversity of a program’s
faculty and the diversity of its students. All of the data are drawn from the tables of pairwise
correlations found in Appendix D, in which any correlations greater than or equal to 0.31 are
highlighted.
The correlations provide insights into the relationships between characteristics that can be
explored further. The panel focused its attention on correlation coefficients greater than or equal
to 0.3 because they are nontrivial and they may display, in the panel’s view, important
relationships between program characteristics. Pairwise correlations uncover these potential
relations of interest. Where associations are detected that, based upon prior knowledge, are
judged indicative of relationships worth further study, adjustments for potential confounding
variables must be made. Such adjustments are beyond the scope of this brief report.
Table 3-1 provides the correlations of student median time to degree and average cohort
completion rate with three measures of faculty research productivity: average publications per
faculty member, average citations per faculty member, and the percent of faculty with grants (see
Appendix C for definitions). There is little relation between the average cohort completion rate
and the productivity measures, with the exception of faculty with grants in physiology. The
correlation of median time to degree and grants is also strong for physiology, and the correlations
of median time to degree with citations per publication are strong for physiology, biomedical
engineering and bioengineering, genetics and genomics, and immunology and infectious disease.
Correlations in these four fields do not meet the 0.3 level with respect to publications per faculty,
although they range from 0.179 to 0.272. The only field with a strong correlation between
median time to degree and publications per faculty is nutrition. Where appreciable correlations
exist between median time to degree and measures of faculty research productivity, greater
research productivity is associated with longer times to degree.
1
Correlations of 0.295 and higher were rounded to 0.3.
17
OCR for page 18
Table 3-1 Correlations of Median Time to Degree and Average Cohort Completion with Publications, Citations, and Grants
Correlation with Median Time to Correlation with Average
Degree Cohort Completion
Percent Percent
Faculty Average Faculty
Average Pubs Average with Pubs Average with
Fields per Faculty Cits/Pubs Grants per Fac Cits/Pubs Grants
Biochemistry, Biophysics, and Structural
Biology 0.052 0.166 0.077 0.123 0.089 0.094
Biomedical Engineering and Bioengineering 0.185 0.369 0.018 -0.184 0.015 0.148
Cell and Developmental Biology 0.014 0.128 0.081 0.087 0.057 -0.041
Genetics and Genomics 0.181 0.364 0.23 0.229 -0.02 0.149
Immunology and Infectious Disease 0.179 0.327 0.189 -0.067 -0.05 -0.02
Integrated Biological and Biomedical Sciences -0.12 0.058 0.04 0.056 0.021 0.014
Microbiology 0.232 0.289 0.302 -0.072 -0.087 -0.201
Neuroscience and Neurobiology 0.059 0.21 0.169 0.036 0.046 -0.03
Nutrition 0.475 0.216 0.202 -0.037 0.085 -0.095
Pharmacology, Toxicology, and Environmental
Health -0.01 0.29 0.058 0.136 -0.095 0.117
OCR for page 19
PAIRWISE CORRELATIONS 19
Table 3-2 correlates median time to degree and average completion rate with GRE
General Test scores and the average number of Ph.D.’s in each program. The correlations
between cohort completion and both average GRE and average PhDs are uniformly low, and in
several fields are negative. The exception is physiology. There is a positive correlation with
respect to median time to degree and both average GRE scores and average Ph.D.’s produced,
but only in nutrition are both strongly correlated. In biomedical engineering and bioengineering
there is a strong correlation between median time to degree and average number of Ph.D.’s, and
in microbiology a strong correlation between median time to degree and average GRE scores.
