Journal of Undergraduate Ethnic Minority Psychology – ISSN 2332-9300

Volume 3, 2017 6

Effects of Race and Ethnicity on Mental

Health and Help-Seeking amongst
Undergraduate University Students
Joy Twentyman1, Matthew Frank, MSW/MPH, Linda Lindsey, PhD

Washington University in St. Louis

Abstract Mental health disorders present a significant health challenge to college students. Furthermore, prevalence rates and
treatment rates differ across races and ethnicities, tending to burden minority groups with worse mental health and less treatment.
This research was interested in two questions: (1) the effects of race and ethnicity on depression and anxiety and (2) the effects of
race and ethnicity on mental health help-seeking behaviors amongst undergraduates at a medium-sized Midwestern university. It
was hypothesized that non-White individuals will report higher rates of symptoms of depression and anxiety and will seek help for
mental health problems less often than their White peers. Results found that rates of symptoms of depression and anxiety did not
differ across race or ethnicity, but rates of help-seeking behaviors did differ across race and ethnicity. Finally, the concordance
between the racial/ethnic compositions of student bodies and that of campus mental healthcare providers was measured for the top
twenty universities in the United States. These topics are important because mental illness affects not only students’ ability to
succeed in school, but also their overall well-being. Furthermore, mental illness can be fatal: treatment, whether by counseling or
by chemicals, is the best way to prevent mortality.

Index Terms— mental health, race, ethnicity

need for greater education about mental health and

INTRODUCTION treatment accessibility on campuses. According to the

M
same 2015 ACHA survey, when asked about what types of
ental health disorders present a significant health issues students have received information from their
challenge to college students. According to data collected schools, results were somewhat positive but need
by American College Health Association (ACHA) in a improvement. For example, 65.3% have received
survey of 74,438 undergraduate students at 108 colleges information about stress reduction, 60.7%
and universities in the spring of 2015, over the last 12 depression/anxiety, and 49.4% suicide prevention. While
months, 13.2% were diagnosed with depression and 15.8% gains have been made (in spring 2010, the percentages
with anxiety. These statistics are enlightening because they receiving information about stress reduction,
demonstrate the broad effect of mental illnesses on depression/anxiety, and suicide prevention were 53.5%,
students’ lives. Over the last 12 months, 16.1% felt so 46.9%, and 30.1%, respectively), schools could be
depressed it was difficult to function, 6.2% seriously communicating about mental health more effectively: for
considered suicide, and 1.1% attempted suicide. Failure to comparison, 80.3% have received information about
recognize or treat depression and other mental illnesses can alcohol/drug use, 78.8% sexual assault/relationship
be fatal. At the same time, statistics alone cannot fully violence prevention, and 52.1% cold/flu/sore throat.
communicate the daily challenges faced by many more Education alone, however, is not sufficient. While 18.5%
students: affected grades, difficulty paying attention in of students report ever having utilized their institutions’
class, trouble maintaining social connections, insomnia and mental health services, these rates are known to vary by
fatigue, decreased confidence and self-esteem, etc. race and ethnicity. Though 13.2% were diagnosed with
The sobering nature of these data emphasizes the depression, 10.5% were treated. For anxiety, 15.8% were
diagnosed and 11.7% treated. While the numerical

Manuscript received June 2016, resubmitted and accepted October 2016.

1
Joy Twentyman graduated from Washington University in St. Louis in May of 2016 with a B.A. in biochemistry/molecular biology and applied mathematics.
(jtwentyman@wustl.edu)
 Journal of Undergraduate Ethnic Minority Psychology - ISSN 2332-9300 7
Volume 3, 2017

