Validity and Reliability of Internalized Stigma of Mental Illness (Cantonese)

Abstract

Purpose:

This study aims to translate and test the reliability and validity of the Internalized Stigma of Mental Illness-Cantonese (ISMI-C).

Methods:

The original English version of ISMI is translated into the ISMI-C by going through forward and backward translation procedure. A cross-sectional research design is adopted that involved 295 participants randomly drawn from a population of Chinese consumers participated in different kinds of community-based mental health services.

Results:

Results show that the Cronbach’s α coefficient of the ISMI-C is .93. With regard to validity test, the ISMI-C shows significant and negative correlation with measures on self-esteem and quality of life. Also, an explorative factor analysis yields five factors that are consistent with previous research results.

Discussion:

This study shows that the ISMI-C is a reliable and valid measure. ISMI-C can facilitate the development of interventions in reducing self-stigma for people with mental illness across Chinese societies.

Keywords

internalized stigma of mental illness validation mental illness

Introduction

Stigma is the mark of disgrace or discredit that separate people with mental illness from others (Byrne, 1999). Stigma consists of public and self-stigmas (Corrigan, Roe, & Tsang, 2011). In terms of mental illness, the stigmatization of those with mental health conditions by the general public is considered to be public stigma. In Hong Kong, people with mental illness commonly experience public stigma. For example, in a recent local survey of 1,007 community residents, nearly half of the respondents reported that they view people with mental illness as quick-tempered, and about 30% of the respondents regarded them as dangerous and will constitute as such for the rest of their lives (Tsang, Tam, Chan, & Cheung, 2003). In addition, in another local survey of 193 people who have mental illness, 75% of the respondents reported that they have experienced feelings of being stigmatized and discriminated (Chung & Wong, 2004).

One consequence of public stigma is that people with mental illness often internalize these stigmatizing beliefs and this becomes self-stigmatization. The problem of self-stigma is common among those with mental illness in Western countries; more than 40% in Europe were found to have internalized stigma (Brohan, Elgie, Sartorius, & Thornicroft, 2010). There is research evidence that indicates self-stigma has negative impacts on the lives of those who suffer from mental illness, such as reduction in self-esteem, self-efficacy, quality of life (QoL), and treatment adherence, while an increase in psychiatric symptoms (Corrigan, Watson, & Barr, 2006; Gerlinger, Hauser, Hert, Lacluyse, & Wampers, 2013; Livingston & Boyd, 2010; Lysaker, Roe, & Yanos, 2007; Ritsher, Otilingam, & Grajiales, 2003; Tang & Wu, 2012).

To enable the development of interventions that are effective in reducing the self-stigma of mental health sufferers in the Hong Kong context, a psychometrically sound measure of self-stigma relevant to Chinese people with severe mental illness is necessary.

The Internalized Stigma of Mental Illness (ISMI) scale has been developed by Ritsher, Otilingam, and Grajiales (2003) in the United States to measure the subjective experiences of self-stigma. The ISMI consists of 29 items. Each item is rated on a 4-point Likert-type scale with 1 representing strongly disagree and 4 representing strongly agree. A higher total score indicates a higher level of internalized stigma. The ISMI consists of five subscales that measure alienation, stereotype endorsement, perceived discrimination, social withdrawal, and stigma resistance (SR). Recent research studies have suggested that the SR subscale is conceptually different from the other subscales (Sibitz, Unger, Woppmann, Zidek, & Amering, 2011) and has been perceived as a separate construct in the ISMI (Brohan et al., 2010; Lysaker et al., 2007; Sibitz et al., 2011; West, Yanos, Smith, Roe, & Lysaker, 2011). So, in the present study, the psychometric properties of the 24-item ISMI and 5-item SR are separately examined. In view of previous research studies (Brohan et al., 2010; Ritsher et al., 2003; West et al., 2011), the cutoff point of the 2.5 ISMI overall mean score is adopted in the present study for the 24-item version of the ISMI.

The ISMI has been translated into different languages and widely used in the United States (Ritsher et al., 2003; West et al., 2011) as well as 14 European countries (Brohan et al., 2010). However, a validated and reliable Cantonese version of the ISMI is lacking. This Cantonese version is especially needed in the Hong Kong context. So it is important to translate the ISMI into Cantonese (ISMI-C) and validate this scale so as to facilitate further research and service development in this area. Thus, the aims of the present study are to translate the ISMI into Cantonese Chinese and test its reliability and validity on a Chinese sample in the Hong Kong context.

