What Predicts Graduate Public Health Student Success? Evidence for Admission Committees in a Post–Affirmative Action Landscape

Abstract

Objective:

A severe staff shortage and a dearth of professionals from underrepresented backgrounds in the public health workforce are contributing to poor health outcomes in the United States. Schools and programs can mitigate these problems by admitting more graduate public health students overall and from underrepresented backgrounds. We identified predictors of foundational graduate public health course grades and graduate grade point average (GPA), sharing evidence to remove application factors that are admission barriers and do not predict student outcomes.

Methods:

We conducted a linear regression analysis on demographic and academic factors from 564 graduate public health applications for students at the University at Buffalo who received their degree from January 1, 2016, to February 1, 2021, analyzing age, race and ethnicity, sex, income, undergraduate degree, verbal and quantitative Graduate Record Examination (GRE) percentiles, and undergraduate GPA. Outcomes were grades in foundational public health courses and cumulative graduate GPA.

Results:

Undergraduate GPA was the best predictor of graduate public health student success, explaining nearly 7% of foundational public health course grades and 29% of graduate GPA. Higher undergraduate GPA contributed to higher course grades and graduate GPA. GRE scores explained <1% of student outcomes.

Conclusions:

Our findings add to the growing body of research showing that standardized test scores may not predict graduate student outcomes and provide further evidence for the field of public health to consider removing this admission barrier. By doing so, institutions could admit more students to graduate public health programs who can bring needed skills to the market, further diversifying the workforce and public health faculty, to better meet population health needs.

The field of public health has experienced a decades-long workforce shortage, which has existed in both the public and private sectors of the workforce. The shortage has been most severe in the public or government sector consisting of local or state health departments and federal agencies.^1-4 Staff losses in the governmental public health workforce are a threat to the health and well-being of the US population, which the Association of Schools and Programs of Public Health indicates must be addressed.^5,6 In addition to a workforce shortage, there is a skill gap in the existing workforce, where employees have largely been trained in non–public health fields such as nursing or management.^7-10 The COVID-19 pandemic further exacerbated these shortages and skill gaps as public health agencies struggled to maintain skilled staff to perform surveillance, contract tracing, prevention programming, and other critical roles to mitigate the pandemic’s effect.^8,11,12 Public health schools and programs have a responsibility to train students to reduce these shortages and strengthen essential skills needed to respond to future public health crises.

While strengthening skills in the workforce can contribute to programming that improves the public’s health, a workforce that is representative of the population it serves can also better contribute to programs that improve health outcomes.^2,13-15 On average, one-third (33%) of the US population identifies as Black, Hispanic, American Indian, Alaska Native, or Pacific Islander, backgrounds that are underrepresented, ¹⁶ yet the public health workforce is primarily non-Hispanic White. ⁸ The public health pipeline is also not as diverse as the population, with students from underrepresented backgrounds representing one-fifth (19%) of graduate public health enrollments at accredited colleges and universities. ¹⁷ The workforce shortage and skill gaps in public health present challenges to delivering programs that can improve health, and the lack of diversity within the workforce exacerbates these challenges.^5,12,18,19 Considering these factors that pose threats to the public health workforce, health care experts recognize that the recent decision by the US Supreme Court ruling against affirmative action in college admissions ²⁰ may lead to even more challenges in responding to public health problems and, ultimately, to more deaths in the population. ²¹ Schools and programs have an essential role in addressing some of the most pressing workforce challenges of our time. In today’s landscape, it is imperative that admissions committees make evidence-based decisions and admit students who can gain public health skills that employers need and continue to better diversify the workforce and reflect the diversity of the population. ²²

While increasing the number of students admitted into graduate public health programs may add skilled candidates to the workforce, this increase may not directly reduce the workforce shortage in the public sector. Other factors, such as governmental hiring practices, access to loan repayment or forgiveness programs, and career choices of graduates, affect the size of the governmental workforce.^3,4 Knowing which pre-admission factors predict graduate public health student success and using these data to remove admission barriers for students from underrepresented or low-income backgrounds are key to training a workforce that is representative of an increasingly diverse population.^16,19 If Graduate Record Examination (GRE) scores do not predict student success in graduate public health programs, removing them for admission may help further diversify candidates who enter the workforce in both the public and private sector.

Each year, approximately 20 000 students submit applications to accredited graduate public health programs, while on average 12 000 students enroll and more than 10 000 students graduate. ¹⁷ Individual schools and programs reject up to 50% of applicants, and little data explain why. Most institutions collect data on background factors such as age, sex, race and ethnicity, and financial status, as well as academic factors such as type of undergraduate degree, undergraduate grade point average (GPA), and standardized test scores from applicants.^23,24 This information is reviewed by admission committees that generally consist of public health faculty who often do not come from underrepresented backgrounds. ²⁵ A lack of diversity within the student pipeline not only affects diversity in the workforce but also further limits diversity among public health faculty who admit students and train future generations of public health professionals.

