Improving Equity Within the Holistic Admission Process Through Race-Neutral Holistic Review: A Before and After Cross-Sectional Study

Abstract

Introduction.

The US Physical therapist (PT) workforce lacks racial and ethnic diversity, which may contribute to health disparities. Holistic review (HR) admission methods may improve workforce diversity by evaluating applicants' academic metrics, experiences, and personal attributes. Holistic review processes must adapt following the 2023 Supreme Court ruling prohibiting race-based admission criteria.

Review of Literature.

Holistic review criteria are program-specific and align with institutional and programmatic goals. While HR is shown to improve student body diversity, most PT education programs heavily emphasize academic metrics, which may disadvantage racial and ethnic minority (REM) applicants from acceptance. This study examines the racial and ethnic distributions of PT education program applicants and matriculants before and after implementing a race-neutral HR rubric where academics and experiences were equally weighted.

Subjects.

The sample included 848 qualified applicants to a PT education program in 2 consecutive admission cycles.

Methods.

This retrospective cross-sectional study analyzed the racial and ethnic distributions of qualified applicants ranked by verbal Graduate Record Exam (vGRE) and HR rubric scores. Chi-square tests assessed differences in racial and ethnic frequencies between qualified applicants Invited to Interview (ITI) or waitlisted for interview and differences between matriculating cohorts.

Results.

There was a statistically significant difference in the frequencies of White/non-Hispanic applicants who were ITI versus waitlisted for interview when ranked by verbal GRE, with White/non-Hispanic applicants more represented in the ITI pool. There were no differences in the racial and ethnic frequencies of applicants who were ITI versus waitlisted for interview after HR implementation, and no differences were found between matriculated cohorts.

Discussion and Conclusion.

Race-neutral HR promotes equitable admissions and reduces bias against REM applicants. Although REM representation increased in the matriculated cohort, adjustments to HR weighting may further enhance diversity. These findings support the broader use of race-neutral HR in PT admissions after the affirmative action ban.

INTRODUCTION

The racial and ethnic diversity of the past, current, and projected future US Physical therapist (PT) workforce is less diverse than the US population, which may contribute to existing health inequities.¹ Researchers have called for holistic review (HR) admission strategies in PT education programs to address the multifaceted issue of health inequities.² Although programs vary in HR procedures, they all follow unified principles requiring selection criteria to include experiences and attributes beyond academic metrics. Each criterion is weighted according to the institution's priorities and must be applied equitably.^3-6 In addition, all programs implementing HR should engage in ongoing assessment to ensure that HR practices are legal, fair, and aligned with current program, institutional, and professional goals.³

Until the recent decision by the Supreme Court of the United States (SCOTUS) on affirmative action in June 2023, race and ethnicity could be considered as one of several admission criteria if clear, mission-aligned program goals regarding diversity-related priorities and educational benefits existed, and only in states where the use of race-based criteria was legal.^6-8 While researchers have shown that previous affirmative action bans at the state level considerably diminished the enrollment of racial or ethnic minority (REM) students in public medical schools, the influence of the recent federal affirmative action ban on HR procedures and outcomes in PT education programs remains unclear.^9,10

Although it is now illegal to use race and ethnicity in college admission–related decisions, the SCOTUS ruling on affirmative action does not preclude institutions from assessing the demographic composition of their student body, or administrators from assessing for and addressing bias within admission procedures. Despite this, individual state regulations and recent federal executive orders rolling back diversity-related initiatives may limit the extent of these practices in higher education. While evidence suggests that admission processes relying heavily on academic metrics may create bias against REM medical school applicants,^11,12 there is little research examining these outcomes in PT education. Although research supports using HR strategies to foster diversity in PT education programs, these findings have limited generalizability due to limitations in admission criteria transparency.^5,13,14 To address this gap, the primary aim of this study was to compare the race and ethnicity distributions of qualified applicants to a PT education program who were ranked and Invited to Interview (ITI) according to traditional review or race-neutral HR in two different admission cycles. A secondary aim was to determine if the HR process resulted in differences in the racial and ethnic distributions of students who enrolled in the PT education program before and after the revised HR process.