Method

Procedure

A cross-sectional research design is adopted in this study. A randomized sample of 295 individuals was recruited from a population of Chinese consumers of community-based mental health services provided by five nongovernment organizations (NGOs) in two Cantonese-speaking regions, Hong Kong and Macau, during the period of 2013–2014. The service units included five psychiatric halfway houses, two sheltered workshops with supported employment services, and five community mental health centers which serve over 6,000 clients. The inclusion criteria for this study include (a) meeting the criteria cited in the Diagnostic and Statistical Manual of Mental Disorder (Fourth Edition, Text Revision; American Psychiatric Association, 2005) for any kind of diagnosis of mental disorder, (b) aged 18 or above to participate, (c) able to speak and understand Cantonese, and (d) receiving one or more community-based mental health services from local NGOs. However, those who have a primary or comorbid diagnosis of intellectual or learning disability were excluded. The current study was approved by the boards of the executive committee of the NGOs involved. All of the participants gave their written informed consent.

A questionnaire survey was conducted with the participants from September 2013 to September 2014. Demographic and clinical characteristics were collected from the participants and/or from the service units involved. The majority of the questionnaires were self-administered, but questions were read out to a few of the illiterate participants by the researcher.

Measuring Scales

Development of ISMI-(Cantonese)

The original English version of the 29-item ISMI was translated into the Cantonese language by using the forward–backward translation procedure. With regard to forward translation, three bilingual research team members (the authors) carried out independent translations of the ISMI from English into Cantonese. Then, the research team members reached a consensus on a tentative version of the ISMI-Cantonese (ISMI-C). After that, the tentative version of the ISMI-C was translated back into English by professional linguistics translators who did not have knowledge of the original English version. Then, the draft ISMI-C was reviewed by a consultation panel that comprised research team members and experienced mental health professionals. Both conceptual and linguistic equivalence to the local language and culture were taken into consideration. Later, the drafted version of the ISMI-C was piloted with 15 consumers of supported employment or psychiatric halfway house services. After the pilot test was reviewed, and the necessary amendments were made by the research team, the ISMI-C was finalized.

The Rosenberg Self-Esteem Scale (RSES) has been widely used in many countries and translated into many languages, including Chinese (Schmitt & Allik, 2005). It consists of 10 items. Each item is rated on a 4-point scale that ranges from strongly agree to strongly disagree. The reliability and validity of the Chinese version of the RSES have been tested and verified to be satisfactory (S. O. Leung & Wong, 2008).

The World Health Organization Quality of Life measure (WHOQOL-BREF; Hong Kong [HK]) has been reported to demonstrate satisfactory estimates of reliability and validity and allows for cross-cultural sensitivity (Skevington, Lotfy, & O’Connell, 2004). The WHOQOL-BREF is suitable for use with those who have mental illness (Orley, Saxena, & Herrman, 1998). The reliability and validity of the Chinese version of the WHOQOL-BREF (HK) have been tested and determined to be satisfactory (K. F. Leung, Wong, Tay, Chu, & Ng, 2005). In this study, 2 items (“how would you rate your QoL?” and “how satisfied are you with your health?”) are adopted from the Chinese version of the WHOQOL-BREF (HK) to measure the overall QoL of the respondents.

Statistical Analysis

The SPSS 22.0 software package was used for statistical analysis. Any values of p < .05 were considered to be statistically significant. Descriptive statistics, including demographical and clinical characteristics, were presented as means, standard deviations (SDs), or percentages. The scores on the measures were calculated and presented as the total and mean scores, SDs, skewness, and kurtosis. The normality of the data was also assessed.

The reliability of the ISMI-C was examined through internal consistency with Cronbach’s α and Guttman split-half coefficient as the measures, and an α equal to or greater than .70 was considered to be satisfactory (Polit & Beck, 2006). Item-level tests were also conducted to assess the scale mean, variance, corrected item-total correlations, and Cronbach’s α if an item was deleted.

The ISMI-C was validated through concurrent and construct validities. The former was assessed by Pearson’s correlation among the ISMI-C, RSES, and WHOQOL-BREF (HK). The latter was assessed by factor analysis for the 29 items in the ISMI-C to determine whether it consisted of five factors as suggested by the original English version of the ISMI (Ritsher et al., 2003). An exploratory factor analysis (EFA) was used as the statistical technique, and the principal component analysis with equamax rotation was used as the method. These were applied on the total respondents to extract the number of factors based on the criterion of eigenvalues greater than 1.00. The Kaiser–Meyer–Olkin (KMO) measure of testing the sampling adequacy and the Bartlett’s test of sphericity were used to verify the suitability of the EFA. The KMO value was compared to that of the standard adequacy (>.80; Kaiser, 1974). The extraction criterion was >.30 for each item loading to ensure that the item loadings were selected based on the most conceptually possible model (Kline, 1986).