Test scores most often submitted with student applications are from quantitative and verbal GREs,^26,27 and many public health institutions seek applicants with a minimum 3.0 undergraduate GPA, although 3.4 is the average of enrolled students.^24,28 These data suggest that admission committees may be rejecting applicants with an undergraduate GPA between 3.0 and 3.4 who could succeed in their programs. Nearly half of public health schools and programs (41%) require or allow students to optionally submit GRE scores,^24,28 even though submitting GRE scores has been shown to be a barrier for students from underrepresented or low-income backgrounds, who on average score 25% lower than other students.^27,29-31 While few published studies have analyzed the GRE’s ability to predict public health student outcomes, several institutions recognize that standardized test scores may be barriers to admission and have recently removed GRE requirements for admission. ²⁴ Institutions that still consider GRE scores when selecting students may be rejecting applicants whose scores are lower than those of the average of admitted students, even though students with lower scores could be just as successful in their programs. If admission committees know which factors predict graduate public health student success, they can admit more students who are most likely to graduate and bring needed skills to the workforce.

Few studies have examined factors associated with graduate public health student success, often demonstrated by grades in foundational courses that address key public health competencies and cumulative graduate GPA, which are essential components of a degree.^32-35 One study in Mexico and one in Poland examined predictors of graduate public health student outcomes, although GRE scores were not applicable in either country. Both studies examined predictive values of demographic and academic factors among master of public health (MPH) students, as well as country-specific standardized test scores. Lamadrid-Figueroa et al ³² examined age, sex, number of children of the graduate student, CENEVAL standardized test scores (Mexico’s graduate standardized test similar to the GRE), and supplemental math test scores from 428 students at the National Institute of Public Health in Mexico. They found that 3 factors combined—CENEVAL scores, math scores, and undergraduate GPA—positively predicted graduate GPA. ³² Panczyk et al ³³ studied 605 MPH student records from the Medical University of Warsaw, finding that high undergraduate GPA was the strongest predictor of high graduate GPA. Having a public health bachelor’s degree from the Medical University of Warsaw also predicted having a higher graduate GPA than students who earned a bachelor’s degree from another institution. ³³ In the United States, Sullivan et al ³⁴ examined the graduate GPA of Boston University’s MPH students before 2019 (when they required GRE scores for admission) and after 2019 (when they removed GRE requirements), finding a similar graduate GPA for both groups. However, this study did not aim to identify predictors of student outcomes. ³⁴ They also found that removing GRE scores increased the enrollment of Black and Hispanic students, providing further evidence that GRE scores are barriers to underrepresented student enrollment. ³⁴ No published studies in the United States offer guidance on factors at time of admission that predict student outcomes in public health, yet a need for these data exists. ²⁴ This study begins to fill that gap in the research.

Methods

This study, conducted at the University at Buffalo (UB), sought to identify pre-admission factors that predict graduate public health foundational course grades and graduate GPA. We performed a secondary data analysis with records from 564 graduate public health students who received their degree from January 1, 2016, to February 1, 2021. UB is a large public research university and member of the Association of American Universities and has been accredited by the Council on Education for Public Health (CEPH), the academic public health accrediting body, since 2008.

In most areas, the sample was generally representative of the larger graduate public health student body. ¹⁷ Consistent with national data, in this study, two-thirds (67%) of students were enrolled in an MPH program and one-third (37%) were enrolled in a master of science (MS), master of arts (MA), or doctor of philosophy (PhD) program. ¹⁷ All students were enrolled in a public health concentration that included biostatistics, bioinformatics, community health and health behavior, environmental health, epidemiology, health services administration, or a customizable MPH that integrates coursework across multiple concentration areas. Rates of rejection, admission, enrollment, and degree completion for graduate public health programs at UB were similar to averages at other accredited schools and programs. ¹⁷ This study was reviewed by UB’s Social and Behavioral Research Institutional Review Board and deemed not human subjects research.

