Mental Illness
The Stigma of Mental Illness: Using Segmentation for Social Change
Marie A. Yeh, Robert D. Jewell, and Veronica L. Thomas
Despite the unequivocal incidence and burden that mental illnesses place on the world, those with mental illness remain not only neglected but also deeply stigmatized across societies. The stigma that surrounds mental illness serves as a barrier to treatment and recovery, leading to serious negative consequences such as school failure, job loss, and suicide. While many large-scale social marketing efforts have found some success in reducing stigma, we contend that the recommended approaches, which utilize the input of people with mental illness and those close to them, are inadequate and that a deeper understanding of those who stigmatize is needed. This research first provides a comprehensive examination of the components that comprise stigma and then uses these components to segment thegeneral population. Theauthors thenpresent recommendations based on differences in the endorsement of stigma among these segments to inform policy and advocacy groups in developing more varied and potentially more effective social marketing campaigns.
Keywords: mental illness, segmentation, cluster analysis, social marketing
A s one of the largest causes of morbidity, mortality, and disability in the United States, mental illnesses affect over 25% of people in any given year (National Institute
of Mental Health 2008). Because of measurable costs in health care and social service, productivity losses due to unemployment and caregiver burdens, increased crime, and immeasurable costs such as opportunities lost and lowered quality of life, the impact of mental illness on society is massive (Kessler et al. 2008; Rice, Kelman, and Miller 1991). Despite the unequivocal incidence and burden that mental illnesses place on the world, those with mental illness remain not only neglected but also deeply stig- matized across societies, with people who suffer from mental illness treated in a wary and even punitive fashion (Hinshaw and Stier 2008; Horton 2007).
The stigmatization of those with mental illness is especially alarming because stigma is a fundamental cause of the health inequalities faced by those with mental illness (Hatzenbuehler, Phelan, and Link 2013). For those who seek treatment, the psychological effort to cope with stigma can have negative consequences on both their mental and physical health as stigma thwarts, undermines, or exacerbates several problems, including the availability of resources, social isolation, and stress (Gross and Muñoz 1995). Many who suffer from mental illness never
seek treatment because of its stigma; these lower levels of help- seeking behavior are a major barrier to recovery (Cooper, Corrigan, and Watson 2003; Eisenberg et al. 2009). When mental disorders remain untreated, consequences such as school/job failure and suicidality can result (McGlashan 1999; Meltzer et al. 2003).
With such dire consequences, changing attitudes and beliefs to reduce stigma associated with mental illness and the people whosufferfromitisacriticallyimportantgoalofgovernmentand consumer groups hoping to prevent such negative consequences for people with mental illness. Indeed, reducing the stigma of mental illness is a stated priority by governmental and non- governmental organizations in the United States and Europe (U.S. Department of Health and Human Services 1999; World Health Organization 2013). As such, numerous social mar- keting campaigns and initiatives have been conducted with reducing stigma as a major goal. However, the evidence of the effectiveness of such efforts leaves much to be desired (Corrigan 2012). We believe the effectiveness of social marketing efforts in this arena could be significantly improved by enhancing social marketers’ application of marketing strategy.
Specifically, we propose that stronger segmentation of the general public according to their endorsements of the cognitive components that comprise the stigma of mental illness can inform the creation of campaigns and policies that address the concerns of specific target segments. The usefulness of seg- mentation on public health and policy issues is not new and has been used to inform environmental legislation and federal tax dollar allocation (Durand, Klemmack, and Roff 1982; Pilling, Crosby, and Ellen 1991). However, how segmentation can be used to inform social marketing and policy to reduce stigma for people with mental illness has not been examined. Specifically, there is a need to understand fully the cognitive representations of stigma among those who stigmatize the mentally ill. To date, existing research has explored the components that comprise
Marie A. Yeh is Assistant Professor, Sellinger School of Business and Management, Loyola University Maryland (e-mail: mayeh@loyola.edu). Robert D. Jewell is Professor, College of Business Administration, Kent State University (e-mail: rjewell1@kent.edu). Veronica L. Thomas is Assistant Professor of Marketing, Towson University (e-mail: vlthomas@ towson.edu). The authors thank the Dean’s Office of the Sellinger School of Business and Management at Loyola University Maryland for their kind financial support for this work. Stacey Menzel Baker served as associate editor for this article.