Results

Demographic and Clinical Characteristics

The demographic and clinical characteristics of the surveyed individuals are shown in Table 1. Most of the participants were female (62.4%), single (61.5%), had attained an education to the secondary school level (72.0%), and had a mean age of 43.93 years (SD = 11.41). About two thirds of the participants were diagnosed with schizophrenia (63.9%), followed by depression (23.7%), and anxiety disorders (9.0%). Most of the participants (80.3%) had suffered from mental illness for 5 years or more, with a mean of 14.73 years (SD = 10.33). The majority (82.40%) admitted into a psychiatric hospital at least once, with a mean of 3.02 (SD = 3.39). Over half (60.0%) were living at home, while one third (33.7%) were in psychiatric halfway houses and a small number (6.3%) in residential homes. One fifth (20.0%) engaged in full- or part-time open employment, while one third (33.2%) worked at sheltered workshops/supported employment and another one third (30.3%) were unemployed. Two thirds (69.4%) relied on governmental subsidies for support of livelihood.

Table 1.

Demographic and Clinical Characteristics of Study Participants.

Characteristics (N = 295)	Valid Data	n	%
Sex	295
Male		111	37.6
Female		184	62.4
Diagnosis	266
Schizophrenia		170	63.9
Depression		63	23.7
Anxiety disorders		24	9.0
Other		9	3.4
Marital status	294
Single		181	61.5
Married		49	16.7
Divorced		49	16.7
Separated, widowed		15	5.1
Education level	293
Postsecondary		34	11.6
Secondary		211	72.0
Primary		47	16.0
Other		1	.3
Occupational status	291
Full-time job		29	10.0
Part-time job		29	10.0
Sheltered workshop/supported employment		97	33.2
Retired, housewife		48	16.5
Unemployed		88	30.3
Place of residence	288
Home		173	60.0
Half-way house		97	33.7
Residential home		18	6.3
Source of income	288
From employment		46	16.0
From family member		42	14.6
From government subsidies		200	69.4
	Valid data	Mean	SD
Age	292	43.93	11.41
Period of illness	262	14.73	10.33
Number of hospitalizations	273	3.02	3.39

Scores on the ISMI and Related Measures

The scores for each measure was calculated, including the total scores, mean, standard deviations, skewness, and kurtosis (see Table 2). The total scores (SD) for the measures were 24-item ISMI-C 56.93 (11.70), SR 14.20 (2.52), overall QoL 3.41 (.99), and RSES 26.32 (4.13). The overall QoL was significantly and negatively skewed, the 24-item ISMI-C and RSES had significantly high kurtosis, while the SR did not show any significant skewness and kurtosis. In order to address the skewness and kurtosis, the original scores of the measures were squared to normalize the distributions. However, the results did not show any significant differences between the original and the transformed. Thus, the original data were used.

Table 2.

Descriptive Statistics (Total Scores, Mean, Standard Deviations, Skewness, and Kurtosis) of the Scores on the Measures.

	Total (SD)	M (SD)	Skewness	Kurtosis
ISMI-C (24 items)	56.93 (11.70)	2.37 (.49)	.34 (ns)	−.20*
SR	14.20 (2.52)	2.84 (.50)	−.11 (ns)	.31 (ns)
RSES	26.32 (4.13)	2.64 (.41)	−.08 (ns)	.62*
Overall QoL	3.41 (0.99)	3.41 (.99)	−.49*	−.38 (ns)

Note. ns = nonsignificant; ISMI-C = Internalized Stigma of Mental Illness scale (Chinese); SR = stigma resistance subscale; Overall QoL = overall quality of life assessed by the WHOQOL-BREF (HK); RSES = Rosenberg Self-Esteem Scale

*p ≤ .05.

The individual items of the 24-item ISMI-C had a mean score that ranged from 2.09 to 2.75 and SD that ranged from .74 to .85. The individual items of the SR had a mean score that ranged from 1.94 to 2.59 and SD from .62 to .84. These results are comparable with those reported by Ritsher et al. (2003).

Internal Consistency Reliability

The Guttman split-half reliability and Cronbach’s α coefficient for the 24-item ISMI-C were .85 and .93, respectively. There were no significant improvements in reliability related to the removal of any individual item in the 24-item ISMI-C. The Cronbach’s α coefficients for the four subscales of the 24-item ISMI-C were alienation (α = .85), stereotype endorsement (α = .75), perceived discrimination (α = .84), and social withdrawal (α = .86). The Cronbach’s α coefficient of the SR was .72. All of the 5 items of the SR, that is, Items 25, 26, 27, 28, and 29, had poor corrected item-total correlation with the 24-item ISMI-C (r < .3).