differences seem small, if the sample is taken to be critical” in better mental health amongst minority groups
representative of all four year students in the United States, (Halpern, 1993). Social support is beneficial to maintaining
this translates to approximately 278,451 untreated cases of good mental health, and minority individuals—particularly
depression and 421,029 untreated cases of anxiety (U.S. those in areas of lower cultural similarity—are thus at
Department of Education, 2014). In comparison, the increased risk.
population of St. Louis city is 319,365 (U.S. Census
Bureau, 2010). These statistics are significant not only in Race, Ethnicity, and Help-Seeking
terms of decreased health and productivity, but also
because, as stated above, mental illness—especially if There are three main components in the process of
help-seeking: recognizing need to seek help, deciding to
unattended—can be fatal.
seek help, and continuing to seek help, if appropriate. The
first two steps are largely mediated by cultural perceptions
Literature Review
(i.e., stigma or lack thereof), while the final step is affected
by communication with and the cultural competency of the
Race, Ethnicity, and Prevalence
professional (Hwang et al., 2008).
According to statistics released by the Substance
Stigma is defined as “negative attitudes, beliefs,
Abuse and Mental Health Services Administration
thoughts, and behaviors that influences the individual, or
(SAMHSA), rates of mental illness vary across race and
the general public, to fear, reject, avoid, be prejudiced, and
ethnicity: 19.0% of whites, 16.8% of blacks/African
discriminate against people with mental health disorders”
Americans, 22.7%, Asians 13.4% of American
(Gary, 2005). This concept is underpinned by symbolic
Indians/Alaska Natives, 24.9% of people of two or more
interactionism: behavior does not occur in a vacuum, but in
races, and 15.3% of Hispanics report experiencing any
the context of interaction with others, both at societal and
mental illness in the past year (Substance Abuse and Mental
interpersonal levels. In particular, though stigma exists in
Health Services Administration, 2015). The variance is
all races, ethnicities, and cultures, “double stigma”
noteworthy, even before accounting for further
particularly affects minority groups due to institutional
discrepancies due to culturally-affected differences in
discriminatory practices and mindsets affecting all levels of
reporting and diagnosis. Many influences inform the
society from politicians to clinicians (Gary, 2005).
disparate prevalence rates, including but not limited to
Additionally, some minority groups may hold differing, if
genetic and epigenetic variants, geographic area, diet,
not stronger, stigmatizing attitudes towards mental illness
family background, socio-cultural factors. In the scope of
than Whites. A variety of explanations are given for these
this paper, the last is of greatest import as source of
differences, including viewing mental illness as a weakness,
increased stressors, since stressful life events are a risk
perceiving mental illness as within the normal spectrum of
factor for mental illnesses, and being a racial or ethnic
behavior (i.e., differing social constructions of the
minority carries a unique set of stressors (American
definitions of normalcy and mental illness), holding non-
Psychological Association, 2013).
Western views on the causes or origins of mental illness,
Migration is the first stressor many minorities face.
and parents seeking perfection in their children (Bignall et
For minorities who immigrate to the United States or their
al., 2014; Kearney et al., 2005). Regardless the cause,
children, acculturation may cause stress as the individual
stigma is deleterious because it hinders those who need
tries to reconcile their original culture with American
mental healthcare from seeking services due to fear of
culture (Office of the Surgeon General, 2001). From a
shame, rejection, or discrimination.
functionalist perspective, this may be understood as a
How individuals understand the concept of mental
disruption of the individual’s equilibrium—their cultural
illness is critical to both deciding and continuing to seek
surroundings—leading to personal and societal
help. Obviously, one will not seek help for something that
disequilibrium. A group of particular interest is adolescents
is not recognized as problematic. To an extent, mental
and young adults (e.g., college students) who are second or
illness is culturally bound and socially constructed. Though
“1.5” (immigrated during infancy) generation immigrants:
depression and anxiety are universal conditions, they may
these individuals may uniquely struggle to reconcile their
be understood and defined differently across different
parents’ cultures with their own identities as Americans.
cultures, individuals, and care providers. Foremost
Minorities who are not recent immigrants still face
amongst these dissimilarities are people’s beliefs about the
cultural stress: by being racially or ethnically different, they
causes of mental illness. For example, individuals who
are at risk of prejudice, discrimination, or simply not fitting
perceive mental illness as a product of weakness or as a
in with the majority culture. Racism and discrimination are
personal problem not relevant to share with outsiders are
obvious stressors that could contribute to worse mental
not likely to seek care, individuals who perceive it as a
health. Failure to fit in to cultural norms or lacking life
spiritual problem may seek help from a clergy person, and
experiences common to the majority culture is perhaps less
individuals who understand it only as a set of physiological
overtly negative than experiencing outright bigotry, but can
symptoms may complain to a primary care physician (PCP)
still be harmful because it can hinder social interaction and
of somatic symptoms.
connectedness. Halpern notes that “group density at the
local level rather than regional or national level…is
 Journal of Undergraduate Ethnic Minority Psychology – ISSN 2332-9300
Volume 3, 2017 8

