Cross-Cultural Validity of the Self-Stigma of Seeking Help (SSOSH)
Scale: Examination Across Six Nations
David L. Vogel, Patrick Ian Armstrong,
Pei-Chun Tsai, Nathaniel G. Wade,
and Joseph H. Hammer
Iowa State University
National and Kapodistrian University of Athens
Nottingham Trent University
National and Kapodistrian University of Athens
Chinese University of Hong Kong
Researchers have found that the stigma associated with seeking therapy—particularly self-stigma—can
inhibit the use of psychological services. Yet, most of the research on self-stigma has been conducted in
the United States. This is a considerable limitation, as the role of self-stigma in the help-seeking process
may vary across cultural groups. However, to examine cross-cultural variations, researchers must first
develop culturally valid scales. Therefore, this study examined scale validity and reliability of the widely
used Self-Stigma of Seeking Help scale (SSOSH; Vogel, Wade, & Haake, 2006) across samples from 6
different countries (England, Greece, Israel, Taiwan, Turkey, and the United States). Specifically, we
used a confirmatory factor analysis framework to conduct measurement invariance analysis and latent
mean comparisons of the SSOSH across the 6 sampled countries. Overall, the results suggested that the
SSOSH has a similar univariate structure across countries and is sufficiently invariant across countries
to be used to explore cultural differences in the way that self-stigma relates to help-seeking behavior.
Keywords: self-stigma, stigma, help seeking, cross-cultural, validity
Self-stigma has been indicated as a considerable deterrent to
receiving quality mental health care (Vogel, Wade, & Haake,
2006). In the literature, self-stigma is defined as the reduction in a
person’s self-esteem or sense of self-worth due to the perception
held by the individual that he or she is socially unacceptable
(Vogel, Wade, & Hackler, 2007). Self-stigma is thought to occur
when people experiencing a mental illness or considering seeking
psychological help self-label as someone who is socially unaccept-
able (i.e., someone needing psychological services is weak) and in
doing so internalize stereotypes, apply negative public attitudes to
themselves, and suffer diminished self-esteem and self-efficacy
(Corrigan & Shapiro, 2010). Research has shown that individuals
who experience self-stigma suffer from lowered self-esteem (Link,
Struening, Neese-Todd, Asmussen, & Phelan, 2001) and increased
depression (Manos, Rusch, Kanter, & Clifford, 2009). In turn,
researchers have noted that individuals who self-stigmatize have
David L. Vogel, Patrick Ian Armstrong, Pei-Chun Tsai, Nathaniel G.
Wade, and Joseph H. Hammer, Department of Psychology, Iowa State
University; Georgios Efstathiou, Department of Psychology, National
and Kapodistrian University of Athens, Athens, Greece; Elizabeth
Holtham, Department of Psychology, Nottingham Trent University,
Nottingham, England; Elli Kouvaraki, Department of Psychology, Na-
tional and Kapodistrian University of Athens; Hsin-Ya Liao, Depart-
ment of Educational Psychology, Chinese University of Hong Kong,
Hong Kong; Zipora Shechtman, Department of Counseling and Human
Development, Haifa University, Haifa, Israel; Nursel Topkaya, Department of
Educational Sciences, Dumlupınar University, Kütahya, Turkey.
The data from Turkey were presented at the 11th National Congress of
Counseling and Guidance, I˙zmir, Turkey.
Correspondence concerning this article should be addressed to David
L. Vogel, Department of Psychology, Iowa State University, W112
Lagomarcino Hall, Ames, IA 50011. E-mail: email@example.com
Journal of Counseling Psychology
2013, Vol. 60, No. 2, 000
© 2013 American Psychological Association
0022-0167/13/$12.00 DOI: 10.1037/a0032055
more negative attitudes toward and less intentions to seek psycho-
logical services for many different forms of therapy, including
individual counseling (Conner et al., 2010; Vogel et al., 2007)
career counseling (Ludwikowski, Vogel, & Armstrong, 2009), and
group counseling (Vogel, Shechtman, & Wade, 2010). Those who
endorse greater self-stigma are also less willing to return for
subsequent sessions even after an initial visit (Wade, Post, Cor-
nish, Vogel, & Tucker, 2011) and have lower treatment compli-
ance (Fung, Tsang, & Corrigan, 2008).
Despite these important findings regarding self-stigma, much of
this research has been conducted only in the United States. This is
a considerable limitation, as self-stigma has been implicated in
avoidance of psychological services for individuals from different
backgrounds and nationalities (e.g., Shechtman, Vogel, & Maman,
2010). Although stigma may be present in most cultures, it may
take different forms depending upon cultural norms (Coker, 2005).