We performed a linear regression analysis that included 4 demographic characteristics: age (continuous, in years), sex (male/female), eligibility for financial aid (yes/no), and underrepresented racial or ethnic background (yes/no). We considered students who identified as Black, Hispanic, American Indian, Alaska Native, or Pacific Islander to be underrepresented. We added 4 academic factors to the model: type of undergraduate degree (bachelor of science or other), quantitative GRE percentile, verbal GRE percentile, and undergraduate GPA. Type of undergraduate degree was dichotomous, and GRE percentiles and undergraduate GPA were continuous. We included type of undergraduate institution in the original analysis; however, we removed this factor as a predictor from subsequent models because we found it was not associated with or did not predict any dependent variables. The analysis examined the extent to which each factor predicted foundational public health course grades and graduate GPA. Another key graduate student outcome is degree completion; however, in this study, 98% of students graduated, leaving little variation in this outcome and making it difficult to detect predictors. For this reason, we did not analyze degree completion. We developed a framework for our analysis (Figure).

Figure 1.

Framework for predictors of success among graduate students in public health. Abbreviations: GPA, grade point average; GRE, Graduate Record Examination.

Foundational course grades came from 5 core courses covering most MPH foundational public health competencies defined by CEPH. Several of these foundational courses were also required courses for UB’s MS, MA, and PhD public health programs, which generally train future public health faculty. Students earn letter grades on a range of F (failing) to A (highest grade). Students must earn grades of B or better to pass each course and most students pass, leading to little variability in letter grades. As a result, letter grades were converted to their equivalent GPA on a scale of 0 to 4.0. A grade of B represents 3.0, B+ represents 3.33, A− represents 3.67, and A represents 4.0. Grades of B− or below, which are failing grades, had an equivalent GPA from 2.67 (B−) to 0 (F).

Graduate GPA was the student’s cumulative graduate GPA at the time of graduation. Given that students need to earn grades of B or better in each course to graduate, most course grades and graduate GPA data skewed high, between 3.0 and 4.0, so we transformed the data for normality by using the inverse distribution function.

We used the SPSS 27.0 statistical analysis program (IBM Corporation, Inc) to conduct all analyses.

Results

The average age of graduate students in the study was 25.7 years. Approximately 70% were female, 25% qualified for financial aid, and just under 8% came from underrepresented backgrounds. Most (71%) applied to graduate school having previously earned a bachelor of science degree. The average undergraduate GPA of students was 3.4, and the average quantitative and verbal GRE percentiles were 57.3 and 63.1, respectively. Most students earned foundational course grades ranging from 3.0 to 4.0 (mean = 3.7 [SD = 0.35]; median = 3.8). The average graduate GPA of the sample was 3.7 (SD = 0.24) and ranged from 2.4 to 4.0. Students with a graduate GPA of 3.0 or higher (98%) earned their degree.

Predictors of Graduate Public Health Student Outcomes

Undergraduate GPA was the best predictor of graduate public health student success, explaining nearly 7% of foundational public health course grades (Table 1) and 29% of graduate GPA (Table 2).

Table 1.

Predictors of foundational public health course grades: results of linear regression analysis on demographic and academic factors from 564 graduate public health applications at the University at Buffalo, 2016-2021

Predictor	Unstandardized B (SE) [95% CI]
Age	0 (0.004) [–0.009 to 0.009]
Sex (men)	–0.039 (0.052) [–0.140 to 0.063]
Eligible for financial aid	0.085 (0.043) [–0.001 to 0.170]
Underrepresented students	0.075 (0.080) [–0.083 to 0.234]
Degree other than bachelor of science	–0.073 (0.047) [–0.166 to 0.021]
Verbal GRE percentile	0.001 (0.001) [–0.001 to 0.004]
Quantitative GRE percentile	0.001 (0.001) [0.002 to 0.003]
Undergraduate GPA a	0.066 (0.066) [0.161 to 0.421]

Abbreviations: GPA, grade point average; GRE, Graduate Record Examination.

^aP value is significant at the .01 level (2-tailed); holding all factors in the model constant, for each unit increase in undergraduate GPA, average course GPA increased by 0.066.

Table 2.

Predictors of graduate GPA: results of linear regression analysis on demographic and academic factors from 564 graduate public health applications at the University at Buffalo, 2016-2021

Predictor	Unstandardized B (SE) [95% CI]
Age	0 (0.010) [–0.008 to 0.030]
Sex (men)	–0.039 (0.107) [–0.380 to 0.042]
Eligible for financial aid	0.085 (0.099) [–0.007 to 0.385]
Underrepresented students	0.075 (0.183) [–0.172 to 0.548]
Degree other than BS	–0.073 (0.103) [–0.327 to 0.078]
Verbal GRE percentile a	0.001 (0.002) [0.001 to 0.009]
Quantitative GRE percentile b	0.001 (0.002) [0 to 0.009]
Undergraduate GPA c	0.291 (0.140) [0.852 to 1.406]

Abbreviations: BS, bachelor of science; GPA, grade point average; GRE, Graduate Record Examination.