TABLE 3-2 Correlations of Median Time to Degree and Average Cohort Completion with GRE Scores
and Number of PhDs
Correlation with
Correlation with Median Average Cohort
Time to Degree Completion
Average
Ph.D.’s Average
GRE 2002 to GRE Ph.D.’s 2002
Fields Average 2006 Average to 2006
Biochemistry, Biophysics, and Structural
Biology 0.114 0.140 0.094 0.046
Biomedical Engineering and Bioengineering 0.251 0.491 0.080 -0.011
Cell and Developmental Biology 0.093 0.074 -0.022 -0.022
Genetics and Genomics 0.179 0.074 -0.108 0.235
Immunology and Infectious Disease 0.033 0.050 -0.216 0.051
Integrated Biological and Biomedical Sciences 0.111 0.145 -0.181 -0.033
Microbiology 0.319 0.270 -0.075 -0.089
Neuroscience and Neurobiology 0.156 0.150 0.007 0.076
Nutrition 0.487 0.309 -0.055 -0.106
Pharmacology, Toxicology, and Environmental
Health 0.179 0.038 -0.058 0.103
Physiology 0.223 0.192 0.261 0.295
The correlations in Table 3-3 demonstrate a strong relationship between underrepresented
minority faculty and underrepresented minority students in six of the eleven fields:
Biochemistry, Biophysics, and Structural Biology;
Immunology and Infectious Disease;
Microbiology;
Nutrition;
Pharmacology, Toxicology, and Environmental Health; and
Physiology.
For a fuller discussion of underrepresentation see Chapter 5.
OCR for page 20
20 RESEARCH-DOCTORATE PROGRAMS IN THE BIOMEDICAL SCIENCES
The same relationship does not hold true for gender. The panel found no meaningful
correlation between the percent of female faculty in a program and the percent of female
students; the correlations are below 0.3 in every biomedical science field. The highest correlation
(0.288) is in nutrition. While the average percentage of female students in all fields except
biomedical engineering and bioengineering is over or near 50 percent, this is not the case with
the average percentage of female faculty (see Appendix E). Only in nutrition is the average
percentage of female faculty over 50 percent; the average percentage of female students is over
75 percent. Participation of women in faculty positions in the biomedical sciences is not a new
issue. Women have consistently been represented on the faculty of biomedical fields at a rate
lower than their proportion in the Ph.D. population.2 Thus, although programs with a higher
percentage of minority faculty do indeed seem to attract minority students at a higher rate, the
same is not true for women.
TABLE 3-3 Correlations of Percent Female Students with Percent Female Faculty and
Percent of Non-Asian Minority Students with Percent Minority Faculty
Correlation with Correlation with
Percent Female Percent Non-Asian
Students Minority Students
Percent Female Percent Minority
Fields Faculty Faculty
Biochemistry, Biophysics, and Structural Biology 0.170 0.489
Biomedical Engineering and Bioengineering 0.118 0.076
Cell and Developmental Biology 0.004 0.247
Genetics and Genomics 0.109 0.290
Immunology and Infectious Disease 0.014 0.150
Integrated Biological and Biomedical Sciences 0.227 0.529
Microbiology 0.233 0.765
Neuroscience and Neurobiology 0.204 -0.002
Nutrition 0.288 0.531
Pharmacology, Toxicology, and Environmental
Health 0.187 0.370
Physiology 0.086 0.570
The correlations in Appendix D permit examination of many other relationships among
the characteristics of doctoral programs, faculty, and students. For example, the relationship
between program size (as measured by average number of Ph.D.’s) and research productivity (as
measured by faculty publications, citations, and grant awards) may be of particular interest to
some university administrators and researchers. Although correlation does not imply causation,
2
Research Training in the Biomedical, Behavioral, and Clinical Research Sciences, National Academies Press,
2011,p. 39.
OCR for page 21
PAIRWISE CORRELATIONS 21
it would make sense that, in fields where laboratories are critical to research productivity,
programs with larger laboratories would be more productive—even when measured on a per
capita basis. This is seen in the relationship between the three measures of research productivity
and number of Ph.D.’s, where several fields with higher values for these productivity variables
also tend to have a larger number of Ph.D.’s (see Appendix E).
OCR for page 22