As such, in order to examine potential differences across cultural
groups, one must have culturally valid measurement tools. In the
Handbook of Counseling Psychology, Miller and Sheu (2008)
suggested that researchers must examine their measures with di-
verse samples to determine which aspects of the measures have
universal utility and which are applicable to only certain groups.
Without such examinations, it is unknown how applicable the
results of any particular cross-cultural study would be, as the
results could be due to true differences in the constructs of interest
or due to measurement error based on changes in the psychometric
properties of the measures when used with groups on which they
were not normed. Therefore, this study examined the cross-cultural
invariance of the psychometrics of the widely used Self-Stigma of
Seeking Help (SSOSH) scale across samples from six different
countries: England, Greece, Israel, Taiwan, Turkey, and the United
The SSOSH is a 10-item scale designed “to assess concerns
about the loss in self-esteem a person would feel if they decided to
seek help from a psychologist or other mental health professional”
(Vogel et al., 2006, p. 326). The scale has been shown to have a
unidimensional factor structure and adequate reliability among
samples drawn from various U.S. populations. For example, inter-
nal consistency estimates have been reported for general samples
of college students (.79–.92; Bathje & Pryor, 2011; Shepherd &
Rickard, 2012; Vogel et al., 2006, 2007), military personnel (.89;
Skopp et al., 2012), and community samples (.81–.91; Hammer &
Vogel, 2010; Wester, Arndt, Sedivy, & Arndt, 2010), as well as a
Middle Eastern American sample (.79; Soheilian & Inman, 2009)
and samples of African American (.84), Asian American (.85),
Latino American (.89), heterosexual (.90), and gay (.85; Vogel,
Heimerdinger-Edwards, Hammer, & Hubbard, 2011) men. Test–
retest reliability estimates in college populations have been re-
ported to be .72 (Vogel et al., 2006). The SSOSH also uniquely
predicts attitudes toward (r ? –.65) and intent to seek (r ? –.37;
Vogel et al., 2007) psychological help. In the original development
sample, the SSOSH was also found to differentiate between those
who sought psychological services and those who did not across a
2-month period (Vogel et al., 2006). However, whereas the psy-
chometric properties of the SSOSH have been researched in the
context of U.S. samples, little is known about the psychometric
properties of the scale with samples drawn from countries other
than the United States.
To address this limitation, we used a confirmatory factor anal-
ysis (CFA) framework to conduct measurement invariance (MI)
analysis (also called factorial invariance analysis or multiple-group
invariance analysis; Miller & Sheu, 2008) of model fit and factor
loadings within samples from six different countries. MI analysis
has been proposed by counseling researchers as a way to examine
measurement equivalence across different groups (Dimitrov, 2010;
Miller & Sheu, 2008), yet this analysis unfortunately remains rare
in the counseling literature. However, studies examining voca-
tional constructs (Hu, Pellegrini, & Scandura, 2011), depression
(Wu, 2010), and measurement of stress reactivity (Schlotz, Yim,
Zoccola, Jansen, & Schulz, 2011) have used this methodology.
Without knowledge of the measurement equivalence of a scale,
counseling psychologists cannot be certain that studies aimed at
understanding help-seeking decisions across different countries are
Participants and Procedures
We used archival data independently collected as part of six
separate studies investigating stigma and help seeking in England,
Greece, Israel, Taiwan, Turkey, and the United Stated. The results
from the samples are currently unpublished (see Shechtman et al.,
2010, for the one exception for the Israeli data).1In all cases, a
university’s Institutional Review Board approved the data collec-
tion procedures before data collection began. The scale was trans-
lated into the native language of each country so that all partici-
pants could complete the scale in their own language. In each case,
at least two translators, bilingual in English and their native lan-
guage, translated and back-translated the scale (one of the trans-
lators was always one of the authors of the current article). The
translators discussed items that showed semantic differences, and
decisions regarding wording choices were made by consensus.
Last, a separate expert faculty member in psychology and/or
education checked the translated version of the SSOSH and re-
vised the wordings where needed to ensure readability.