© 2017, American Marketing Association Journal of Public Policy & Marketing ISSN: 0743-9156 (print) Vol. 36 (1) Spring 2017, 97–116
1547-7207 (electronic) DOI: 10.1509/jppm.13.12597
stigma but has not focused specifically on those who stig- matize, potentially resulting in an incomplete understanding of stigmatizers and how best to reduce stigma. Research has also tended to focus on testing specific aspects of the mental illness stigma process rather than undertaking a compre- hensive view of the beliefs that constitute it.
This research examines the multifaceted aspects of stigma that have been identified as the primary contributors to stigma creation and illustrates and highlights how the general public may be segmented along their stigmatizing beliefs. Using cluster analysis on survey data, we use cognitions that represent all of the components of mental illness stigma to uncover different segments. We then discuss recommendations re- garding target market selection and the positioning of the issue of mental illness for social marketing campaigns and public policy actions on the basis of the target market selected.
Stigma “Stigma” refers to personal attributes that convey undesirable characteristics and stereotypes that “deeply discredit” (Goffman 1963, p. 15) the possessor. Mental illness stigma is the con- vergence of several interrelated components wherein people are first distinguished as different with a label like “mentally ill” (Link and Phelan 2001). This label is associated with negative attributes that comprise a stereotype that pervades public thinking about the issue, leading to the devaluation of people with mental illness. This devaluation then leads to a sense of “us versus them.” This separation leads to rejection and exclusion, resulting in discrimination and a loss of status. Thus, the stigma of mental illness can be defined as the process of beliefs and attitudes associated with the perception of mental illness that result in social distancing behaviors and sometimes outright discrimination (Jones et al. 1984; Link and Phelan 2001).
Stigmatization occurs on societal, interpersonal, and indi- vidual levels, with dynamically interrelated manifestations of different stigmas, such as structural, public, and self-stigma (Bos et al. 2013). The focus of stigma reduction in social marketing efforts is frequently on combatting public stigma, which is also thefocus of this research.Public stigma comprises people’s social and psychological reactions to someone whom they perceive to have a stigmatized condition (Bos et al. 2013). Theoretically, public stigma has been conceptualized as having cognitive and behavioral core features that include stereotypes (cognitive knowledge structures), prejudice (cognitive and emo- tional consequences of stereotypes), and discrimination (behav- ioral consequences of prejudice) (Corrigan 2000). Thus, the origin of stigmatization lies in the cognitive representations that people hold of those with mental illness (Bos et al. 2013).
Stigma Reduction Efforts Many stigma reduction efforts utilizing social marketing have been attempted over the years. The Center for Mental Health Services developed the Elimination of Barriers Initiative in 2003 as a national campaign to reduce the stigma of seeking care (Mulligan 2005). In 2007, the Substance Abuse and Mental Health Agency (SAMHSA) and the Ad Council in the United States launched the “What a Difference a Friend Makes” campaign (SAMHSA 2010; Suicide Prevention Resource Center 2007). Similar programs have been developed around the world, such as the “Like Minds, Like Mine” campaign in
New Zealand, “Open Doors” in Germany, and “Changing Minds” in the United Kingdom (Crisp et al. 2005; Gaebel et al. 2008; Vaughan and Hansen 2004).
While some research has indicated small reductions in stigma (Evans-Lacko et al. 2013) as a result of these social marketing campaigns, the campaigns have been criticized by experts as ineffective (Corrigan 2012). We propose that two factors may inhibit the effectiveness of social marketing campaigns to reduce stigma: (1) the general nature of the social marketing campaigns’ messages, with no specific target market; and (2) the method of message development used to inform the message strat- egy. Specifically, we contend that social marketing programs inadequately conduct market segmentation, targeting, and positioning.
Manypreviouscampaignshavepurportedtoutilizesocialmar- keting strategies applying commercial marketing techniques such as segmentation and targeting in an attempt to influence behaviors and beliefs about mental illness to alleviate stigma (Andreasen 1994). However, in practice, social marketers’ application of segmentation and targeting fall short. While some segmenting is done in these social marketing cam- paigns, it is done in a very general sense (Corrigan 2012). Specifically, many past campaigns have created general messages distributed to the general public. For example, the “What a Difference a Friend Makes” campaign sponsored by SAMHSA targeted people between 18 and 25 years of age and encouraged them to support friends experiencing mental health problems (SAMHSA 2010). Yet people in this age group are diverse and can and should be further segmented.