Concurrent Validity

The results of the Pearson’s correlation analysis conducted on the 24-item ISMI-C, SR, QoL, and RSES are reported in Table 3. As expected, the results show that the total score of the 24-item ISMI-C is moderately, significantly, and negatively correlated with the total score of the overall QoL (r = −.31, p ≤ .001) and RSES (r = −.45, p ≤ .001).

Table 3.

Correlation Among ISMI-C Subscales, RSES, and Overall QoL.

	ISMI-C (24 Item)	Alienation	Stereotype	Perceived Discrimination	Social Withdrawal	SR
ISMI-C (24 items)	1	.854**	.801**	.857**	.872**	−.013 (ns)
Alienation	.854**	1	.595**	.633**	.645**	−.106 (ns)
Stereotype	.801**	.595**	1	.567**	.555**	−.095 (ns)
Perceived discrimination	.857**	.633**	.567**	1	.735**	.071 (ns)
Social withdrawal	.872**	.645**	.555**	.735**	1	.092 (ns)
SR	−.013 (ns)	−.106 (ns)	−.095 (ns)	.071 (ns)	.092 (ns)	1
RSES	−.447**	−.434**	−.372**	−.292**	−.404**	.476**
Overall QoL	−.306**	−.288**	−.207^**	−.277^**	−.264^**	.361^**

**p ≤ .01; *p ≤ .05.

Also, as expected, the four subscales of the 24-item ISMI-C, that is, alienation, stereotype endorsement, perceived discrimination, and social withdrawal, are significantly, positively, and moderately to highly correlated with each other, and significantly, positively, and moderately correlated with the overall QoL and RSES.

On the other hand, the SR is not related to the four subscales of the 24-item ISMI-C and 24-item ISMI-C total score but positively and moderately related to the overall QoL (r = .36) and RSES (r = .48).

Construct Validity

Construct validity was tested on the factor-based validity of the ISMI-C. In the study of the original ISMI carried out by Ritsher et al. (2003), a factor analysis was conducted on 24 items due to the small sample size (N = 127). In this study, a factor analysis is implemented on the 29-item ISMI-C as the sample size is large enough to do so (N = 295). The KMO measure value for the sampling adequacy of the 29-item ISMI-C was .91, and the χ² value of the Bartlett’s test of sphericity was 3,732.26 (p < .001). These results demonstrate common factors in the 29-item ISMI-C, and thus an EFA was performed based on the criterion of eigenvalues greater than 1.00. The factor analysis yielded five factors with factor loadings that ranged from .39 to .82. These five factors could explain for 55.79% of the total variance in the results (see Table 4). The factor loadings on the five factors were highly consistent with the results of the factor analysis carried out by the Ritsher et al. (2003), except for 2 items, that is, Items 21 and 25. Items 21 and 25 were heavily loaded on different factors than those reported by Ritsher et al. (2003).

Table 4.

Exploratory Factor Analyses: Factor Loadings of Five-Factor Model of 29-Item ISMI-C.