The following are sample characteristics by group: England:
Participants (N ? 450) ranged in age from 18 to 64 (M ? 36.13,
SD ? 11.22). The sample was 63% female and 37% male. Greece:
Participants (N ? 1,376) ranged in age from 18 to 30 (M ? 21.25,
SD ? 1.94). The sample was 63% female and 33% male (4% did
not report their sex). Israel: Participants (N ? 299) ranged in age
from 18 to 42 (M ? 24.04, SD ? 3.86). The sample was 49%
female and 51% male. Taiwan: Participants (N ? 299) ranged in
age from 18 to 29 (M ? 20.15, SD ? 1.50). The sample was 66%
female and 34% male. Turkey: Participants (N ? 506) ranged in
age from 18 to 42 (M ? 21.40, SD ? 2.39). The sample was 66%
female and 33% male (1% did not report their sex). United States:
Participants (N ? 655) ranged in age from 18 to 35 (M ? 19.35,
SD ? 1.75). The sample was 56% female and 43% male (1% did
not report their sex).
1The data from Turkey has also been presented at the National Congress
of Counseling and Guidance, I˙zmir, Turkey (Topkaya, 2011).
VOGEL ET AL.
of Seeking Help scale (SSOSH; Vogel et al., 2006). The SSOSH
is a 10-item scale measuring how much participants feel their
self-esteem would be threatened by seeking counseling (see Ap-
pendix for scale items). Responses are on a 5-point scale ranging
from 1 (Strongly Disagree) to 5 (Strongly Agree). Five items are
reverse-scored so that higher scores indicate greater self-stigma.
The evidence of the SSOSH’s reliability and validity was previ-
ously discussed in the introduction section.
Self-stigma was assessed using the Self-Stigma
Descriptive Statistics and Meta-Analytic Alpha
Table 1 shows means, standard deviations, and alpha reliability
estimates for each sample. As can be seen in the table, all groups
had adequate internal reliability estimates for research purposes
with a 10-item scale (see Ponterotto & Ruckdeschel, 2007). Fur-
thermore, using the varying coefficient equation developed by
Bonett (2010), we calculated the 95% confidence interval for the
alpha reliability coefficient of each sample as well as the meta-
analytic reliability across samples (see Table 1). Using this calcu-
lation, the alpha reliability estimate was .83, with 95% confidence
intervals between .82 and .84.
To examine the measurement invariance (MI) of the SSOSH
across countries, we used the sequential constraint imposition
approach as described by Dimitrov (2010). Specifically, employing
the full-information maximum likelihood estimation in LISREL 8.8
(Jöreskog, Sörbom, Du Toit, & Du Toit, 2003), we examined the
three most frequently assessed forms of MI (configural invariance,
metric invariance, and scalar invariance) using multiple-group
confirmatory factor analysis (Miller & Sheu, 2008). This approach
produces a series of nested models that can be compared to
examine whether configural, metric, and scalar invariances are
present across samples. Researchers have recently suggested that
the best way to compare MI models is to examine changes in
specific model fit indices (Cheung & Lau, 2012; Meade, Johnson,
& Braddy, 2008). For example, Meade et al. (2008) suggested that
changes in fit indices, such as the comparative fit index (CFI), are
less sensitive to issues such as (a) sample size and (b) number of
indicators and suggested a cutoff of ?–.002 for the ?CFI for
programs such as LISREL that use the normal theory weighted
least squares chi-square. However, because these model fit criteria
are sensitive to increasing model complexity (Meade et al., 2008)
and, therefore, may not be as accurate when more than two groups
are added to the model, in the subsequent MI analyses we used the
?CFI ?–.002 criteria to compare the U.S. sample to that of each
of the other countries.
Configural invariance is present
when the overall fit of the model for each country is present (i.e.,
the item factor loadings all significantly load on the hypothesized
factor [a single factor in the case of SSOSH]) and the overall
pattern of factor loadings are similar (i.e., the multiple-group
model fits the data). Model fit indices reported separately for each
country are presented in Table 2. Each country’s model fit the data,
and the item loadings were all significant at p ? .001 for each
sample (see Table 3 for item loadings). Furthermore, the multiple-
group models comparing each of the countries to the U.S. sample
each showed an overall acceptable fit to the data (see Table 4 for
the full invariance testing results and Table 5 for a summary of the
?CFI model results). Thus, configural invariance was supported,
suggesting that a univariate construct provides an acceptable fit
within each country.
Metric invariance is present when the
specific item factor loadings are similar across groups. To examine
the degree of metric invariance present, we compared a fully
invariant model, where each model factor path was set to be equal
across groups, to the previous configural model, where all the
paths were allowed to freely estimate across the different groups
(see Tables 4 and 5). The results showed that England and Taiwan
both were fully invariant with the U.S. sample (i.e., all of the items
loaded similarly). In contrast, Greece, Israel, and Turkey showed
?CFIs ? –.002, and thus full metric invariance was not supported.
