Disclosure: The authors declared no conflict of interest.
Weight bias in graduate school admissions
Article first published online: 20 JUN 2013
Copyright © 2012 The Obesity Society
Volume 21, Issue 5, pages 918–920, May 2013
How to Cite
Burmeister, J. M., Kiefner, A. E., Carels, R. A. and Musher-Eizenman, D. R. (2013), Weight bias in graduate school admissions. Obesity, 21: 918–920. doi: 10.1002/oby.20171
- Issue published online: 20 JUN 2013
- Article first published online: 20 JUN 2013
- Accepted manuscript online: 30 NOV 2012 09:03AM EST
- Manuscript Accepted: 6 NOV 2012
- Manuscript Received: 19 JUL 2012
Whether weight bias occurs in the graduate school admissions process is explored here. Specifically, we examined whether body mass index (BMI) was related to letter of recommendation quality and the number of admissions offers applicants received after attending in-person interviews.
Design and Methods:
Participants were 97 applicants to a psychology graduate program at a large university in the United States. They reported height, weight, and information about their applications to psychology graduate programs. Participants' letters of recommendation were coded for positive and negative statements as well as overall quality.
Higher BMI significantly predicted fewer post-interview offers of admission into psychology graduate programs. Results also suggest this relationship is stronger for female applicants. BMI was not related to overall quality or the number of stereotypically weight-related adjectives in letters of recommendation. Surprisingly, higher BMI was related to more positive adjectives in letters.
The first evidence that individuals interviewing applicants to graduate programs may systematically favor thinner applicants is provided here. A conscious or unconscious bias against applicants with extra body weight is a plausible explanation. Stereotype threat and social identity threat are also discussed as explanations for the relationship between BMI and interview success.
Individuals with excess body weight face bias and discrimination in a wide variety of settings (1). One unstudied realm in which weight bias may occur is graduate school admissions. Although few, several studies have investigated the impact obesity might have on education attainment. Two early studies by Canning and Mayer (2, 3) showed that despite equivalent performance in high school, students with obesity were less likely to be accepted to elite universities; the authors speculated that these differences may have been due to biases against obese students by educators. Additionally, studies have suggested that parents affect education attainment by being less likely to provide financial support for their heavier daughters (4). Several studies have used large national survey samples to study weight bias in education attainment while controlling for relevant covariates. They have shown that that adolescents who are obese are less likely obtain a college degree (5). This relationship seems especially strong for females as two other surveys found that even after controlling for relevant covariates, adolescent girls who are obese are less likely to enter college (6) and finish fewer years of school over all (7). Considered together, these studies suggest there is a link between obesity and obtaining a college education.
Weight bias in graduate school admissions could manifest in two forms that are not necessarily part of the undergraduate admissions process: letters of recommendation and in-person interviews. For example, letters of recommendation written for female applicants to academic positions were found to have more stereotypical feminine adjectives (8), fewer standout adjectives (e.g., excellent, unparalleled; (9)) were shorter, and included more negative phrases (10) compared with males' letters. Indeed, past research has suggested a pattern of bias against women (although not necessarily based on weight) in academic hiring despite equal qualifications (11). Additionally, experiencing weight stigma from teachers and professors was reported by 30% of women in one study (12) and several studies have shown that educators have lower expectations for individuals with obesity (13). Thus, it is possible that professors writing letters for graduate school applicants exhibit similar biases. Additionally, applicants could be disfavored because of their weight during in-person interviews, as has been found in research on employment and hiring (14).
This study tested potential sources of weight bias in graduate school admissions. Our first hypothesis was that weight bias would be evident in the letters of recommendation applicants submitted to graduate programs in psychology. We predicted that the number of stereotypically weight-related adjectives (e.g., hardworking, intelligent), standout adjectives (e.g., excellent, outstanding), positive personality descriptors (e.g., caring, friendly), and the quality of applicants' letters would decrease as body mass index (BMI) increased. Hypothesis 2 was that applicants' BMI would negatively predict their success in the interview process.
Methods and Procedures
Applicants (N = 274) to the psychology graduate programs (i.e., clinical, industrial/organizational, developmental, neural/cognitive) of a large Midwestern university in 2010 were contacted via email in the summer of 2011 to participate in a study about graduate admissions. See Table 1 for participant information. As an incentive, participants were entered into a drawing for four $25 gift cards. Procedures were approved by the university human subjects review board. Of the 274 applicants, 97 (35.3%) responded. Applicants came from 23 different US states and 3% were international applicants.
|Total samplea||Normal weight||Overweight||Obese||Male||Female|
|(N = 97)||(BMI < 25; n = 54)||(25 ≤ BMI < 30; n = 29)||(BMI ≥ 30; n = 14)||n = 45||n = 51|
|Type of program|
|Quantitative academic history|
In an online survey, participants provided consent and information about all the applications they submitted, including whether they received interviews (in-person or phone) and offers of admission, rejection, or withdrew their application. Participants also reported their height and weight.