Item				EFA
Item		Mean	SD	1	2	3	4	5
Factor 1: perceived discrimination
14	People discriminate against me because I have a mental illness	2.47	.814	.739	.318	.164	.095	−.026
15	Others think that I can’t achieve much in life because I have a mental illness	2.46	.792	.686	.301	.156	.184	.134
16	People ignore me or take me less seriously just because I have a mental illness	2.41	.793	.643	.289	.366	.059	−.107
21	Negative stereotypes about mental illness keep me isolated from the “normal” world	2.48	.806	.545	.361	.379	.191	−.115
18	Nobody would be interested in getting close to me because I have a mental illness	2.30	.753	.545	.110	.415	.298	−.052
17	People often patronize me, or treat me like a child, just because I have a mental illness	2.36	.781	.491	.193	.238	.396	−.150
Factor 2: alienation
4	I am embarrassed or ashamed that I have a mental illness	2.42	.798	.162	.767	.074	.067	−.045
5	I am disappointed in myself for having a mental illness	2.49	.780	.250	.755	.088	.173	.096
3	People without mental illness could not possibly understand me	2.75	.787	.186	.626	.324	.114	.048
2	Having a mental illness has ruined my life	2.34	.824	.236	.597	.302	.251	.114
6	I feel inferior to others who don’t have a mental illness	2.47	.810	.253	.588	.190	.269	.212
1	I feel out of place in this world because I have a mental illness	2.39	.768	.203	.573	.199	.321	.176
Factor 3: social withdrawal
22	I refrain from social situations in order to protect my family or friends from embarrassment	2.35	.818	.376	.156	.656	.151	.012
25	I feel comfortable being seen in public with an obviously mentally ill person	2.62	.780	.195	.017	−.650	−.211	.246
19	I don’t talk about myself much because I don’t want to burden others with my mental illness	2.60	.804	.226	.279	.624	.047	.059
24	I avoid getting close to people who don’t have a mental illness to avoid rejection	2.29	.786	.388	.120	.606	.174	.125
20	I don’t socialize as much as I used to because my mental illness might make me look or behave “weird”	2.37	.854	.323	.301	.578	.172	.014
23	Being around people who don’t have a mental illness makes me feel out of place or inadequate	2.34	.812	.396	.293	.536	.145	.051
Factor 4: stereotype endorsement
9	Mentally ill people tend to be violent	2.09	.741	.164	−.040	.116	.665	−.028
8	People can tell that I have a mental illness by the way I look	2.25	.748	.144	.208	.060	.644	−.142
11	People with mental illness cannot live a good, rewarding life	2.22	.783	.297	.195	−.026	.584	.281
7	Stereotypes about the mentally ill apply to me.	2.29	.761	-.120	.317	.245	.571	.080
10	Because I have a mental illness, I need others to make most decisions for me	2.42	.772	.029	.191	.211	.556	.066
12	Mentally ill people shouldn’t get married	2.14	.805	.354	.017	.148	.415	.327
13	I can’t contribute anything to society because I have a mental illness	2.19	.814	.388	.249	.154	.391	.190
Factor 5: stigma resistance
27	I can have a good, fulfilling life, despite my mental illness	2.52	.82	.012	.083	.021	.069	.817
26	In general, I am able to live life the way I want to	2.43	.78	.109	.076	−.161	−.099	.717
29	Living with mental illness has made me a tough survivor	2.59	.84	−.067	.084	−.070	.027	.696
28	People with mental illness make important contributions to society	1.94	.62	−.126	−.057	.159	.071	.692
	Eigenvalue			3.65	3.65	3.30	2.94	2.65
	% Variance explained			12.59	12.53	11.38	10.44	9.14
	Cumulative % variance explained			12.59	25.12	36.50	46.64	55.79

Note. EFA= Exploratory factor analysis; ISMI-C= Internalized Stigma of Mental Illness scale (Chinese). Each item is rated on a 4-point Likert-type scale with 1 representing strongly disagree and 4 representing strongly agree, and 2.5 is the midpoint. Bold figures indicate loadings > 0.3.

Discussion and Applications to Social Work

In this study, the SR is not correlated with the total score of the 24-item ISMI-C and its four subscales. These results suggest that the 24-item ISMI-C and SR should be conceptualized as two separate scales as suggested by previous studies (Brohan et al., 2010; Sibitz et al., 2011; Tang & Wu, 2012).

In the present study, the Cantonese version of the 24-item ISMI-C and SR are shown to have good reliability. The Cronbach’s α coefficient for the 24-item ISMI-C was >.90, thus indicating that the ISMI-C has high internal consistency reliability. The Cronbach’s α coefficient for the four subscales of the ISMI-C ranged from .75 to .86, which is comparable to the reliability reported in the original ISMI scale (Ritsher et al., 2003). The Cronbach’s α coefficient for the SR was .72, thus suggesting acceptable internal consistency reliability.

The Cantonese version of the ISMI is therefore shown to have good concurrent and construct validities. As expected, the 24-item ISMI-C as well as its four subscales demonstrate significant, moderate, and negative correlation with the RSES and overall QoL as assessed by using the WHOQOL-BREF (HK). These results are consistent with those of the previous studies (Gerlinger et al., 2013; Livingston & Boyd, 2010; Lysaker et al., 2007; Ritsher et al., 2003; Tang & Wu, 2012). On the other hand, the SR has demonstrated significant, moderate, and positive correlation with the RSES and overall QoL as assessed by using the WHOQOL-BREF (HK), which is also consistent with the results reported by previous studies (Lysaker et al., 2007; Ritsher et al., 2003; Sibitz et al., 2011; Tang & Wu, 2012). These research results indicate that the 24-item ISMI-C and SR have satisfactory concurrent validity.

An EFA of the 29-item ISMI-C, including the 24-item ISMI-C and the 5-item SR, has yielded five factors based on the criterion of eigenvalues greater than 1. The extracted factors are highly consistent with the five factors reported by Ritsher et al. (2003), which are alienation, stereotype endorsement, perceived discrimination, social withdrawal, and SR. The factor loadings on the five factors are comparable with the factor analysis reported by Ritsher et al. (2003), except for Items 21 and 25. Although these 2 items have the strongest loading on the unexpected factors, they have the second highest loading on the expected factors as suggested by Ritsher et al. (2003). In fact, in the factor analysis of the 24-item ISMI (minus the 5-item SR) as reported by Ritsher et al. (2003), 13 of the 24 items have their strongest loadings on the unexpected factors but also their second highest loadings on the expected factors. Therefore, the above-mentioned research results indicate that the ISMI-C shows good construct validity.