To examine if partial metric variance was supported, we relaxed
the constraints in the models where the MIs and expected param-
eter changes(EPCs) were
Roznowski, & Necowitz, 1992) in a sequential fashion (Dimitrov,
2010). This led to freeing of one item for Israel (Item 7), as well
as two items for Greece (Items 7 and 8) and Turkey (Items 5 and
7), before the comparison of these partially invariant models (see
Tables 4 and 5) with the original configural model showed no
difference in the model fit (?CFI ? –.002). Therefore, metric
invariance was supported, with the majority of items (80%–100%
of the items per group) showing invariant factor loadings (see
Steenkamp & Baumgartner, 1998). As such, the scales seem to be
largely measuring the same construct, allowing for examination of
relationships between the SSOSH and other meaningful factors
One problem with the above multiple-group invariance testing is
that it can be sample-specific, and it does not provide confidence
intervals to understand the true possible range of factor loadings
across groups. Therefore, we conducted a bootstrap procedure to
create more stable means and confidence intervals around the
mean. The first step in the bootstrap procedure was to create 1,000
bootstrap samples from the original data sets for each country (i.e.,
1,000 samples for England, 1,000 samples for Greece, etc.) by
random sampling with replacement. The second step was to run the
factor model 1,000 times with these bootstrap samples to yield
Means, Standard Deviations, and Internal Consistencies for
Variables by Demographic Group
? [95% CI]
.89 [.87, .90]
.77 [.75, .79]
.80 [.76, .83]
.84 [.81, .87]
.82 [.80, .84]
.88 [.87, .89]
CI ? confidence interval.
CROSS-CULTURAL VALIDITY OF THE SSOSH SCALE
1,000 estimations of each factor path coefficient. The final step
was to use LISREL’s saved output of the 1,000 estimations of each
factor path coefficient to calculate the mean and 95% CI of the
factor coefficient. The means and 95% CIs are reported in Table 3.
The above analyses suggest that the
SSOSH is largely conceptualized similarly within different coun-
tries. However, it has also been suggested that to examine mean
differences between groups, scalar invariance should also be pres-
ent (Miller & Sheu, 2008). Scalar invariance is present when a
sufficient number of item intercepts are similar across groups. (i.e.,
Steenkamp & Baumgartner, 1998, have suggested that at least two
invariant items per factor are needed for meaningful comparisons
to be made). To examine if scalar invariance was present, we
compared a fully invariant model, where each item intercept (full
invariance) was set to be equal across models, to the previous
nested metric models (see Tables 4 and 5). The results showed that
none of the countries were fully invariant compared to the U.S.
sample (i.e., ?CFIs ? –.002), and thus full metric invariance was
not supported. To examine if partial metric variance was sup-
ported, we relaxed the constraints in the models where the MIs and
EPCs were substantial (see MacCallum et al., 1992) in a sequential
fashion (Dimitrov, 2010). This led to freeing of three paths for
Israel (Items 3, 4, and 10) and Taiwan (Items 5, 9, and 10) and five
paths for Turkey (Items 5, 6, 7, 8, and 10), Greece (Items 1, 2, 7,
8, and 10), and England (Items 2, 3, 4, 7, and 10), before the
comparison of these partially invariant models (see Tables 4 and 5)
with the original configural model showed sufficiently small dif-
ferences in the model fit (?CFI ? –.002). Therefore, partial scalar
invariance was supported, with at least half of the items (50%–
70% of the items per group) showing invariant factor loadings. As
such, based on Steenkamp and Baumgartner’s (1998) mathemati-
cal proof, sufficient invariance appears to be present to allow
between-country differences to be examined.
Latent mean comparison.
The latent means for England (M ?
3.03, SE ? 0.03), Greece (M ? 2.76, SE ? 0.02), Israel (M ? 2.89,
0.03), and the United States (M ? 3.03, SE ? 0.03) were then
examined by comparing an invariant model, where each of the
latent means were set to be equal across models, to the previous
partial scalar models, where all the latent means were allowed to
freely estimate (see Tables 4 and 5). The results showed that the
English and U.S. samples were invariant (i.e., equal means), while
the latent mean scores for Greece, Israel, Turkey, and Taiwan were
each variant with the latent mean from the U.S. sample (i.e.,
?CFIs ? –.002). In each case, the samples from Greece, Israel,
Turkey, and Taiwan showed lower latent means (i.e., less self-
stigma) than did the U.S. sample.
Self-stigma is a key factor in the decision to seek help and,
therefore, of direct importance to researchers and clinicians devel-
oping interventions to reach out to underserved populations (see
Vogel et al., 2010). However, limited research on the reliability
and validity of scales to measure self-stigma outside of U.S.
populations has curtailed generalizability (Miller & Sheu, 2008).