LITERATURE REVIEW

Holistic review

Holistic review provides a structure for schools to uniquely and systematically assess a candidate's alignment with program, school, and broader professional goals and to narrow the applicant pool to a qualified and diverse group of accepted candidates.^3-6,14 In an era where considering race and ethnicity in higher education admissions is prohibited, aligning a more inclusive definition of diversity with a program's mission and institutional values may offer a viable approach to achieving diversity-related objectives through race-neutral holistic admission strategies. The American Physical Therapy Association (APTA) adopted a definition of underrepresented minority (URM) in PT education that includes underrepresented REM student populations and individuals from geographically underrepresented areas, lower economic or educationally disadvantaged backgrounds, and people with disabilities.¹⁵ This broader definition of URM in PT education supplies a framework for programs to consider other dimensions of diversity outside of race or ethnicity, such as a history of disadvantage in various social constructs for recruitment, admission, or assessment purposes.² While the URM definition alone does not capture all admission criteria in HR, incorporating URM factors into HR may provide additional pathways to a diverse student body.

Some researchers have reported an association between a background of economic, educational, or geographical disadvantage and REM status in health professions students. Racial and ethnic minority medical school applicants had higher instances of disadvantaged backgrounds and lower economic status than non-REM applicants.¹¹ These findings are similar to research findings in Doctor of Physical Therapy (DPT) students, where REM students had higher frequencies of disadvantaged backgrounds than non-REM students.^13,16 While these findings do not indicate that substituting other URM factors for race or ethnicity is the ideal approach in HR, they suggest that incorporating URM factors may still effectively promote diversity and equitable admissions practices in the post-SCOTUS ruling era.

Holistic review and the Underrepresented Minority Definition

Researchers called for using the expanded URM definition and HR strategies to address the deficits in PT workforce diversity^2,16; however, there is a gap in the literature examining the implementation of HR strategies among PT education programs. These practice recommendations are grounded in survey results that showed programs do not commonly assess and report URM factors outside of race and ethnicity, nor do programs regularly use URM factors as admission criteria.^2,16 Survey results also indicate that programs typically use a mix of academic and noncognitive factors in admissions. Still, they tend to heavily prioritize academic metrics, such as grade point average (GPA) and Graduate Record Examination (GRE) scores, with most programs considering these academic criteria as the key predictors of student success.^2,17 The emphasis on academic metrics in program admission practices may contribute to the growing mismatch between the diversity of the PT workforce and societal needs. Heavily weighting admission in favor of academic criteria could limit diversity by excluding qualified REM applicants.

Evidence points to differences in preadmission academic metrics between REM and non-REM students. In medical education research, REM applicants and first-generation college students were shown to have lower preadmission academic metrics and acceptance rates compared to non-REM or non-first-generation applicants.^11,12,18 These findings are consistent with research on PT program applicants, where undergraduate GPA, GRE scores, and PT program acceptance rates were lower among REM applicants.^19,20 Other research has demonstrated lower GRE scores among URM DPT students with a history of disadvantage than their non-URM peers.²¹ These study results imply that test scores may hinder URM applicants' acceptance into DPT programs when academic metrics dominate admission decisions.

Outcomes of Holistic Review in Doctor of Physical Therapy Education

There is a lack of evidence examining the long-term student and program outcomes of HR. Yet, there is a growing body of evidence describing the diversity-related outcomes of HR. In one study exploring racial and ethnic distribution outcomes of HR in several health professional education programs, researchers reported an increase in REM matriculants in all programs despite varying HR methods.⁴ Other researchers have reported the diversity-related outcomes of HR approaches in PT education programs through the lens of the expanded URM definition, showing that using URM factors other than race or ethnicity in HR led to increased student body diversity.^5,13

Emerging evidence supports integrating the APTA URM definition into HR to foster DPT student body diversity. However, a gap remains in the literature describing best practices in HR procedures in DPT education in this post-SCOTUS era, where using race and ethnicity as admission criteria is unlawful. In addition, there is limited generalizable evidence related to admission criteria selection and weight because of the institution-specific nature of HR and the lack of transparency in HR review procedures. This study aims to address this gap by exploring the demographic outcomes after using a race-neutral HR rubric (holistic rubric) to score and rank qualified applicants compared to scoring and ranking by academic metrics alone, and by providing more transparency to HR criteria and their weights in the admission process.