Corrigan (2012), in his analysis of public service announce- ments used to address mental health stigma, recommends that social marketing efforts be targeted toward key groups who may play critical roles in the lives of people with mental ill- ness, such asemployers, landlords, and media executives.While we agree that this more specific targeting strategy is a step in the right direction, it does not go far enough. Employers, for example, are likely a very heterogeneous population in terms of their endorsement of mental illness stigma. Moreover, for those employers focused solely on the profitability of their business, Corrigan’s suggestion to consider making legal accommodations for individuals with mental illness may be alarming.
Rather than grouping people solely by their role, we need to conduct more research to identify specific segments of people who engage differentially in the stigmatization process. When we attempt to reduce stigma, we need to understand more fully the heterogeneity that exists among those without mental ill- ness, because within this group can be found those who stig- matize, or, as we call them, stigmatizers. Stigmatizers craft lay theories grounded in their cognitive, affective, and behavioral reactions toward those with mental illness, leading to the devel- opment of stereotypes, which, in the aggregate, create a public stigma that pervades the culture (Bos et al. 2013). The com- ponents that comprise stigma are multifaceted and multi- dimensional, suggesting significant variation in the factors that may drive stigma. Therefore, to effectively reduce stigma, marketers need to target stigmatizers by understanding the cognitions and beliefs that comprise the lay theories that lead to the development of mental-illness stigma, and create messages that specifically address those aspects rather than employing general messages for general audiences.
98 The Stigma of Mental Illness
Current social marketing efforts, however, do not focus on understanding stigmatizers in an in-depth way. Message de- velopment for social marketing campaigns has relied heavily on the use of consensus development and community-based participatory research. Consensus development is often used by public policy and advocacy groups and requires gathering experts and relevant consumers affected by an issue to inform policy decisions (Clement et al. 2010; Murphy et al. 1998). Similarly, community-based participatory research—the in- clusion of mental health care consumers in all decisions in the creation of antistigma campaigns and the research surrounding it—has been recommended for all programs that challenge the public stigma of mental illness (Corrigan and Shapiro 2010). These consumers include people with mental illness, their family members, and service providers. Such approaches make sense because people with mental illness and their family members must have a voice in developing messages to ensure that social marketing campaigns do not further exacerbate stigma for those experiencing it as well as to create messages that ring true for their experience of stigma. Their inclusion in the social marketing process also serves to empower people with mental illness and those who love and work with them in fighting stigma, a valuable and relevant goal in and of itself in the fight to eliminate stigma.
However, the goal of stigma change is to reduce stigma held by people without mental illness to improve the experience of the stigmatized, as they are the victims of stigma. It is the stigmatizers who perpetuate stigma, and, therefore, they should be the target of social marketing campaigns. Social marketers must go beyond the consensus development and community- based participatory research approaches and study stigmatizers specifically when attempting to determine which components of stigma should be targeted with social marketing messages. Yet some existing studies of mental illness stigma among the general public have not asked participants whether they have a mental illness, which has surely resulted in inclusion of people with mental illness in these studies. The studies that have asked for this information have still included people with mental illness in their analysis, using their diagnosis as an indication for fa- miliarity with mental illness. Therefore, research specific to the general public without mental illness needs to be conducted to understand who stigmatizes those with mental illness, what their exact prejudices are, and how differences may exist within stigmatizers, to find distinct segments to inform the message creation process for social marketing campaigns.
The Components of Public Stigma of Mental Illness
In addition to failing to understand the aspects of stigma held by stigmatizers, current campaigns fall short in that they fail to fully consider all of the aspects that could potentially comprise stigma. Link and Phelan (2001) suggest that the process of mental illness stigmatization consists of beliefs that comprise a stereotype, affective reactions to the stereotype, the phenom- enon of othering those with mental illness, and behavioral en- dorsements about how stigmatizers might treat people with mental illness. We suggest that to find distinct groups that would inform differential social marketing actions, each component of this process of mental illness stigmatization should be used. We further explicate each component of the
process by consulting the literature on mental illness stigma, paying specific attention to studies of the general public in order to inform selection of constructs and their corresponding mea- sures that would differentiate groups. Interestingly, we do not find agreement about what cognitions comprise stigma; rather, variation exists regarding each of these components. Table 1 highlights the literature review results.