The present research addressed this need by testing the measure-
ment invariance of the widely used Self-Stigma of Seeking Help
(SSOSH) scale across samples from six different countries. Spe-
cifically, in the configural invariance analysis, we found that the
single-factor construct held across all countries. Similarly, the
internal consistencies across country samples (.77–.89) were con-
sistent with previous reports based on samples of college students
(.79–.92; Vogel et al., 2006) and nonmajority samples (.79–.89;
Soheilian & Inman, 2009; Vogel et al., 2011). Furthermore, the
metric (factor loading) invariance analysis supported the invari-
ance of the majority of items across countries (80%–100% of the
items invariant across countries). Given these findings, it seems
that, overall, the SSOSH assesses a construct that can be mean-
ingfully measured across many cultural groups.
Fit Indices of the Confirmatory Factor Analysis Results of the
SSOSH Across Cultures
Country Scaled ?2
CFI RMSEA [95% CI] SRMR
.069 [.054, .084]
.074 [.066, .082]
.071 [.053, .090]
.066 [.048, .085]
.061 [.047, .075]
.044 [.031, .057]
index; RMSEA ? root-mean-square error of approximation; CI ? confi-
dence interval; SRMR ? standardized root-mean residual.
???p ? .001.
SSOSH ? Self-Stigma of Seeking Help; CFI ? comparative fit
Mean Factor Loadings From Bootstrap of the 10 Items of the SSOSH Across Cultures
M [95% CI]
M [95% CI]
M [95% CI]
M [95% CI]
M [95% CI]Item
M [95% CI]
.80 [.73, .86]
.67 [.59, .74]
.71 [.62, .80]
.52 [.43, .62]
.45 [.34, .55]
.84 [.77, .90]
.78 [.69, .87]
.81 [.74, .88]
.46 [.35, .55]
.68 [.60, .75]
.66 [.60, .71]
.47 [.41, .53]
.61 [.54, .67]
.31 [.24, .37]
.24 [.17, .31]
.76 [.71, .82]
.51 [.44, .57]
.73 [.68, .78]
.28 [.21, .35]
.58 [.41, .52]
.58 [.46, .70]
.57 [.46, .68]
.68 [.55, .82]
.33 [.19, .46]
.35 [.23, .48]
.69 [.58, .79]
.74 [.64, .84]
.74 [.63, .84]
.47 [.33, .60]
.33 [.21, .46]
.69 [.57, .81]
.68 [.58, .79]
.69 [.57, .80]
.34 [.20, .46]
.29 [.16, .41]
.79 [.69, .91]
.63 [.53, .72]
.72 [.60, .83]
.64 [.53, .75]
.45 [.34, .56]
.64 [.54, .76]
.58 [.48, .68]
.62 [.52, .72]
.35 [.24, .45]
.56 [.45, .65]
.71 [.61, .81]
.68 [.59, .76]
.72 [.63, .82]
.49 [.39, .59]
.30 [.19, .40]
.80 [.76, .83]
.70 [.64, .76]
.76 [.71, .81]
.40 [.33, .48]
.42 [.35, .50]
.81 [.76, .85]
.79 [.74, .83]
.80 [.75, .84]
.50 [.42, .57]
.62 [.56, .67]10
All loadings are significant at p ? .001. SSOSH ? Self-Stigma of Seeking Help; CI ? confidence interval.
VOGEL ET AL.
Although the majority of the items were invariant, three items
did show some variance in three countries (see the Appendix for a
list of the items). Item 7 (“I would feel okay about myself if I made
the choice to seek professional help”) was the most variant, show-
ing a difference with the normed sample (U.S.) and Greek, Israeli,
and Turkish. The wording of this item may not have translated as
clearly across countries. For example, it is possible the phrase “feel
okay” did not have a clear translation in these languages. Item 5
(“My view of myself would not change just because I made the
choice to see a therapist”) also showed a difference for Turkey but
was invariant for other groups. Finally, Item 8 (“If I went to a
therapist, I would be less satisfied with myself”) was variant for
the Greek sample but invariant for the other groups. Some addi-
tional examination of these items may be warranted, and research-
ers using the SSOSH in these countries may need to assess if these
items function the same for their samples.
We also examined the invariance of the latent means and found
that, whereas the English and U.S. samples did not differ with each
other, the other samples (Greek, Israeli, Taiwanese, and Turkish)
each showed lower levels of self-stigma than did the U.S. sample.