Subjects

We used a selective, purposive sample²² of qualified candidates who applied to an entry-level residential, 3-year PT education program within a private, graduate health-professions institution in the southwest United States in the 2020–2021 or 2021–2022 admissions cycles. Qualified candidates were individuals who submitted a complete application to the institution through the PT Centralized Application Service (PTCAS), who met or were on target to meet required prerequisites and observation hours, and who met the minimum undergraduate GPA requirements established by the program: 2.9 Science GPA and 3.0 undergraduate cumulative GPA. To be included in the study, an applicant had to be placed in the qualified applicant pool (QAP) by an admissions official for the 2020–2021 or the 2021–2022 application cycles. Data from applicants who did not meet QAP criteria were excluded from the study.

METHODS

This study was approved by the Midwestern University Institutional Review Board as exempt. This non-experimental, observational study used a retrospective before-and-after cross-sectional design and analysis of program admissions records to observe the effect of using the holistic rubric on the demographics of applicants who were ITI versus applicants who were not ITI. The cross-sectional design allowed for analysis of the sample at the point of interest in the application cycle in this study: the ranking and sorting of the QAP.²³ The before-and-after method allowed the researchers to observe the effect of the holistic rubric on QAP sorting outcomes between cycles.²⁴

Admission Process

The QAP was ranked by verbal GRE (vGRE) scores in the 2020–2021 cycle and by holistic rubric scores in the 2021–2022 cycle. In both cycles, the admissions committee extended invitations to interview on a rolling basis in rank order. All applicants who accepted an invitation received 3 standardized virtual interviews, scored by 6 raters and summed. All applicants also completed a scored on-demand essay during the interview day. In the 2020–2021 (vGRE ranking) cycle, total admission ranking scores were calculated with vGRE weighted 15%, total interview score 70%, and essay score 15%. In the 2021–2022 (holistic rubric) cycle, total admission scores were calculated with the holistic rubric weighted at 20%, interview at 70%, and essay score at 10%. Within the holistic rubric, the academic and experiences categories were equally weighted at 10% each. In addition, the interview and essay followed the same standardized questions and scoring procedures in both cycles. Candidates were ranked by total admission score, and rolling offers of acceptance were made in rank order.

Holistic Rubric

To align admission practices with professional recommendations, the program admission committee developed a holistic rubric to score and rank each qualified applicant. The committee considered institutional policies and goals, current evidence, prior student outcomes, the definition of URM in PT education, the Association of American Medical Colleges HR guidelines, and the Experience-Attributes-Metrics model to establish rubric criteria aligned with program, institutional, and professional values.^{2,3,15,20,25-36}

The race-neutral HR rubric had two categories: academic background and experiences, weighted equally. All data evaluated in the HR rubric were available within the standard PTCAS application questions, including the personal statement. Applicants who met the rubric criteria earned points in the specific category of the rubric, whereas applicants who did not meet established criteria or who left specific questions blank did not earn points in that area of the rubric.

The academic background category included scored criteria such as undergraduate cumulative and science GPAs, 12-month grade trend with no C's earned in the last academic year, undergraduate degree focus rigor, and history of prior graduate-level education success such as completing other graduate-level coursework. Submission of GRE scores was optional for applicants and only evaluated in the HR review if both GPAs were below 3.2, earning no GPA points in the rubric. The optional nature of GRE scores and the condition when GRE scores could improve the strength of the application were published on the PT program website.

The experiences category included service history, health care, and employment experience, history as a primary caregiver while pursuing their education, experience of educational, economic, or geographic disadvantage, and other experiential factors aligned with program goals and professional values. Specific definitions of admission criteria related to disadvantaged backgrounds used in the HR review are reported in previously published research that described admission practices at this institution.²¹ Portions of the holistic rubric were pilot-tested to assess the impact of weighing these factors on the demographic distribution of the 2020–2021 QAP, leading to refinements and faculty approval for implementation in the 2021–2022 cycle.⁵

Data Collection

The primary outcome variables for this study were frequencies of racial and ethnic characteristics of the applicants within the QAP groups of the two admission cycles. Undergraduate GPAs and applicant sex were collected for descriptive purposes. The original data source for demographic and academic data was the PTCAS application. Admission counselors verified undergraduate GPA values with official transcripts. Biological sex and race or ethnicity were self-selected responses that applicants completed within the PTCAS application, and while not verified, all applicants signed an attestation of application accuracy.³⁷ We obtained deidentified data from admissions officials for applicants within the QAP for each admission cycle within this study, including the final application status of the applicant, self-reported race and ethnicity, undergraduate cumulative and science GPAs, and biological sex. Because no rejections were made at this stage, the applicants of the QAPs were sorted into either the Waitlist for Interview (WLI) group, consisting of all applicants who did not receive an invitation to interview by the end of the interview cycle, or the ITI group consisting of all applicants who received an invitation to interview at some point in the cycle, including applicants who declined an interview, for each admission cycle.