The Stereotype of Mental Illness To gain a comprehensive view of the first major component of the stigmatization process, the beliefs that comprise the ste- reotype of people with mental illness need to be explicated. Several beliefs have been repeatedly examined by mental health stigma researchers, such as beliefs that people with mental illness are dangerous, unpredictable, and incompetent. The perception of people with mental illness as being to blame for and needing to be ashamed of their illness as well as their prognosis for recovery also comprise the stereotype.
The most consistently studied belief in mental illness stigma is the conceptualization of people with mental illness as dan- gerous (Bos et al. 2013; Corrigan et al. 2003; Jones et al. 1984; Link and Phelan 2001; Star 1952, 1955). As first articulated by Star (1952, 1955), stigmatizers believe that people with men- tal illness are more violent, and thus more dangerous, than peo- ple without mental illness. Perceptions of dangerousness are widespread and continue to be consistently endorsed by the general public in the United States (Cohen and Struening 1962; Corrigan, Kuwabara, and O’Shaughnessy 2009; Crisp et al. 2000; Kobau et al. 2010; Perry et al. 2007; Pescosolido et al. 1996, 1999, 2010; Taylor and Dear 1981) as well as worldwide, in places like Germany (Angermeyer and Matschinger 2003a, b), the Netherlands (Van ’t Veer et al. 2006), India (Kermode et al. 2009), Japan, and Australia (Griffiths et al. 2006). Media portrayals of people with mental illness have further compounded this perception (Wahl 1992). Sometimes dangerousness is the only measure of stigma toward mental illness (Silton et al. 2011).
The perception of people with mental illness as unpredict- able and difficult has also often been studied (Angermeyer and Matschinger 2003a, b; Crisp et al. 2000; Griffiths et al. 2006; Kermode et al. 2009; Kobau et al. 2010). Mental illnesses are disorders that affect one’s mood, thinking, and behavior such that theyimpair one’snormal social functioning.Thus,themood, thinking, and social interactions of a person with mental illness often violate accepted norms of behavior, making the person unpredictable to others. The unexpectedness of their behavior may make others perceive people with mental illness as diffi- cult tointeract withor talkto.Thismayplayintostigmatizers’ be- liefs about fear, but apart from believing that people with mental illness may be dangerous, perceptions of people with mental illness as difficult to interact with also drive social distance.
Another consistently studied belief about people with mental illness is that they are incompetent (Star 1952, 1955). Per- ceptions of incompetence is a particularly important issue as, historically, some have argued that mental illness impairs decision making, rendering an individual with mental illness incompetent to make their own treatment decisions and therefore dependent and helpless (Pescosolido et al. 1999). Competence comprises not just an individual’s ability to manage their lives, but also their ability to work successfully (Cohen and Struening 1962; Kermode et al. 2009; Kobau
Journal of Public Policy & Marketing 99
T ab
le 1.