These varying presentations are relevant to sized Midwestern university. Though the primary goal was
clinicians and other help-providers. If an individual decides to better understand racial and ethnic correlates of help-
to see to a practitioner, the provider’s ability to seeking, it was necessary first to study the context of effects
communicate effectively is critical to correct diagnosis and on mental health status itself in order to understand
treatment adherence. Central to effective communication implications for help-seeking. That is, all things being
about mental illness is “the development of congruence equal, it would be expected that levels of help-seeking
between the patient’s and doctor’s explanatory models within a group would reflect the level of symptoms of
[conceptual constructions that explain clinical phenomena] depression and anxiety experienced in that group. It was
for the patient’s sickness” (Ashton et al., 2003; Bignall et hypothesized that non-White individuals will report higher
al., 2014). A provider who is able to listen to, understand,
rates of symptoms of depression and anxiety while seeking
and act in light of a patient’s explanatory model is not only
help for mental health problems less often than their White
more likely to make a correct diagnosis, but may also be
peers.
perceived as more trustworthy and relatable. Consequently,
the patient may be more likely to adhere to treatment. For
example, it has been shown that minority college students Research Design and Methods
are less likely to schedule return mental healthcare
appointments than their White peers. However, return visit Undergraduate students at the university were
rates improve if they share language and ethnicity with the surveyed online using a 40-item questionnaire hosted on
provider (Kearney et al., 2005). Likewise, it was found that Qualtrics. The survey was publicized by posting in
university counseling service utilization was positively university-restricted Facebook groups, with a goal of
correlated with the ethnic makeup of the care providers receiving at least 100 responses. The survey remained open
(Hayes et al., 2011). from 17 October 2015, until 4 November 2015.
In a profession dominated by non-minority care The first section of the survey asked required
providers, the burden of responsibility lies largely with the demographics questions. Non-undergraduates were
clinician. One way to improve patient/provider discarded. The rest of the survey was response-optional.
communication is by improving the provider’s cultural The second section consisted of the Patient Health
competency, which the Office of the Surgeon General Questionnaire: Modified (PHQ-M), a nine-item assessment
defines as “recognition of patients’ cultures and for major depressive disorder (Spitzer, 1999). The third
[development of] a set of skills, knowledge, and policies to section consisted of the Generalized Anxiety Disorder 7-
deliver effective treatments” (2001). This is important Item Scale (GAD-7), a seven-item assessment for general
because even if patient and provider speak the same
anxiety disorder (Spitzer, 2006). Both the PHQ-M and
language, they may use and interpret terms, idioms, and
GAD-7 are well-validated assessments that score
metaphors differently. In addition, ethnic groups have
respondents’ symptoms using a four-choice scale (“not at
preferred styles of communicating. For example, people
from individualistic cultures tend to be more direct, all,” “several days,” “more than half the days,” “nearly
assertive, and expressive, whereas people from collectivist every day”) which is scored from 0 to 3. Scores for the
cultures tend to be indirect, deferential to authority, and PHQ-M range from 0 to 27, 11 or greater indicating a
accommodating (Ashton et al., 2003). positive score; scores for the GAD-7 range from 0 to 21, 8
Cultural competency is important in all stages of or greater indicating a positive score. The fourth section
the clinical interaction. Healthcare professionals are not asked questions related to awareness of mental health
perfectly objective, dispassionate scientists (nor should they resources on campus, past help-seeking behaviors, and
be): they too are part of a culture, even if it is the majority future help-seeking behaviors. The final two questions
culture. Their diagnostic and treatment choices are guided were free response, allowing respondents to comment on
by their own backgrounds. A provider’s bias, even if (1) why they would or would not be willing to seek help
unconscious, may affect diagnosis. Likewise, his or her from a variety of sources and (2) anything they wished to
ignorance of symptoms that present differently across races discuss relevant to the survey. As discussed below, these
and ethnicities may affect diagnosis. A clinician with free response sections provided insight into attitudes,
greater cultural competency will be better able to opinions, and beliefs about mental healthcare. Surveys
understand the patient’s complaints and to help the patient completed through the seventh question of the GAD-7
in a relevant, meaningful way. (Q23) were included in results.
Data were analyzed using GraphPad Prism. p <
Research Question 0.05 established significance in all tests. Levels of
depression and anxiety were analyzed based on raw score
This research was designed to address the and positive/negative status based on established cutoffs for
following issues: the effects of race and ethnicity on (1) the PHQ-M and GAD-7. Group scores were compared
depression and anxiety and (2) mental health help-seeking using Mann-Whitney test or non-parametric ANOVA;
behaviors in undergraduates at a selective, private, medium- group statuses were compared using Fisher’s exact test.
 Journal of Undergraduate Ethnic Minority Psychology - ISSN 2332-9300 9
Volume 3, 2017