Data Analysis

An a priori sample size estimate was completed using G*Power 3.1.9.7 (Heinrich-Heine-University Düsseldorf, Düsseldorf, Germany). A chi-square contingency table test was completed with a large effect size (w = 0.50), an alpha of 0.05, a power of 0.95, and an estimated degrees of freedom of 3. This was based on historical cohort racial and ethnic frequencies and the minimum cell count required for analysis, yielding a minimum sample size of 69.^38,39 To preserve the demographic characteristics of the QAP and reduce the potential for selection bias, all applicants who met the study criteria were included in the study. A total of 441 applicants were excluded from the study for not meeting qualifying GPA requirements, 238 and 203 from each cycle, respectively.

The primary aim of this study was to examine for differences in the racial and ethnic distributions of applicants within the two independent groups of the QAP for each application cycle, and a secondary aim was to examine differences in the racial or ethnic distributions of applicants who ultimately matriculated into the cohort in each admission cycle. Descriptive and inferential statistics for each research aim were calculated using SPSS version 28.0.1 (IBM Corp., Armonk, NY). To examine differences in the racial and ethnic distributions between independent groups for both aims, a chi-square test of homogeneity using the crosstabs function within SPSS was completed, with significance accepted for a two-sided P < .05.^39-41 The dependent variable for the first two analyses was QAP status, a nominal, dichotomous variable grouped as ITI versus WLI. The dependent variable for the third analysis was the matriculating cohort year, dichotomized as 2021 versus 2022. The independent variable for all analyses was the polytomous variable of race or ethnicity, grouped based on categories from the PTCAS application, with low-frequency (<5) categories combined into the “Other Underrepresented Minorities” category, and applicants with “unknown” race or ethnicity were removed from analysis to satisfy the sample size assumptions of the chi-square test.^39,41 All assumptions for the tests were satisfied for the 3 analyses. When indicated, post hoc testing was completed to identify the location of differences, using multiple z-tests of two proportions, and a Bonferroni adjustment was made to reduce the incidence of type I error associated with multiple comparisons, accepting significance at P < .0125.^42-44

RESULTS

Descriptive Analysis

Admissions data from 848 qualified applicants, 437 in the 2020–2021 (ranked by vGRE) cycle QAP and 411 in the 2021–2022 (ranked by holistic rubric score) cycle QAP, were analyzed by cycle (Table 1). In both admission cycles, most of the QAP was female (62% and 60.1%) and White/non-Hispanic (64.5% and 60.1%). The mean GPAs of the QAPs were similar between both cycles (Table 1).

Table 1.

Demographic and Academic Characteristics of the Qualified Applicant Pools

Factor	2020-2021 Cycle n = 437	2021-2022 Cycle n = 411
	n (%)	n (%)
Sex (n = 848)
Male	166 (38)	164 (39.9)
Female	271 (62)	247 (60.1)
Race/Ethnicity (n = 848)
American Indian	2 (.5)	2 (.5)
Asian	50 (11.4)	46 (11.2)
Black	12 (2.7)	11 (2.7)
Hispanic	66 (15.1)	65 (15.8)
Native Hawaiian	0	2 (.5)
Pacific Islander	0	1 (.2)
Two or more races	23 (5.3)	18 (4.4)
White	282 (64.5)	263 (64)
Race/Ethnicity unknown	2 (.5)	2 (.5)
Academic metrics	M (SD)	M (SD)
Undergraduate cGPA	3.50 (.25)	3.55 (.24)
Undergraduate sGPA	3.31 (.31)	3.37 (.32)

Abbreviations: cGPA = cumulative grade point average; sGPA = science grade point average (courses in biology, chemistry, anatomy, and physiology).

Table 2 presents each cycle's demographic and academic characteristics according to the ITI and WLI groups. In the 2020–2021 cycle, when ranked by vGRE, the proportion of White, non-Hispanic applicants in the ITI pool was 70.2%, and 56.4% in the WLI pool. When ranked by holistic score in the 2021–2022 cycle, White, non-Hispanic applicants comprised 64.6% of the ITI pool and 61.9% of the WLI group. The difference in White, non-Hispanic applicant representation between the ILI and WLI pools was 13.8% when ranked by vGRE (2020–2021 cycle) and 2.7% by holistic rubric score (2021–2022 cycle).