Metric Invariance Comparisons of SSOSH Across Countries
Latent mean invariance
250.52 222.54 70.063.986
275.65 253.3279 .063.984
Latent mean invariance
493.39 379.5270 .066.973
Latent mean invariance
Latent mean invariance
218.31131.7070 .043 .993
246.93 149.95 79 .043.992
Latent mean invariance
256.23 178.0170.052 .988
comparative fit index.
aInvariance is ?CFI ? –.002.
?p ? .05.
SSOSH ? Self-Stigma of Seeking Help; S-B ?2? Satorra–Bentler chi-square; RMSEA ? root-mean-square error of approximation; CFI ?
??p ? .01.
???p ? .001.
CROSS-CULTURAL VALIDITY OF THE SSOSH SCALE
This finding is consistent with assertions that perceptions of stigma
could vary across cultural groups that place more emphasis on
independence versus interconnectedness (Angermeyer & Dietrich,
2006). For example, a cultural focus on others could lessen the
importance of the self and, therefore, might elicit less self-stigma.
In addition, the SSOSH scale may reflect specific Western cultural
expectations, including being able to solve problems on one’s own,
being independent, and being in control of one’s emotions. In other
words, the act of seeking mental health services may be viewed as
a sign of individual weakness (Vogel et al., 2006). The current
version of the SSOSH, having been developed in a Western
country, might, therefore, overrepresent individualistic notions
(e.g., Item 10 “I would feel worse about myself if I could not solve
my own problems”) and underrepresent collectivistic notions of
self-stigma. Given previous findings that self-stigma may be re-
lated to perceptions of shame and views of mental health treatment
in Eastern and Middle Eastern cultures (e.g., Shechtman et al.,
2010; Soheilian & Inman, 2009), the understanding of the concept
of self-stigma might be further enhanced by adding some addi-
tional items that reflect a more collectivistic orientation. For ex-
ample, in a study of Middle Eastern Americans, Soheilian and
Inman (2009) suggested that self-stigma could occur among Arab
individuals when they internalize the prejudices related to seeking
help present in the larger society and their family unit. Thus,
individuals may feel increased self-stigma due to the desire to
protect not only their own reputation but also that of their family.
In the future, researchers may want to investigate the effect of
adding items (e.g., “I would feel as though I let my family down
by not solving my problems without professional help”) that
directly reflect concerns about failing other important people such
as family members.
Limitations and Conclusions
Overall, this study confirmed that the SSOSH scale largely
maintained accuracy of measurement across the studied cultural
groups. Therefore, researchers who wish to use the SSOSH to
examine self-stigma across cultural groups can, with greater con-
fidence, interpret detected mean differences on the SSOSH as
reflecting true latent differences in self-stigma, rather than mea-
surement error due to measurement invariance. Furthermore, re-
searchers can also more confidently evaluate the strengths of
relationships between self-stigma and other theoretically and cul-
turally relevant factors, across cultural groups. As such, future
research should examine the cultural applicability of help-seeking
decision-making models that include self-stigma.
Despite the important findings of the current investigation, some
limitations should be noted. First, even though our study accessed
samples from six countries, future researchers could examine
whether the results generalize to other countries in the world.
Furthermore, within-country differences (within-country sub-
groups) were not examined (e.g., between individuals from differ-
ent religious affiliations or ethnic groups). Researchers may want
to attend to these subgroups. Researchers may also want to exam-
ine these relationships among immigrants within a given country
to see if differences exist between groups originating from differ-
ent countries but sharing the same nationality. In addition, future
research should consider the impact of generational status, which
has previously been found to relate to attitudes and stigma asso-
ciated with seeking help (Ta, Holck, & Gee, 2010). Another
limitation is that these samples were mostly drawn from college
student populations (i.e., only the sample from England included
noncollege students) within each country. As such, future research
could examine whether measurement invariance of the SSOSH
scale holds for individuals from community populations across
different cultures. In many countries, college students do not
necessarily represent the majority and are often more privileged.
Such privilege may influence the degree to which individuals
internalize stigma compared to others living in their country.
Angermeyer, M. C., & Dietrich, S. (2006). Public beliefs about and
attitudes towards people with mental illness: A review of population
studies. Acta Psychiatrica Scandinavica, 113, 163–179. doi:10.1111/j
Bathje, G. J., & Pryor, J. B. (2011). The relationships of public and
self-stigma to seeking mental health services. Journal of Mental Health
Counseling, 33, 161–176.