R ev ie w
of S ti gm
at iz at io n C on
st ru ct s U ti li ze d in
P re vi ou
s S tu d ie s
C it at io n
S te re ot yp
ic al
B el ie fs
A ff ec ti ve
B el ie fs
O th er in g
B eh av
io ra l
E n d or se m en ts
D an
ge ro u s
U n p re d ic ta b le /
D if fi cu lt
In co m p et en t
B la m e
S ec re cy
T re at m en t
E ff ic ac y
P it y
F ea r
A n ge r
D if fe re n ce
W il li n gn
es s
to C oe rc e
W il li n gn
es s
to H el p
A n g er m ey er
an d M at sc h in g er
(2 0 0 3 b )
x x
x x
x x
x x
x x
A n g er m ey er
an d M at sc h in g er
(2 0 0 3 a)
x x
x x
x x
x A n g er m ey er
et al . (2 0 0 3 )
x x
x A n g er m ey er
et al . (2 0 1 3 )
x x
x x
x x
x x
x A n g li n , L in k , an d P h el an
(2 0 0 6 )
x x
x B o ri n st ei n (1 9 9 2 )
x x
x x
x B o y d et
al . (2 0 1 0 )
x x
x x
C o h en
an d S tr u en in g (1 9 6 2 )
x x
x x
x x
x x
C o le m an
et al . (2 0 0 9 )
x C o rr ig an
et al . (2 0 0 3 )
x x
x x
x x
x C o rr ig an
et al . (2 0 0 9 )
x x
x x
x C ri sp
et al . (2 0 0 0 )
x x
x x
x E v an s- L ac k o et
al . (2 0 1 3 )
x x
x x
x x
x x
x G ae b el
et al . (2 0 0 8 )
x x
x G ri ff it h s et
al . (2 0 0 6 )
x x
x x
H u g o et
al . (2 0 0 3 )
x x
Jo rm
an d G ri ff it h s (2 0 0 8 )
x x
x Jo rm
et al . (1 9 9 9 )
x x
K er m o d e et
al . (2 0 0 9 )
x x
x x
K o b au
et al . (2 0 1 0 )
x x
x x
x x
x M eh ta
et al . (2 0 0 9 )
x x
x x
x x
x x
B la ck
et al . (1 9 9 3 )
x x
x x
N g , M ar ti n , an d R o m an s (1 9 9 5 )
x x
N o rm
an et
al . (2 0 0 8 )
x x
x x
x x
O ja n en
(1 9 9 2 )
x x
x P er ry
et al . (2 0 0 7 )
x x
x x
P es co so li d o et
al . (1 9 9 6 )
x x
x x
P es co so li d o et
al . (1 9 9 9 )
x x
x P es co so li d o et
al . (2 0 0 7 )
x x
P es co so li d o et
al . (2 0 1 0 )
x x
x x
P es co so li d o et
al . (2 0 1 3 )
x x
x x
x x
P h el an
an d B as o w
(2 0 0 7 )
x S il to n et
al . (2 0 1 1 )
x T ay lo r an d D ea r (1 9 8 1 )
x x
x x
x x
x x
V an
’t V ee r et
al . (2 0 0 6 )
x x
x x
x W al k er
et al . (2 0 0 8 )
x x
x W o lf f et
al . (1 9 9 6 )
x x
x x
x
100 The Stigma of Mental Illness
et al. 2010; Pescosolido et al. 1996; Taylor and Dear 1981; Van ’t Veer et al. 2006).
Perhaps the next most studied belief about people with mental illness is the notion of blame (Angermeyer and Matschinger 2003b; Cohen and Struening 1962; Corrigan, Kuwabara, and O’Shaughnessy 2009; Crisp et al. 2000; Evans- Lacko, Henderson, and Thornicroft 2013; Griffiths et al. 2006; Kermode et al. 2009; Kobau et al. 2010; Pescosolido et al. 1996, 2007; Taylor and Dear 1981). People with mental illness are more likely to be blamed for their illness than people with physical illnesses. Attribution theory states that when people make attributions about the cause and controllability of a per- son’sailment, theymakeinferencesaboutresponsibility(Weiner, Perry, and Magnusson 1988). High levels of attribution of personal responsibility for the onset of mental illness can evoke anger and prejudicial attitudes. Greater inclinations to blame the mentally ill have been associated with greater tendencies toward approval of structural discrimination (Angermeyer and Dietrich 2006; Angermeyer and Matschinger 2003a, b). Blame also carries over to beliefs about the ability of people with mental illness to “snap out of it” or recover without treatment on their own. These beliefs relate to control over mental illness, giving credence to the belief that people with mental illness could re- cover if only they wanted to.
Related to the concept of blame is the concept of shame (Griffiths et al. 2006; Parcesepe and Cabassa 2013; Walker et al. 2008). Shame involves negative feelings such as sadness, em- barrassment, or humiliation caused by having done something wrong. Often, people with mental illness cope with stigma by concealing their mental illness from others (Link et al. 1997). From the perspective of stigmatizers, however, shame is related to whether the stigmatizer feels a person with mental illness should be ashamed, which manifests through the endorsement of secrecy. Griffiths et al. (2006), for example, as a part of assessing mental illness stigma, ask whether the general public would advise a close friend or relative to not tell anyone about a mental illness.