Tests were used in the same manner to compare awareness Thus, a larger variance score corresponds to worse
and help-seeking behaviors across groups. concordance between the racial/ethnic compositions of the
U.S. News and World Report was consulted to student body and campus mental healthcare providers. It
identify the highest ranked U.S. colleges and universities in should be noted that although students per clinician is a
2015. The top twenty private institutions were noted: more palatable measure because it is not a fraction, it was
Princeton University, Harvard University, Yale University, not possible to use this because multiple schools have zero
Columbia University, Stanford University, University of providers in some categories. The institutions were then
Chicago, Massachusetts Institute of Technology (MIT), ranked from lowest value to highest score. Low variance
Duke University, University of Pennsylvania (U Penn), was interpreted as the racial/ethnic composition of the
California Institute of Technology, Johns Hopkins clinical staff being a relatively good fit to the composition
University, Dartmouth College, Northwestern University, of the student body, whereas high variance was interpreted
Brown University, Cornell University, Vanderbilt as representing a relatively poor fit.
University, Washington University in St. Louis (Wash U),
Rice University, University of Notre Dame, and Emory Results
University. Although CalTech is tied for tenth place,
CalTech does not provide a publicly available list of its University Demographics
campus healthcare providers, so CalTech was excluded, and
Georgetown University, the ranked twenty-first, was According to the Office of the University
included instead. Although University of California— Registrar, the undergraduate student body is 49.25% male
Berkeley falls in the top twenty schools overall, it was and 50.75% female. 51.3% are White, 28.6% Asian, 7.0%
excluded because of the substantial difference in size of its African American, and 1.0% Native
student body compared to those of other campuses. This American/Alaskan/Hawaiian. 5.5% identify themselves as
size difference also held true for the other highly-ranked Hispanic/Latino. 19.0% have international student status
public universities, such as University of California—Los (2015). According to the university website, the on-campus
Angeles, University of Virginia, University of Michigan— Student Health Services (SHS) has 13 mental healthcare
Ann Arbor, and University of North Carolina—Chapel Hill, providers, including psychiatrists, psychologists, social
which is why data collection was ceased at 20 universities, workers, and licensed professional counselors. 4 are male
rather than continuing to the top 25 or more. (30.8%) and 9 female (69.2%). 10 are white (76.9%), 2
The Common Data Sets, providing enrollment African American (15.4%), and 1 Asian (7.7%).
demographics, were obtained from the universities’ 100 responses were recorded (the maximum
websites. Data from the most recent year available as of 25 amount allowed per survey using a Qualtrics account
January 2016 were used. Specifically, numbers of full-time obtained through university undergraduate account
undergraduate students and of White, African licensing). 22 identified as male, 74 as female, and 4 as
American/Black, and Asian students were collected. Then, other (Fig. 1a). 72 identified as Caucasian/White, 5 African
campus mental healthcare provider data were collected American/Black, 15 East Asian, 5 South/West Asian, 1
from campus health clinic websites. In particular, numbers Native American/Native Alaskan/Native Hawaiian or
of mental healthcare providers, defined as individuals with Pacific Islander/First Nations, and 2 multiracial (Fig. 1b).
graduate, professional, or medical degrees in counseling, For analysis, groups were combined: East and South/West
therapy, social work, psychology, or psychiatry, were Asian were combined into Asian (20), and all non-White
collected. Many university websites provide biographies groups were combined into non-White (28). Statistical
and/or pictures of their clinicians from which the comparisons were made across all groups, across all groups
races/ethnicities were inferred. In cases in which in which the Asian groups are combined, and between
racial/ethnic identity was ambiguous or for universities that White and non-White. Because the university has a large
provided no information about clinicians other than their Jewish population for whom being Jewish is an important
names, further internet searches were conducted to find part of ethnic and cultural identity, this was also requested
personal or work websites, curriculum vitae, or social under demographics: 19 identified as Jewish, 77 as non-
networking sites containing photographs and other relevant Jewish. This survey hoped to explore differences between
information about those individuals. international and domestic students; unfortunately, only 4
From these data the variance between the respondents identified as international. Likewise, only 4
racial/ethnic composition of the student body and the respondents identified as Hispanic/Latino/a, so no
clinical staff of each institution was calculate using the comparisons were made in either category.
following expression, where CPS represents clinicians per
student:
!"#$%&"!'()*+ 5 !"#$%&"!70/89 5 !"#$%&"!/;)/< 5
123'()*+ − 123,-+./00 + 12370/89 − 123,-+./00 + 123/;)/< − 123,-+./00
!"#$%&"!,-+./00 !"#$%&"!,-+./00 !"#$%&"!,-+./00
123,-+./00
 Journal of Undergraduate Ethnic Minority Psychology – ISSN 2332-9300
Volume 3, 2017 10