Items 21 and 25 are heavily loaded on different factors than those reported by Ritsher et al. (2003), which may be due to problems of cross-cultural adaptation of the ISMI that has its origins in the Western context and deserves further explanation as follows.

Traditional Chinese cultural values contribute to the stigmatization toward mental illness and the development of internalized self-stigma in people with mental illness (Fabrega, 1991; Lam et al., 2010). Due to the stigmatization of mental illness in traditional Chinese cultural values, especially those of Confucianism origins, people with mental illness are perceived as incompetent members of society, those who fail to comply with the five cardinal relations of Confucianism (respect between individuals in the following five types of relationships: ruler and subject, father and son, husband and wife, older and younger brother, and friends), and subsequently those who have become a “disgrace” in the family and social systems (Lam et al., 2010). In addition, the concept of “fate,” which reflects acceptance of one’s role in the unchangeable destiny set by transcendence, facilitates the internalization of public stigma and development of self-stigma (Lam et al., 2010). Moreover, relatives of individuals with mental illness also experience “disgrace” in the community, which causes their rejection and condemnation of mental illness sufferers. Under such cultural influences, people with mental illness are more likely to endorse public stigma and develop self-stigma themselves. Subsequently, these self-stigmatized individuals experience “shame,” “disgrace,” “loss of face (reputation),” and “devaluation” in their family, guilty feelings of being dependent, and they worry that they will constitute the family burden. Thus, it is better for a measuring scale on self-stigma of mental illness to assess shame, disgrace, loss of face, devaluation, “guilty feelings,” “being a burden,” and related experiences on self-stigma. So cross-cultural adaptation and modification of the ISMI are needed in order to measure the culturally specific self-stigma experiences of Chinese people with mental illness in the local context.

In order to adapt the ISMI-C to a Chinese context, the following amendments are suggested (see Table 5 for the summary). First, due to the influence of Chinese culture, Item 21 is reclassified into Factor 1 (perceived discrimination) rather than Factor 4 (stereotype endorsement) as indicated in Ritsher et al. (2003). In order to render Item 21 to be more closely related to Factor 1 (perceived discrimination), Item 21 (negative stereotypes about mental illness keep me isolated from the normal world) should be changed (negative stereotypes about mental illness have forced me to be isolated from the normal world).

Table 5.

Suggested Revisions to Items in ISMI for Cultural Relevance.

Factor 1: perceived discrimination
14	People discriminate against me because I have a mental illness
15	Others think that I can’t achieve much in life because I have a mental illness
16	People ignore me or take me less seriously just because I have a mental illness
21	Negative stereotypes about mental illness isolated from the “normal” world. Negative stereotypes about mental illness have forced me to be isolated from the normal world
18	Nobody would be interested in getting close to me because I have a mental illness
17	People often patronize me, or treat me like a child, just because I have a mental illness
Factor 2: shame
4	I am ashamed that I have a mental illness. I am ashamed that I have a mental illness
5	I am disappointed in myself for having a mental illness
3	People without mental illness could not possibly understand me
2	Having a mental illness has ruined my life
6	I feel inferior to others who don’t have a mental illness
1	I feel out of place in this world because I have a mental illness
Factor 3: social withdrawal
22	I stay away from social situations in order to protect my family or friends from embarrassment
25	I comfortable being seen in public with an obviously mentally ill person. I don’t feel comfortable being seen in public with an obviously mentally ill person
19	I don’t talk about myself much because I don’t want to burden others with my mental illness
24	I avoid getting close to people who don’t have a mental illness to avoid rejection
20	I don’t socialize as much as I used to because my mental illness might make me look or behave “weird”
23	Being around people who don’t have a mental illness makes me feel out of place or inadequate
Factor 4: stereotype endorsement
9	Mentally ill people tend to be violent
8	People can tell that I have a mental illness by the way I look
11	People with mental illness cannot live a good, rewarding life
7	Stereotypes about the mentally ill apply to me
10	Because I have a mental illness, I need others to make most decisions for me, Because I have a mental illness, I have become dependent and need to rely on the care of others
12	Mentally ill people shouldn’t get married
13	I can’t contribute anything to because I have a mental illness I can’t contribute anything to my family or society because I have a mental illness
Factor 5: stigma resistance
27	I can have a good, fulfilling life, despite my mental illness
26	In general, I am able to live life the way I want to
29	Living with mental illness has made me a tough survivor
28	People with mental illness make important contributions to society

Note. Bold sentences indicate the suggested revised items in ISMI.