Table 2.

Demographic and Academic Characteristics of the ITI and Waitlisted for Interview Pools

	2020-2021 Cycle QAP Ranked by vGRE		2021–2022 Cycle QAP Ranked by HR Rubric
Factor	ITI Pool n = 258 n (%)	WLI Pool n = 179 n (%)	ITI Pool n = 314 n (%)	WLI Pool n = 97 n (%)
Sex
Male	99 (38.4)	67 (37.4)	120 (38.2)	44 (45.4)
Female	159 (61.1)	112 (62.6)	194 (61.8)	53 (54.6)
Race/Ethnicity
American Indian	2 (.8)	0	1 (.3)	1 (1)
Asian	22 (8.5)	18 (15.6)	31 (9.9)	15 (15.5)
Black	6 (2.3)	6 (3.4)	8 (2.5)	3 (3.1)
Hispanic	31 (12)	35 (19.6)	53 (16.9)	12 (12.4)
Native Hawaiian	0	0	1 (.3)	1 (1)
Pacific Islander	0	0	1 (.3)	0
Two or more races	15 (5.8)	8 (4.5)	15 (4.8)	3 (3.1)
White	181 (70.2)	101 (56.4)	203 (64.6)	60 (61.9)
Race/ethnicity unknown	1 (.4)	1 (.6)	1 (.3)	2 (2.1)
Academic metrics	M (SD)	M (SD)	M (SD)	M (SD)
Undergraduate cGPA	3.54 (.26)	3.44 (.21)	3.61 (.23)	3.37 (.18)
Undergraduate sGPA	3.36 (.33)	3.23 (.25)	3.45 (.31)	3.13 (.19)

Abbreviations: cGPA = cumulative grade point average; HR = holistic review; ITI = invited to interview; sGPA = science grade point average (courses in biology, chemistry, anatomy, and physiology); QAP, qualified applicant pool; vGRE, verbal Graduate Record Exam; WLI, waitlist for interview.

Table 3 presents the demographic and academic characteristics of the matriculating cohorts. The representation of REM students was slightly higher in the cohort ranked by holistic score (40.74%) than in the cohort ranked by vGRE (37.04%). The mean cumulative GPA of the cohort ranked by vGRE (3.39) was slightly lower than the mean GPA of the cohort ranked by holistic rubric score (3.51).

Table 3.

Demographic and Academic Characteristics of the Matriculated Cohorts

Factors	Percent of Matriculating Cohort (%)
	2020-2021 Cohort n = 54	2021-2022 Cohort n = 54
Race/Ethnicity
American Indian	1.8	0
Asian	9.26	12.96
Black	3.7	3.7
Hispanic	14.81	16.67
Native Hawaiian	0	0
Pacific Islander	0	0
Two or more races	7.41	7.41
White	62.96	59.26
Race/ethnicity unknown	0	0
Academic metrics	M(SD)	M(SD)
Undergraduate cGPA	3.40 (.29)	3.52 (.22)
Undergraduate sGPA	3.22 (.36)	3.31 (.30)

Abbreviations: cGPA = cumulative grade point average; sGPA = science grade point average (courses in biology, chemistry, anatomy, and physiology).

Analysis of Qualified Applicants Ranked by Verbal Graduate Record Exam

Two qualified applicants with “unknown” race and ethnicity were excluded from analysis; therefore, race and ethnicity data from 435 qualified applicants in the 2020–2021 cycle were analyzed with a chi-square test to assess for differences in the racial or ethnic distributions of qualified applicants in the ITI pool versus the WLI pool who were ranked and invited according to vGRE score. The multinomial distributions of race and ethnicity (White, Asian, Hispanic, and Other Underrepresented Minorities) were not equal between the ITI and WLI groups, χ²(3) = 11.892, P = .008 (Table 4). Post hoc testing revealed a statistically significant difference in the proportion of White applicants in the ITI (70.2%) group versus the WLI group (56.4%), P = .003 (Table 5). There were no significant differences in the proportion of Asian, Hispanic, and Other URM applicants between the ITI and WLI groups.

Table 4.

Pearson Chi-Square Results

Comparison	Pearson Chi-Square Value	Degrees of Freedom	Significance, 2 Sided (P)
ITI vs WLI when ranked by vGRE (2020–2021)	11.892a	3	.008
ITI vs WLI when ranked by holistic rubric (2021–2022)	3.119b	3	.374
Matriculated cohorts 2021 vs 2022	.530c	3	.912

Abbreviations: ITI = invited to interview; WLI =, waitlist for interview.