Race, Ethnicity, and Help-Seeking

No differences were reported across gender or race

and ethnicity with respect to awareness of mental health
treatment options available on campus. 89% were aware of
SHS’s free counseling appointments; 89% were also aware
of the University’s Peer Counseling Service.
As expected, more females than males have
thought that they needed help for emotional, psychological,
or psychiatric matters in the last 12 months (n=92,
p=0.0021). Likewise, more females have also ever received
a formal diagnosis of mental illness (n=93, p=0.0351).
Fig. 1. Survey Respondent Demographics. (a) However, no differences were reported between males and
Gender. (b) Race/ethnicity. (c) Jewish. females with respect to whether or not providers were
ethnically and/or culturally similar.
Race, Ethnicity, and Prevalence Significantly more White individuals reported
thinking they needed help in the last 12 months (n=96,
No statistically significances differences in scores p=0.0227), having ever seen a mental healthcare provider
or statuses were observed between genders or across any (n=99, p=0.0013), and having ever received a formal
races or ethnicities. The only statistically significant diagnosis (n=96, p=0.0026) compared to non-White
difference found was in PHQ-M scores and statuses individuals (Fig. 3a-c). White individuals also reported that
between Jewish and non-Jewish individuals (Fig. 2a-b). provider(s) they saw were more ethnically and/or culturally
Jewish individuals had lower PHQ-M scores (n=96, similar to themselves than did non-White individuals (n=57,
p=0.0214) and fewer positive statuses (p=0.0207) than non- p=0.0008) (Fig. 3d). No difference was observed in the
Jewish individuals. number of appointments attended.
No differences were reported across any groups
(a) with respect to having visited SHS for mental healthcare or
number of visits made. However, among those reporting
they had visited SHS, White individuals reported that the
SHS provider(s) they saw were more ethnically and/or
culturally similar to themselves than non-White individuals
(n=30, p=0.0044) (Fig. 4).
Respondents were asked if they have ever sought
help related to mental health from other sources, including
clergypersons, PCPs, peer counseling resources, academic
resources, student leaders/mentors, family and friends,
emergency help lines, or other. While no statistically
significant differences were observed across groups, the
(b most commonly reported source was family and friends (45
) out of 47). The next two most common groups were PCPs
(12) and clergypersons and academic resources (both 11).
When asked what sources respondents would be
willing to consult for help, amongst the sources listed above
as well as SHS mental healthcare providers and non-SHS
mental healthcare providers, family and friends were the
most commonly reported (80 out of 97), followed by SHS
providers and non-SHS providers (56 and 54, respectively).
These were followed by peer counseling services (35) and
PCPs (25). These choices were reflected in the free
response comments, where the most common theme
Fig. 2. PHQ-M Scores and Statuses Across Jewish and Non-Jewish
Individuals. (a) Dotted line represents cutoff line (11) for positive score. indicated that individuals preferred professionals with
Jewish individuals had significantly lower PHQ-M scores (𝑛 = 96, 𝑝 = expertise in the area and who were bound to confidentiality.
0.0214, Mann-Whitney test) and (b) fewer positive scores (𝑝 = 0.0207, Other common themes indicated preference for someone
Fisher’s exact test) than non-Jewish individuals. Presented as mean with
SEM. who is familiar or trustworthy and concern for
 Journal of Undergraduate Ethnic Minority Psychology - ISSN 2332-9300 11
Volume 3, 2017

(a)
(b)

(c)
(d)

Fig. 3. Differences in Help-Seeking from Any Source. Compared to non-White individuals, White individuals are more likely to have (a) thought
they needed help for emotional, psychological, or psychiatric matters in the past 12 months (𝑛 = 96, 𝑝 = 0.0227, Fisher’s exact test), (b) ever seen a
mental healthcare provider (𝑛 = 99, 𝑝 = 0.0013, Fisher’s exact test), and (c) ever received a formal diagnosis (𝑛 = 96, 𝑝 = 0.0026, Fisher’s exact
test). (d) Furthermore, White individuals report that provider(s) they saw were more ethnically and/or culturally similar to themselves than did non-
White individuals (𝑛 = 57, 𝑝 = 0.0008, Mann-Whitney test). Respondents were asked to rank whether the provider(s) were not similar (0), somewhat
similar/depends on provider seen (if multiple) (1), or similar (2). Presented as mean with SEM.

racial/ethnic composition of the student body and of the

mental healthcare clinical staff), as follows: (1) Dartmouth,
(2) U Penn, (3) Georgetown, (4) Cornell, (5) Duke, (6)
Brown, (7) Notre Dame, (8) Harvard, (9) Rice, (10) Wash
U, (11) Northwestern, (12) Emory, (13) Stanford, (14)
Princeton, (15) Johns Hopkins, (16) U Chicago, (17)
Vanderbilt, (18) MIT, (19) Columbia, and (20) Yale (Tab.
1).