To assess academic strength, data were obtained from participants' applications including letters of recommendation (N = 276), graduate record exam (GRE) scores, and undergraduate grade point average (GPA). Letters of recommendation were cleared of identifying information before being coded by two graduate students (6% of the letters were coded twice for reliability). For the portion of letters coded by both coders, the average of the ratings was used for analyses. Number and type of positive work-related adjectives (e.g., amazing, outstanding; inter-rater r = 0.98), positive personality characteristics (e.g., caring, friendly; inter-rater r = 0.96), and stereotypically weight-related adjectives (e.g., hardworking, intelligent; inter-rater r = 0.96) were recorded for each letter. A list of words for each category was based on words used in previous research (8–10), words related to common weight stereotypes, and by reviewing other psychology doctoral letters of recommendation not included in the sample of coded letters. Negative statements (e.g., awkward speaker, nervous; inter-rater r = 0.52) and total letter length (inter-rater r = 0.92) were also recorded. Inter-rater reliability for negative statements was likely low because negative statements were quite uncommon in letters which are by nature positive appraisals (only 19 negative statements in 276 letters). Coders scored letters subjectively on overall quality from 1 (very poor) to 5 (very strong) based on the letters' performance on the other five categories and an overall assessment (inter-rater r = 0.78).
A summary of participants' data is presented in Table 1. Hypothesis 1, that letter quality and weight bias-related wording would be inversely related to applicant BMI was not supported. Participants' BMI was unrelated to letter quality (r = −0.01, P = 0.45), letter length (r = −0.12, P = 0.13), positive personality traits (r = −0.04, P = 0.36), stereotypically weight-related adjectives (r = 0.03, P = 0.37), or negative comments (r = 0.22, P = 0.50). Contrary to the hypothesis, BMI was significantly positively related to the number of standout adjectives (e.g., excellent, outstanding; r = 0.21, P = 0.02). Taken together, these results suggest that letter writers were not biased against applicants based on BMI.
Hypothesis 2 that BMI would predict post-interview offers of admission was tested using hierarchical linear regression. Correlation analyses tested for relevant control variables. BMI was not related to demographics (i.e., age, race, and gender), GRE score, number of applications, or number of interviews. BMI was related to undergraduate GPA (r = −0.20, P = 0.02); thus, GPA was controlled for in step 1 of the regression; BMI was entered in the second step. Hypothesis 2 was supported. BMI significantly predicted number of post-interview offers of admission, β = −0.27, t(54) = −2.04, P < 0.05, accounting for a significant proportion of variance in offers of admission, ΔR2 = 0.07, F(1,57) = 4.2, P < 0.05. A second regression showed BMI did not significantly predict total offers of admission without in-person interviews β = −0.01, t(54) = −0.10, P < 0.918. Together, these results suggest that BMI was a factor in applicant admission only when in-person interviews were conducted.
Additionally, an ANOVA tested whether categorical BMI was related to interview success. Results showed that overweight/obese applicants (BMI ≥ 25) received fewer offers of admission (M = 0.69, SD = 0.78) after in-person interviews than normal weight applicants [BMI < 25; M = 1.20, SD = 0.72; F(1,58) = 6.17, P = 0.016].
Post hoc analyses revealed that the relationship between BMI and interview success was significant for females (r = −0.39, P < 0.05) but not for males (r = −11, P = 0.58), suggesting that BMI may be a more important factor in interview success for female applicants.
This study examined weight bias in the graduate school admissions process. There was no evidence that applicant BMI negatively influenced the quality of letters written on their behalf. In fact, higher BMI was significantly related, albeit weakly, to more standout adjectives (e.g., excellent, outstanding) in the candidates' letters. It is plausible that letter writers, who had developed meaningful relationships with applicants, looked beyond superficial qualities, such as weight, to comment on the actual quality of the applicants' work.
Despite similar credentials and equally or slightly more positive letters of recommendation, participants who were heavier received fewer offers of admission to graduate schools in psychology following in-person interviews. Consistent with prior research, there was some evidence that this effect was stronger in female applicants (4). Past research has suggested a similar pattern of bias against women in academic hiring (11). One plausible explanation for this finding is weight bias. For example, heavier applicants may have received lower performance ratings by interviewers or were deemed less favorable because of their weight. Further, conscious or unconscious concerns about a candidate's work ethic or self-discipline could easily shift the dynamics of the interview process. Previous research has shown a similar effect for body weight in experimental studies that manipulated the weight of job applicants in simulated hiring scenarios (14).
An alternative explanation is that weight-related factors could have actually compromised applicants' interview performance. Substantial research finds that the performance of stigmatized groups can be compromised in contexts where their social identity is devalued (i.e., social identity threat; (15)) or stereotypes of the stigmatized group are likely to be relevant (i.e., stereotype threat; (15)). If heavier applicants felt devalued or that their credentials might be interpreted through the lens of weight-based stereotypes, their interview performance could have been diminished.
A potential limitation is that the number of participants is modest and the total number of participants with obesity is smaller still. However, several analyses were based upon a total of 954 applications and 134 in-person interviews. A second limitation is that we did not assess socio-economic status which could have varied with BMI and been a factor that influenced interviewers' opinions. Additionally, the accuracy of participants' self-reported weight and height as well as the number of in-person interviews and offers could not be verified. Finally, because this study was conducted with applicants to a doctoral program in psychology, the findings might not generalize to other disciplines. However, when compared with those in other fields, psychology faculty might be expected to be more aware of the cognitive influences or heuristics that can bias interviews. Therefore, these findings may actually be a conservative estimate of true weight-based discrimination in the graduate admissions process.
Weight-based discrimination has been documented in employment and pre-baccalaureate admissions (1). The graduate admissions process does not appear to be immune to similar effects. While graduate programs strive to admit the most qualified candidates, bias in the admissions process can threaten the integrity of the graduate admissions system and limit opportunities for individuals with higher BMIs. Further research is needed in this important area.