^a0 cells have expected cell counts <5. The minimum expected cell count is 15.14.

^b0 cells have expected cell counts <5. The minimum expected cell count is 7.92.

^c0 cells have expected cell counts <5. The minimum expected cell count is 6.

*P < .05.

Table 5.

Post hoc Pairwise Comparisons Using Multiple Z-Tests of 2 Proportions

Race/Ethnicity n = 435	Pearson Chi-Square Value	Significance, 2 Sided (P)
White	8.704	.003a
Asian	5.280	.022
Hispanic	4.683	.030
Other underrepresented minorityb	.163	.686

Abbreviations: vGRE = verbal Graduate Record Exam.

^aSignificance accepted for P < .0125 (Bonferroni adjustment).

^bRacial subgroups with n < 5 within each pool were combined to meet sample size requirements.

Analysis of Qualified Candidates Ranked by Holistic Rubric Score

Three qualified applicants with “unknown” race and ethnicity were excluded from this analysis. Therefore, race and ethnicity data from 408 qualified applicants were analyzed with a chi-square test to assess for differences in the racial or ethnic distributions of qualified applicants in the ITI versus the WLI pool, ranked and sorted using the holistic rubric score. The multinomial distributions of race and ethnicity (White, Asian, Hispanic, and Other Underrepresented Minorities) were not significantly different between the ITI and WLI groups χ²(3) = 3.119, P = .374 (Table 6). Observed frequencies and percentages of the race and ethnicity of applicants within each applicant group are presented in Table 6.

Table 6.

Cross-tabulation of Race and Ethnicity When Qualified Applicants Were Ranked by HR Rubric Score

Race/Ethnicity n = 408	Qualified Applicant Pool
	ITI Frequency (%)	WLI Frequency (%)
White	203 (64.9)	60 (63.2)
Asian	31 (9.9)	15 (15.8)
Hispanic	53 (16.9)	12 (12.6)
Other underrepresented minoritya	26 (8.3)	8 (8.4)
Total	313 (100)	95 (100)

Abbreviations: HR =, holistic review; ITI = invited to interview; WLI = waitlist for interview.

^aRacial subgroups with n <5 within each pool were combined to meet sample size requirements.

Analysis of Matriculating Cohorts

One hundred eight total qualified applicants ultimately matriculated into the PT education program in 2021 and 2022. A chi-square test was completed to examine for differences in the racial and ethnic distributions of each matriculating cohort. The multinomial distributions of race and ethnicity (White, Asian, Hispanic, and Other Underrepresented Minorities) were not significantly different χ² (3) = .530, P = .912 (Table 4).

DISCUSSION AND CONCLUSION

Admission Ranking Procedures

The difference in the distribution of White/non-Hispanic applicants in the ITI pool and WLI pools in the 2020-2021 admission cycle was statistically significant. However, there was no difference between the ITI and WLI pools in the 2021–2022 cycle. These results suggest that the admission procedure of ranking the QAP by vGRE created a disadvantage for students from REM groups, despite being otherwise qualified. These results also suggest that the HR rubric resulted in a more equitable ranking process despite being race-neutral. The decision to implement a race-neutral HR process before the SCOTUS ruling was based on the program goals of creating a sustainable HR process that could navigate the political uncertainty of affirmative action while addressing perceived admission inequities. The tradition of using vGRE as QAP ranking criteria in this program was based on trends in PT program admissions, historical program precedence, and evidence showing a relationship between GRE scores and DPT student success while providing a method to narrow a large applicant pool quickly.^20,29,45-47 Although evidence supports using GRE scores as admission criteria in PT education programs, there is conflicting evidence regarding the importance of these criteria in predicting student success.^{29,32,35,48,49} In the 2020–2021 cycle, substantial importance was placed on the single metric of vGRE, which may have contributed to the study results, given the known test score disparities between REM and non-REM students.^11,13,19,20

Matriculated Cohort Diversity

There was no statistical difference in the racial and ethnic distributions of the matriculated cohorts between admission cycles despite the 2021–2022 cycle sample having slightly greater REM representation. Both matriculated cohorts have lower White/Non-Hispanic representation (62.9% in 2021 and 59.2% in 2022) than the national averages reported in PT aggregate program data for private institutions in the same years (72.1% and 69.6%).⁵⁰ These results may point to admission procedures that foster diversity, but other factors could influence these findings outside the HR rubric. Moreover, the minimal difference between cohorts could be related to the relatively lower weight of the HR rubric in the overall admissions score. Although the single HR rubric score was equally weighted between academics (50%) and experiences (50%) at the sorting stage of the QAP, in the final ranking stage for program acceptance, the total HR rubric score was weighted lower (20%) compared to other factors; therefore, its utility in fostering diversity at this stage in the admissions cycle may be limited.