Discussion

Fig. 4. Differences in Help-Seeking from SHS. White individuals report Race, Ethnicity, and Prevalence
that SHS provider(s) they saw were more ethnically and/or culturally similar
to themselves than did non-White individuals (𝑛 = 30, 𝑝 = 0.0044, Mann-
Whitney test). Respondents were asked to rank whether the provider(s) were Most individuals were aware of sources on
not similar (0), somewhat similar/depends on provider seen (if multiple) (1), campus from which they could seek help for mental health
or similar (2). Presented as mean with SEM. issues. This indicates that education campaigns during
freshman orientation, at SHS, through student groups, etc.
have been
burdening or bothering one’s peers with personal problems.
largely successful.
The lack of differences in measured levels of
Mental Healthcare at Top 20 Universities
depression and anxiety—particularly where differences
may be expected to exist, such as across gender—may be
The institutions were ranked, from lowest to
due to small sample size. This explanation could extend to
highest variance (i.e., most to least similarity between the
the lack of differences across race and ethnicity.
 Journal of Undergraduate Ethnic Minority Psychology – ISSN 2332-9300
Volume 3, 2017 12

Additionally, shared environmental influences, such as high have been flattened.

academic pressure, economic stressors related to attending a That no difference is seen in the number of
private university, etc., may also abrogate some expected individuals who have seen SHS or the number of visits they
between-group differences. Finally, it may be explained by made is also in contrast to other studies (Kearney et al.,
the arbitrary nature of the survey questions (see 2005). This also may be due to small sample size, since
Limitations). those who have seen a provider at SHS fall in a subcategory
It is interesting that Jewish individuals have lower of those who have ever seen a provider. That a difference
PHQ-M scores and fewer positive scores. This could be is, however, observed across race and ethnicity in whether
due to social factors, such as social connectedness, specific individuals consider their SHS provider ethnically and/or
cultural values, and high local minority population density, culturally similar is significant: as described above, cultural
amongst the Jewish population at the university. It could background and competency are critical components of
also be affected by unmeasured factors, such as genetic patient retention and treatment adherence (Davidson et al,
predisposition or socioeconomic status, common to a large 2004; Kearney et al., 2005). Based on data provided by
portion of the group. SHS and the Office of the Registrar, an area of
improvement could include hiring staff more representative
Race, Ethnicity, and Help-Seeking of the racial and ethnic composition of the student body.
The fact that outside of friends and family, people
The lack of difference in having seen ethnically most often seek PCPs and members of the clergy for help
and/or culturally similar providers (both in general and at with mental health concerns indicates need for these
SHS) between males and females is important because it professions to receive training in this area. Though
supports the hypothesis that the effect observed across insufficient data was collected to determine whether non-
racial and ethnic groups is not due to random chance, since White individuals disproportionately seek help from non-
different races and ethnicities should be approximately mental healthcare professionals, based on previous
equally represented across genders. That is, the lack of research, it is reasonable to hypothesize that someone who
difference observed between genders helps validate the experiences a somatic presentation of depression (for
survey methodology. example, more commonly reported by Asians than
It is in accord with existing literature that White Caucasians) might see their PCP, while someone who
individuals more often report thinking they need help, attributes mental illness to spiritual problems might consult
having ever seen a mental healthcare practitioner, and a religious leader (Halpern, 1993; Kearney et al., 2005).
having ever received a formal diagnosis than non-White Likewise, those who cannot access healthcare specialists
individuals, in addition to describing these providers as due to economic, language, or other barriers might have no
more ethnically and/or culturally similar to themselves than other choice outside a PCP. These providers function as
do non-White individuals (Office of the Surgeon General, bridges between patients and the clinical care they need:
2001). This makes sense in light of possible differences in they are critical in recognizing symptoms and providing the
cultural perceptions and expectations regarding mental initial referrals and advice patients (or parishioners, as the
illness and treatment as described above (Bignall et al., case may be) may need.
2014). Since White and non-White individuals were not
found here to have significantly different levels of Mental Healthcare at Top 20 Private Universities
symptoms of depression and anxiety, these data are
concerning because they indicate that non-White groups The variance measure could also be calculated
perceive less need for help and receive less help than their without factoring in the proportion of students of each
White counterparts. In reality, since many studies show race/ethnicity, as follows:
that non-White individuals have increased risk for mental
health disorders, the disparity may be even more serious
/ / /
!"#$%&'( − !"#*+(,-.. + !"#1.-23 − !"#*+(,-.. + !"#-4&-5 − !"#*+(,-..
(Halpern, 1993; SAMHSA, 2015). However, no difference
!"#*+(,-..
was seen in the number of visits made, which is in contrast
to evidence presented previously (Kearney et al., 2005).
This may be due to small sample size, since fewer non-
White individuals report ever having seen a mental
Although this calculation does not account for the
healthcare professional. That is, adherence is a limited
proportional composition of each race/ethnicity, it is still
metric because it depends on the completion of an initial
worthwhile to consider because does give greater weight to
visit, which non-White individuals are less likely to do.
institutions that have a very high number of clinicians per
Additionally, since the sample size did not allow for
student for any one race/ethnicity. For example, Wash U
analysis to account for differences between individuals who
does well with
thought they did or did not need help, possible effect may
 Journal of Undergraduate Ethnic Minority Psychology - ISSN 2332-9300 13
Volume 3, 2017

Tab. 1. Variance in Racial/Ethnic Composition of Student Bodies Compared to Campus Mental

Healthcare Providers at Top 20 Private Universities.