Second, due to the influence of Chinese culture, Item 25 is unexpectedly heavily and negatively loaded on Factor 3 (social withdrawal) rather than Factor 5 (SR) as reported by Ritsher et al. (2003). In order to render Item 21 to be more closely related to Factor 3 (social withdrawal), Item 25 (I feel comfortable being seen in public with an obviously mentally ill person) should be changed (I don’t feel comfortable being seen in public with an obviously mentally ill person).

Third, Item 10 (Because I have a mental illness, I need others to make most of the decisions for me) under Factor 4 (stereotype endorsement) should be changed (Because I have a mental illness, I have become dependent and need to rely on the care of others). Also, Item 13 (I can’t contribute anything to society because I have a mental illness) should also be changed (I can’t contribute anything to my family or society because I have a mental illness). These amendments will then enable the ISMI-C to assess the experiences of being dependent and a family burden after diagnosis of mental illness under the influence of Chinese culture and values.

Fourth, in order to allow the ISMI-C to assess the experiences of shame, disgrace, and guilt of Chinese individuals with mental illness, several amendments are suggested, such as for Factor 2 (alienation). Factor 2 itself should be designated as shame as opposed to “alienation”. Also under this factor, Item 4 (I am embarrassed or ashamed that I have a mental illness) should be changed (I am ashamed that I have a mental illness). Consequently, these amendments enable the ISMI-C to assess the experiences of shame that is a common and important element of self-stigmatization of mental illness under the influence of Chinese culture and values.

As self-stigmatization has adverse effects on the life of mental illness sufferers, it is important to develop therapeutic interventions that reduce self-stigma for these individuals. Research studies in the Western countries have suggested that it is feasible to reduce self-stigma through various strategies, such as psychoeducation with cognitive–behavioral therapy (Mittal, Sullivan, Chekuri, Allee, & Corrigan, 2012). In particular, some short-term therapeutic groups that are conducted with ISMI as the outcome assessment tool are found to be effective in reducing self-stigma for people with mental illness (Luckstead et al., 2011; Yanos, Roe, West, Smith, & Lysaker, 2012). This study shows that the ISMI-C is a reliable and valid measure of self-stigma for Chinese people with mental illness, and the development of the ISMI-C can certainly facilitate the development of interventions that are effective in reducing the self-stigmatization of these individuals. On the other hand, ISMI-C can also facilitate comparison studies on the experiences of self-stigma across cultures and societies.

Although the ISMI-C has been shown to be a feasible measure in this study, there are several methodological limitations and these warrant attention. First, the generalizability of the research result is limited by the studied sample who is Chinese consumers of community-based mental health services provided by local NGOs in Hong Kong and Macau. This studied sample does not represent the whole Chinese population who have a mental illness, especially those who live in regions outside of Hong Kong and Macau and cannot understand Cantonese. Also, this studied sample does not represent the whole Chinese population with mental illness, such as those with neurocognitive disorders, comorbid diagnosis of intellectual disability, hospital inpatients, and under 18 years of age. So more research studies are needed to determine the reliability and validity of the ISMI-C for people of different clinical and demographical characteristics. Second, although the interrater reliability of the ISMI has been tested to be satisfactory by other studies (Ritsher et al., 2003), it is lacking in the present study. Also, the sensitivity to change of the ISMI has not been previously assessed. Therefore, further studies are recommended which examine these psychometric properties of the ISMI-C among Chinese consumers of mental health services.

Conclusion

This study has shown that the Cantonese version of the ISMI scale (ISMI-C) is a reliable and valid measure of self-stigma for Chinese people with mental illness who are participating in community-based mental health services. The development of the ISMI-C can facilitate comparison studies on the experiences of self-stigma across cultures as well as the development of interventions that are effective in reducing self-stigma for people with mental illness across Chinese societies.

Footnotes

Acknowledgements

Special thanks are given to the following non-government organizations for their kind support to this research project: Caritas Hong Kong Social Service Division, Christian Family Service Centre and Richmond Fellowship of Hong Kong. The authors would like to sincerely thank the participants and staffs for their considerate cooperation.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was generously sponsored by the Research Committee of Hong Kong Baptist University.

References

American Psychiatric Association. (2005). Psychiatric research report. Washington, DC: Author.

Brohan

Elgie

Sartorius

Thornicroft

(2010). Self-stigma, empowerment and perceived discrimination among people with Schizophrenia in 14 European countries: The GAMIAN-Europe study. Schizophrenia Research, 122, 232–238.

Byrne

(1999). Stigma of mental illness: Changing minds, changing behaviour. British Journal of Psychiatry, 174, 1–2.

Chung

K. F.

Wong

M. C.

(2004). Experience of stigma among Chinese mental health patients in Hong Kong. Psychiatric Bulletin, 28, 451–454.