Holistic Review Position

The results of this study also highlight the importance of the placement of HR review procedures in the admission process to promote equitable admissions. Despite the resources required to complete an HR, practice recommendations are for applying HR strategies equitably to applicants.^3,26 In the 2021–2022 cycle, HR was applied early in the application process so that every qualified candidate received a holistic file review. This contrasts with the prior cycle, where HR strategies were only applied to the interview and writing sample and, therefore, limited to applicants who completed the interview process. This study showed no difference in the distribution of race/ethnicity populations between the ITI and WLI groups when ranking applicants by HR rubric score, suggesting that the pool was narrowed equitably.

When the QAP was ranked and sorted by HR, the ITI pool's mean GPAs were slightly higher compared to the mean GPAs when ranked and sorted by vGRE, although the significance of these findings was not explored. These results provide emerging evidence that the HR rubric may have preserved the diversity of the QAP during the sorting phase without compromising academic quality. These results support previous research in DPT education, where researchers compared the diversity and academic metrics of the QAP as ranked by HR score to the same QAP ranked by academic metrics and found that the HR led to greater diversity in the top rankings of the QAP with similar GRE and GPA mean scores in both sorting methods.⁵

Limitations

The findings of this study should be understood in light of its limitations. Five cases of unknown REM classification were excluded from the analysis, which could limit the findings; however, this represents less than 1% of the sample. The study design and its sampling methods also limit the results to observational findings, and their generalizability is restricted.^51,52 In addition, this study was based on secondary data that included unverified self-reported demographic factors obtained from the PTCAS application. Although the PTCAS application is unverified, applicants sign an attestation of accuracy before submission. Despite this, some applicants may have not earned HR points for background factors due to incomplete applications.

The intervention of the HR rubric was implemented at the same time as removal of GRE as an admission requirement, which makes it difficult to discern the effect of the rubric. However, during rubric pilot testing, we examined the effect of sorting the QAP based on GPA alone, which showed that the unequal representation of REM applicants between the QAP and ITI pools was not remedied with removal of the GRE alone.

Also, the number of applicants in the QAP decreased between cycles, and the proportion of the QAP ITI increased between cycles. Although this may limit the study's results, these findings are consistent with national DPT admission trends in the post-COVID era, in which application volumes declined, and applicants had greater school choice due to the ongoing proliferation of DPT programs nationwide.

Implications for Practice and Future Research

To our knowledge, this was the first study in PT education to demonstrate that a race-neutral HR rubric can lead to more equitable demographic distributions of REM applicants during qualified candidate sorting procedures. This is also one of the very few studies that demonstrate the integration of the definition of URM in PT education into admission criteria. Although the race-neutral HR process was implemented before the SCOTUS ruling, this study's results support practice change in HR methods in the postaffirmative action era, showing that using alternative admission criteria can maintain the racial and ethnic diversity of the QAP, promoting equitable admissions and mitigating bias.

The results of this study also provide a framework for physical therapy administrators to address bias in their admissions procedures, which may ultimately address health inequities. Although we focused on bias within the QAP sorting procedures, the use of minimum GPA requirements for qualification or use of interview procedures may create additional bias not examined in this study. Future research that examines interventions for bias associated with program minimum academic requirements or interview procedures is warranted.

While HR methods may foster increased diversity in PT education, the influence of HR on student academic and program outcomes such as National Physical Therapy Examination success is unknown. It is also unknown how HR methods ultimately influence workforce and patient outcomes.

CONCLUSION

This study showed a significant difference in the distribution of White/non-Hispanic qualified applicants who were or were not ITI after admission ranking procedures that relied on academic metrics alone. There was no significant difference after implementing a race-neutral HR rubric score to rank the QAP. These findings support using race-neutral holistic scoring methods to foster diversity and promote equitable admission practices when the use of race or ethnicity in admission decisions has been deemed unconstitutional at the federal level.