^aP value is significant at the .05 level (2-tailed); holding all factors in the model constant, for each unit increase in verbal GRE percentiles, average graduate GPA increased by 0.001.

^bP value is significant at the .01 level (2-tailed); holding all factors in the model constant, for each unit increase in quantitative GRE percentiles, average graduate GPA increased by 0.001.

^cP value is significant at the .01 level (2-tailed); holding all factors in the model constant, for each unit increase in undergraduate GPA, average graduate GPA increased by 0.291.

Higher undergraduate GPA contributed to higher foundational public health course grades: for each unit increase in undergraduate GPA, average course grades increased by 0.066 (Table 1). GRE scores explained <1% of student outcomes in foundational public health course grades.

Verbal and quantitative GRE percentiles were poor predictors of graduate GPA, explaining little of the variance: for each unit increase in verbal or quantitative GRE percentile, the average graduate GPA increased by 0.001 (Table 2). Undergraduate GPA explained most of the variance in graduate GPA: for each unit increase in undergraduate GPA, the average graduate GPA increased by 0.291. In addition, students who had an undergraduate GPA close to 3.0 among the sample earned a graduate GPA of 3.0 or higher and graduated at the same rate as students who were admitted with a higher undergraduate GPA (close to 3.4 or higher).

Discussion

Findings from our study provide data on factors that may predict and—importantly—may not predict outcomes among graduate public health students, offering evidence to admission committees to consider removing GRE scores, which may be a major barrier for underrepresented and low-income students,^31,34 from the admissions process. In our study, GRE scores did not predict foundational course grades and predicted very little of graduate GPA. Undergraduate GPA, however, was a reliable predictor of foundational public health course grades and public health graduate GPA.

GRE scores may not be good predictors of graduate public health student outcomes for students in any graduate public health degree, yet may present barriers for underrepresented and low-income students.^29-31,34 Admission committees that continue to consider them when selecting students are potentially rejecting good candidates from underrepresented or low-income backgrounds who, if admitted, might reduce the workforce shortage and increase diversity in the workforce and among public health faculty.

Undergraduate GPA did predict graduate public health course grades and graduate GPA and is generally one of many factors considered by admission committees. ²⁴ However, in our study, students with an undergraduate GPA of 3.0 graduated from their public health program at rates similar to students with a 3.4 or higher, the average undergraduate GPA of enrolled students. ²⁸ These data suggest that students with an undergraduate GPA close to 3.0 may also have the ability to succeed in a graduate public health program.

Admission committees can use our data to consider evidence-based changes to their admissions practices and reprioritize factors they consider when selecting students. Specifically, our findings support removing GRE scores for graduate public health admissions. Doing so may increase the number of students overall and the number of students from underrepresented and low-income backgrounds who apply to their programs and can succeed in them. By continuing to consider GRE scores, schools and programs may be limiting the number and type of graduate public health students they train, subsequently confining the future public health workforce to its current size, which is insufficient to meet the public’s needs, and its current racial and ethnic make-up, which is not representative of the US population.

Limitations

Our study had several limitations. First, we found little variation in course grades and in graduate GPA because most students earn the high grades needed to complete their program and earn their degree. This limitation may explain why the models did not show additional predictors. Undergraduate GPA was a consistent predictor of graduate public health student outcomes; other factors may also be contributors yet were not detected because of limited outcome variability. Larger sample sizes may be needed to detect other predictors. Second, although the sample in this study was representative of the overall graduate public health student body, our study was conducted at a single institution. UB is representative of other large public universities that offer accredited graduate public health programs, making these findings generalizable to other large public research universities with graduate public health programs. Third, we considered additional analyses that included concentration area as a predictor, but enrollments were low in some concentration areas, so we did not conduct these analyses. However, given limited data on predictors of graduate public health student outcomes, our study offers important contributions to the field. Future research on student outcomes across different types of institutions and concentration areas would be beneficial.

Conclusion

The lack of predictive value of GRE scores, the limited number of factors found to predict student outcomes, and the elimination of affirmative action programs highlight the importance of admission committees using holistic reviews to admit students. Holistic reviews that consider all parts of a student’s application beyond quantitative factors alone, such as undergraduate GPA, might help admission committees select students who are most likely to succeed in their programs, who graduate and bring their skills to the public health workforce, and who diversify the workforce and public health faculty. Although our study explored pre-admission demographic factors and academic background as predictors of student outcomes, future studies could analyze other factors such as work experience, undergraduate degree discipline, choice of graduate degree concentration, or content from personal statements as part of holistic applicant reviews to further inform admissions practices in a post–affirmative action landscape.