Bonett, D. G. (2010). Varying coefficient meta-analytic methods for alpha
reliability. Psychological Methods, 15, 368–385. doi:10.1037/a0020142
Cheung, G. W., & Lau, R. S. (2012). A direct comparison approach for
testing measurement invariance. Organizational Research Methods, 15,
Coker, E. M. (2005). Selfhood and social distance: Toward a cultural
understanding of psychiatric stigma in Egypt. Social Science & Medi-
cine, 61, 920–930. doi:10.1016/j.socscimed.2005.01.009
Conner, K. O., Copeland, V. C., Grote, N. K., Koeske, G., Rosen, D.,
Reynolds, C. F., III, & Brown, C. (2010). Mental health treatment
Invariance CFI Comparisons for SSOSH Models Across Groups
Country Configural CFI
2, 3, 4, 7, 10
1, 2, 7, 8, 10
3, 4, 7
5, 6, 7, 8, 10
5, 9, 10
Dashes indicate that there were no invariant items. CFI ? comparative fit index; SSOSH ? Self-Stigma of Seeking Help. Invariance is ?CFI ?
bVariant mean differences.
VOGEL ET AL.
seeking among older adults with depression: The impact of stigma and
race. American Journal of Geriatric Psychiatry, 18, 531–543.
Corrigan, P. W., & Shapiro, J. (2010). Measuring the impact of programs
that challenge the public stigma of mental illness. Clinical Psychology
Review, 30, 907–922.
Dimitrov, D. M. (2010). Testing for factorial invariance in the context of
construct validation. Measurement and Evaluation in Counseling and
Development, 43, 121–149. doi:10.1177/0748175610373459
Fung, K. M., Tsang, H. W., & Corrigan, P. W. (2008). Self-stigma of
people with schizophrenia as predictor of their adherence to psychoso-
cial treatment. Psychiatric Rehabilitation Journal, 32, 95–104. doi:
Hammer, J. H., & Vogel, D. L. (2010). Men’s help seeking for depression:
The efficacy of a male-sensitive brochure about counseling. Counseling
Psychologist, 38, 296–313. doi:10.1177/0011000009351937
Hu, C., Pellegrini, E. K., & Scandura, T. A. (2011). Measurement invari-
ance in mentoring research: A cross-cultural examination across Taiwan
and the U.S. Journal of Vocational Behavior, 78, 274–282. doi:10.1016/
Jöreskog, K. G., Sörbom, D., Du Toit, S., & Du Toit, M. (2003). LISREL
8: New statistical features. Lincolnwood, IL: Scientific Software Inter-
Link, B. G., Struening, E. L., Neese-Todd, S., Asmussen, S., & Phelan,
J. C. (2001). Stigma as a barrier to recovery: The consequences of stigma
for the self-esteem of people with mental illnesses. Psychiatric Services,
52, 1621–1626. doi:10.1176/appi.ps.52.12.1621
Ludwikowski, W., Vogel, D. L., & Armstrong, P. I. (2009). Attitudes
towards career counseling: The role of public and self-stigma. Journal of
Counseling Psychology, 56, 408–416. doi:10.1037/a0016180
MacCallum, R. C., Roznowski, M., & Necowitz, L. B. (1992). Model
modifications in covariance structure analysis: The problem of capital-
ization on chance. Psychological Bulletin, 111, 490–504. doi:10.1037/
Manos, R. C., Rusch, L. C., Kanter, J. W., & Clifford, L. M. (2009).
Depression self-stigma as a mediator of the relationship between depres-
sion severity and avoidance. Journal of Social and Clinical Psychology,
28, 1128–1143. doi:10.1521/jscp.2009.28.9.1128
Meade, A. W., Johnson, E. C., & Braddy, P. W. (2008). Power and
sensitivity of alternative fit indices in tests of measurement invariance.
Journal of Applied Psychology, 93, 568–592. doi:10.1037/0021-9010
Miller, M. J., & Sheu, H. (2008). Conceptual and measurement issues in
multicultural psychology research. In S. D. Brown & R. W. Lent (Eds.),
Handbook of counseling psychology (4th ed., pp. 103–120). New York,
Ponterotto, J. G., & Ruckdeschel, D. E. (2007). An overview of coefficient
alpha and a reliability matrix for estimating adequacy of internal con-
sistency coefficients with psychological research measures. Perceptual
and Motor Skills, 105, 997–1014.
Schlotz, W., Yim, I. S., Zoccola, P. M., Jansen, L., & Schulz, P. (2011).