Note. Variance factor was calculated as

0123451067814 01234510F=<GH 01234510<08<5
(?@A67814 B?@A9:4;<== )DE (?@AF=<GH B?@A9:4;<== )DE (?@A<08<5 B?@A9:4;<== )D
012345109:4;<== 012345109:4;<== 012345109:4;<==
?@A9:4;<==
Low variance represents greater racial/ethnic concordance between a campus’ student body and its mental
healthcare providers.

respect to African American/Black clinicians per student Limitations

(185) but poorly with respect to Asian clinicians per student
(1249). Because the Asian population is a minority of the The small sample size (n=100), and thus low
student body, this gets scaled down in the first model. power, is the greatest limitation in obtaining significant
However, this could have a disproportionate negative effect results. Consequently, in some instances, cohorts had to be
on students. Likewise, MIT has no African combined (e.g., merging all non-White groups) to analyze
American/Black mental healthcare providers. Because MIT data. In other instances, entire groups were excluded from
only reports 251 African American/Black students (5.6%), analysis because did not include enough members (e.g., in
this stark absence of matched providers is scaled out. gender comparisons, only male and female respondents
However, if even a single one of those students wishes to were included).
seek mental healthcare from a campus provider, there are The demographics of the survey respondents do
no matched providers available. As discussed above, this is not exactly reflect those of the university’s student body.
not just a matter of preference: congruence between patient Some of this effect is due to non-response bias, which was
and provider ethnicity has been shown to be associated with not controlled. It is also likely a consequence of which
higher utilization and retention rates (Hayes et al., 2011; groups of people are likely to have and actively use a
Kearney et al., 2005). Facebook account, prefer other social media platforms, or
use social media infrequently or never. Furthermore, the
nature of survey distribution may augment non-response
bias since the Facebook groups to which the survey was
 Journal of Undergraduate Ethnic Minority Psychology – ISSN 2332-9300
Volume 3, 2017 14

posted are not necessarily representative of the university’s The computational method itself, as described in
population as a whole, and Facebook itself can influence the discussion section, may be interpreted in different ways.
what posts an individual sees. There is no test for significance, so rankings should be
An inherent limitation is the format of the PHQ-M understood in broad strokes of where improvement may
and GAD-7. Several respondents noted in the comment need to occur, rather than as a strict benchmarks of
section that they felt the four options given in the PHQ-M performance. Initially, Friedman’s test/Kendall’s W was
and GAD-7 did not sufficiently describe their experience: in used to measure concordance between the racial/ethnic
particular, they indicated desire for a greater range of composition of the student body and that of the mental
intermediate response options. Other tools, including the healthcare staff. Unfortunately, this was unsuccessful
Outcome Questionnaire 45, Counseling Center Assessment because the small samples produced highly ambiguous
for Psychological Symptoms, and Brief Symptom Inventory results. Thus, concordance was instead calculated by
were considered because of their greater utilization in calculating a measure of variance.
literature, more specific questions and responses, and
specificity to college students (Hwang & Goto, 2008; REFERENCES
Kearney et al., 2005; Soet & Sevig, 2006). However, the American College Health Association. (2010).
PHQ-M and GAD-7 were chosen for several reasons. First, Undergraduate students: Reference group data
their format and scoring is consistent, simplifying both report, Spring 2010. Retrieved from
response and data analysis. Second, and most important, http://www.acha-ncha.org/docs/ACHA-NCHA-
they are relatively brief: the other assessments contained II_ReferenceGroup_DataReport_Spring2010.pdf
45, 70, and 53 items, respectively, and it was feared that American College Health Association. (2015).
respondents would not finish a survey that long. Undergraduate students: Reference group data
A particular barrier to understanding help-seeking report, Spring 2015. Retrieved from
behaviors is cost. SHS offers nine free counseling sessions http://www.acha-ncha.org/docs/NCHA-
per year for undergraduates. Following appointments and II_WEB_SPRING_2015_UNDERGRADUATE_
all appointments with psychiatrists are billed through REFERENCE_GROUP_DATA_%20REPORT.pd
insurance. These costs may pose a serious challenge to f
students’ ability to seek help, even if they are willing to do American Psychiatric Association. (2013). Diagnostic and
so. Statistical Manual of Mental Disorders: DSM-5.
The Common Data Set is inherently limited in that Washington, D.C: American Psychiatric
Association.
it includes international students as a separate
Ashton, C.M., Haidet, P., Paterniti, D.A., Collins, T.C.,
race/ethnicity. This poses a particular challenge for
Gordon, H.S., O’Malley, K.,…,Street, R. L.
analyzing data from schools that have high international (2003). Racial and ethnic disparities in the use of
student populations. For example, the university’s health services: Bias, preferences, or poor
Common Data Set states that the university is communication? Journal of General Internal
approximately 18% Asian. On the other hand, the Medicine, 18(2), 146–152.
documentation from the registrar’s office states that the Best Colleges. (2015). U.S. News and World Report.
university is approximately 28% Asian. This difference is Bignall, W., Jacquez, F., & Vaughn, L. (2014). Attributions
accounted for by Asian international students. Since not of mental illness: An ethnically diverse community
all schools provide enrollment information outside that perspective. Community Ment Health J, 51(5),
given in the Common Data Set, for the sake of consistency, 540-545.
the Common Data Set was used for all variance Davidson, M. M., Yakushka, O. F., & Sanford-Martens, T.
calculations, despite its flaws. Another limitation in data C. (2004). Racial and ethnic minority clients'
collection is that the racial/ethnic identity of some mental utilization of a university counseling center: An
healthcare providers was difficult to identify even after archival study. Journal of Multicultural
exhaustive internet searches as described under Research Counseling & Development, 32259-271.
Design and Methods. In all cases, an educated guess was Gary, F. A. (2005). Stigma: Barrier to mental health care
made, but errors may have occurred. In theory, errors among ethnic minorities. Issues in Mental Health
should be evenly distributed across all races/ethnicities, but Nursing, 26(10), 979-999.
because some of the samples were small, errors may be doi:10.1080/01612840500280661
Halpern, D. (1993). Minorities and mental health. Social
magnified. Finally, minority groups having lower
Science & Medicine, 36(5), 597-607.
representations, such as Hispanic/Latino, Native American,
Hayes, J.A., Soo Jeong, Y., Castonguay, L.G., Locke, B.D.,
or multi-racial were excluded from analysis because the McAleavey, A.A., & Nordberg, S. (2011). Rates
sample sizes were too small. In reality, these groups may and predictors of counseling center use among
be among the most severely underrepresented at university college students of color. Journal of College
health centers. Counseling, 14(2), 105-116.
 Journal of Undergraduate Ethnic Minority Psychology - ISSN 2332-9300 15
Volume 3, 2017