Corrigan

P. W.

Roe

Tsang

W. H.

(2011). Challenging the stigma of mental illness: Lessons for therapists and advocates. Chichester, England: Wiley-Blackwell.

Corrigan

P. W.

Watson

A. C.

Barr

(2006). The self-stigma of mental illness: Implications for self-esteem and self-efficacy. Journal of Social and Clinical, 25, 875–884.

Fabrega

(1991). Psychiatric stigma in non-Western societies. Comprehensive Psychiatry, 32, 534–551.

Gerlinger

Hauser

Hert

M. D.

Lacluyse

Wampers

(2013). Personal stigma in schizophrenia spectrum disorders: A systematic review of prevalence rates, correlates, impact and interventions. World Psychiatry, 12, 155–164.

Kaiser

H. F.

(1974). An index of factorial simplicity. Psychometrika, 39, 31–36.

10.

Kline

(1986). A handbook of test construction. London, England: Methuen.

11.

Lam

C. S.

Tsang

H. W. H.

Corrigan

P. W.

Lee

Y. T.

Beth

Kan

… Larson

J. E.

(2010). Chinese lay theory and mental illness stigma: Implications for research and practices. Journal of Rehabilitation, 76, 35–40.

12.

Leung

K. F.

Wong

W. W.

Tay

M. S.

Chu

M. M.

S. S.

(2005). Development and validation of the interview version of the Hong Kong Chinese WHOQOL-BREF. Quality of Life Research, 14, 1413–1419.

13.

Leung

S. O.

Wong

P. M.

(2008). Validity and reliability of Chinese Rosenberg self-esteem scale. New Horizons in Education, 56, 62–69.

14.

Livingston

J. D.

Boyd

J. E.

(2010). Correlates and consequences of internalized stigma for people living with mental illness: A systematic review and meta-analysis. Social Science and Medicine, 71, 2150–2161.

15.

Luckstead

Drapalski

Calmes

Forbes

DeForge

Boyd

(2011). Ending self-stigma: Pilot evaluation of a new intervention to reduce internalized stigma among people with mental illness. Psychiatric Rehabilitation Journal, 35, 51–54.

16.

Lysaker

P. H.

Roe

Yanos

P. T.

(2007). Toward understanding the insight paradox: Internalized stigma moderates the association between insight and social functioning, hope, and self-esteem among people with schizophrenia spectrum disorders. Schizophrenia Bulletin, 33, 192–199.

17.

Mittal

Sullivan

Chekuri

Allee

Corrigan

P. W.

(2012). Empirical studies of self-stigma reduction strategies: A critical review of the literature. Psychiatric Services, 63, 974–981.

18.

Orley

Saxena

Herrman

(1998). Quality of life and mental illness. British Journal of Psychiatry, 172, 291–293.

19.

Polit

D. F.

Beck

C. T.

(2006). Essentials of nursing research: Methods, appraisal, and utilization. Philadelphia, PA: Lippincott Williams & Wilkins.

20.

Ritsher

J. B.

Otilingam

P. G.

Grajiales

(2003). Internalized stigma of mental illness: Psychometric properties of a new measure. Psychiatry Research, 121, 31–49.

21.

Schmitt

D. P.

Allik

(2005). Simultaneous administration of the Rosenberg self-esteem scale in 53 nations: Exploring the universal and culture-specific features of global self-esteem. Journal of Personality and Social Psychology, 89, 523–642.

22.

Sibitz

Unger

Woppmann

Zidek

Amering

(2011). Stigma resistance in patients with Schizophrenia. Schizophrenia Bulletin, 37, 316–323.

23.

Skevington

S. M.

Lotfy

O’Connell

K. A.

(2004). The world health organization’s WHOQOL-BREF quality of life assessment: Psychometric properties and results of the international field trial—A report from the WHOQOL group. Quality of Life Research, 13, 299–310.

24.

Tang

I. C.

H. C.

(2012). Quality of life and self-stigma in individuals with Schizophrenia. Psychiatric Quarterly, 83, 497–507.

25.

Tsang

W. H.

Tam

K. C.

Chan

Cheung

W. M.

(2003). Stigmatizing attitudes towards individuals with mental illness in Hong Kong: Implication for their recovery. Journal of Community Psychology, 31, 383–396.

26.

West

M. L.

Yanos

P. T.

Smith

S. M.

Roe

Lysaker

P. H.

(2011). Prevalence of internalized stigma among persons with severe mental illness. Stigma Research Action, 1, 3–10.

27.

Yanos

P. T.

Roe

West

M. L.

Smith

S. M.

Lysaker

P. H.

(2012). Group-based treatment for internalized stigma among persons with severe mental illness: Findings from a randomized controlled trial. Psychological Services, 9, 248–258.