The Perceived Stress Reactivity Scale: Measurement invariance, stabil-
ity, and validity in three countries. Psychological Assessment, 23, 80–
Shechtman, Z., Vogel, D. L., & Maman, N. (2010). Seeking psychological
help: A comparison of individual and group treatment. Psychotherapy
Research, 20, 30–36. doi:10.1080/10503300903307648
Shepherd, C. B., & Rickard, K. M. (2012). Drive for muscularity and
help-seeking: The mediational role of gender role conflict, self-stigma,
and attitudes. Psychology of Men & Masculinity, 13, 379–392. doi:
Skopp, N. A., Bush, N. E., Vogel, D. L., Wade, N. G., Sirotin, A. P.,
McCann, R. A., & Metzger-Abamukong, M. J. (2012). Development and
initial testing of a measure of public and self-stigma in the military.
Journal of Clinical Psychology, 68, 1036–1047. doi:10.1002/jclp.21889
Soheilian, S. S., & Inman, A. G. (2009). Middle Eastern Americans: The
effects of stigma on attitudes toward counseling. Journal of Muslim
Mental Health, 4, 139–158. doi:10.1080/15564900903245766
Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement
invariance in cross-national consumer research. Journal of Consumer
Research, 25, 78–90. doi:10.1086/209528
Ta, V. M., Holck, P., & Gee, G. C. (2010). Generational status and family
cohesion effects on the receipt of mental health services among Asian
Americans: Findings from the National Latino and Asian American
Study. American Journal of Public Health, 100, 115–121.
Topkaya, N. (2011). The validity and reliability studies of the Turkish
version of the Self-Stigma of Seeking Help Scale. Paper presented at the
11th National Congress of Counseling and Guidance, I˙zmir, Turkey.
Vogel, D. L., Heimerdinger-Edwards, S. R., Hammer, J. H., & Hubbard, A.
(2011). “Boys don’t cry”: Examination of the links between masculine
norms and help-seeking attitudes for men from diverse cultural back-
grounds. Journal of Counseling Psychology, 58, 368–382. doi:10.1037/
Vogel, D. L., Shechtman, Z., & Wade, N. G. (2010). The role of public and
self-stigma in predicting attitudes toward group counseling. Counseling
Psychologist, 38, 904–922. doi:10.1177/0011000010368297
Vogel, D. L., Wade, N. G., & Haake, S. (2006). Measuring the self-stigma
associated with seeking psychological help. Journal of Counseling Psy-
chology, 53, 325–337. doi:10.1037/0022-018.104.22.1685
Vogel, D. L., Wade, N. G., & Hackler, A. H. (2007). Perceived public
stigma and the willingness to seek counseling: The mediating roles of
self-stigma and attitudes towards counseling. Journal of Counseling
Psychology, 54, 40–50. doi:10.1037/0022-022.214.171.124
Wade, N. G., Post, B., Cornish, M., Vogel, D. L., & Tucker, J. (2011).
Predictors of the change in self-stigma following a single session of
group counseling. Journal of Counseling Psychology, 58, 170–182.
Wester, S. R., Arndt, D., Sedivy, S. K., & Arndt, L. (2010). Male police
officers and stigma associated with counseling. Psychology of Men &
Masculinity, 11, 286–302. doi:10.1037/a0019108
Wu, P. C. (2010). Measurement invariance and latent mean differences of
the Beck Depression Inventory II across gender groups. Journal of
Psychoeducational Assessment, 28, 551–563.
CROSS-CULTURAL VALIDITY OF THE SSOSH SCALE
Appendix Download full-text
Self-Stigma of Seeking Help Scale Items
they consider seeking help for. This can bring up reactions about what
seeking help would mean. Please use the 5-point scale to rate the degree
to which each item describes how you might react in this situation.
Received November 9, 2012
Revision received January 22, 2013
Accepted January 22, 2013 ?
1 ? Strongly Disagree2 ? Disagree3 ? Agree & Disagree Equally4 ? Agree5 ? Strongly Agree
1. I would feel inadequate if I went to a therapist for psychological help.
2. My self-confidence would NOT be threatened if I sought professional help.
3. Seeking psychological help would make me feel less intelligent.
4. My self-esteem would increase if I talked to a therapist.
5. My view of myself would not change just because I made the choice to see a therapist.
6. It would make me feel inferior to ask a therapist for help.
7. I would feel okay about myself if I made the choice to seek professional help.
8. If I went to a therapist, I would be less satisfied with myself.
9. My self-confidence would remain the same if I sought professional help for a problem I could not solve.
10. I would feel worse about myself if I could not solve my own problems.
Items 2, 4, 5, 7, and 9 are reverse scored.
VOGEL ET AL.