Hwang, W., & Goto, S. (2008). The impact of perceived

racial discrimination on the mental health of Asian
American and Latino college students. Cultural
Diversity and Ethnic Minority Psychology, 14(4),
326-335.
Hwang, W., Myers, H., Abekim, J., & Ting, J. (2008). A
conceptual paradigm for understanding culture's
impact on mental health: The cultural influences
on mental health (CIMH) model. Clinical
Psychology Review, 28(2), 211-227.
Kearney, L.K., Draper, M., & Barón, A. (2005). Counseling
utilization by ethnic minority college students.
Cultural Diversity and Ethnic Minority
Psychology, 11(3), 272-285.
Office of the Surgeon General, Center for Mental Health, &
National Institute of Mental Health. (2001).
Mental Health, Culture, Race, and Ethnicity: A
Supplement to Mental Health: A Report of the
Surgeon General. Rockville, MD: U.S.
Department of Health and Human Services.
United States Census Bureau. (2010). QuickFacts beta: St.
Louis city, Missouri. Retrieved from
http://quickfacts.census.gov/qfd/states/29/29510.ht
ml
United States Department of Education, National Center for
Education Statistics. (2014). Enrollment in
elementary, secondary, and degree-granting
postsecondary institutions, by level and control of
institution, enrollment level, and attendance status
and sex of student: Selected years, fall 1990
through fall 2024 [Data file]. Retrieved from
http://nces.ed.gov/programs/digest/d14/tables/dt14
_105.20.asp?current=yes
Soet, J., & Sevig, T. (2006). Mental health issues facing a
diverse sample of college students: Results from
the college student mental health survey. Naspa
Journal (Online), 43(3), 410-431.
Spitzer, R.L. (1999). Patient health questionnaire:
Modified. Retrieved from
http://www.integration.samhsa.gov/images/res/8.3.
4%20Patient%20Health%20Questionnaire%20(PH
Q-9)%20Adolescents.pdf
Spitzer, R.L. (2006). Generalized anxiety disorder 7-item
(GAD-7) scale. Retrieved from
http://www.integration.samhsa.gov/clinical-
practice/GAD708.19.08Cartwright.pdf
Substance Abuse and Mental Health Services
Administration. (2015). Racial/ethnic differences
in mental health service use among adults.
Rockville, MD: U.S. United States Department of
Health and Human Services. Retrieved from
http://www.samhsa.gov/data/sites/default/files/MH
ServicesUseAmongAdults/MHServicesUseAmong
Adults.pdf