Stereotypical beliefs also encompass thoughts about treatment efficacy or the prognosis for people with mental illness—in other words, the likelihood of recovery (Angermeyer and Matschinger 2003a, b; Crisp et al. 2000; Kobau et al. 2010; Pescosolido et al. 1996, 2007). These beliefs relate to the question of whether available treatments for mental illness are effective. If one can recover from mental illness, then one is deserving of help and assistance. However, if stigmatizers believe people with mental illness cannot recover, then we would expect increased social distance and discriminatory behavior as stigmatizers perceive people with mental illness to be a lost cause.
Affective Reactions to the Stereotype The endorsement of stereotypical beliefs has been found to lead to emotional reactions. Three types of emotional reactions to people with mental illness have been well studied: fear, anger, and pity(Angermeyer and Matschinger 1997,2003a,b; Corrigan et al. 2003). Fear is driven by perceptions of dangerousness and unpredictability. Fear of people with mental illness has re- peatedly been demonstrated as a driver for social distance (Angermeyer and Matschinger 2003a, b; Martin, Pescosolido, and Tuch 2000; Martin et al. 2007). Corrigan et al. (2003) find that willingness to help or hire people with mental illness is
significantly negatively correlated with perceptions of dan- gerousness, with fear functioning as a mediating mechanism between the two. Thus, the affective reaction of fear is a com- ponent of people’s cognitive representation of those with mental illness, accounting for the inclusion of fear in assessing the stigma of mental illness.
Anger and pity as emotional reactions are thought to be the result of attributions of blame. High levels of attribution of personal responsibility for the stigmatized condition evokes emotional responses and stigmatizing behavior (Bos et al. 2013), including punishing thoughts, such as willingness to coerce. Conversely, low attribution of personal responsibility can evoke sympathy or pity (e.g., “He can’t help himself be- cause of his mental illness”) and greater tendencies to help (Corrigan et al. 2003; Weiner, Perry, and Magnusson 1988). Pity may also be related to perceptions of people with mental illness as incompetent or, more benevolently, perceptions that they need help.
Othering of People with Mental Illness Theoretically, the stereotype of people with mental illness combined with negative affect work together to create an “us- versus-them” mentality among stigmatizers (Devine, Plant, and Harrison 1999; Link and Phelan 2001). People with mental illness are viewed as separate and apart from people without mental illness. This idea of othering can be thought of as a belief that people with mental illness are different from those without mental illness, which increases social distance (LinkandPhelan 2001). A person must first be distinguished as different, then labeled as a part of a group that is associated with negative at- tributes,beforestigmatizationcanoccur.Thus,peoplewithmental illness are considered “other” and members of an out-group.
Behavioral Endorsements Toward People with Mental Illness Believing that people with mental illness are dangerous and incompetent leads to a willingness to coerce people with mental illness into treatment (Cohen and Struening 1962; Taylor and Dear 1981). That is, stigmatizers endorse the belief that people with mental illness should be forced into treatment, believing that coercion is justified in order to protect society and to protect themselves, especially if they uphold the view that people with mental illness are incapable of making their own decisions.
Concern and the desire to help people with mental ill- ness have also been studied as components of mental illness stigma (Cohen and Struening 1962; Corrigan, Kuwabara, and O’Shaughnessy 2009; Evans-Lacko, Henderson, and Thornicroft 2013; Taylor and Dear 1981). Cohen and Struening (1962) describe these behaviors as benevolence, which they explain as “a sort of Christian kindliness toward unfortunates” (p. 353). More recent studies have examined people’s willingness to help those with mental illness directly or indirectly (e.g., advo- cating for resources for people with mental illness; Corrigan, Kuwabara, and O’Shaughnessy 2009; Evans-Lacko, Henderson, and Thornicroft 2013).
Study The goal of this study is to analyze the beliefs, affective re- actions, othering thoughts, and behavioral inclinations that
Journal of Public Policy & Marketing 101
comprise stigma among the general population. We believe that understanding stigmatizers’ cognitions will lead to the most effective messaging strategies to combat the perpetuation of stigma. We demonstrate that while stigmatizers are generally approached as a homogeneous group, they are quite heteroge- neous in their endorsement of stigmatizing thoughts and feel- ings. Thus, stigmatizers can be segmented into distinct groups, which can be selected for differential targeting for purposes of integrated communications. We then use measures of these thoughts and feelings to identify segments for whom differential social marketing strategy can be applied and test for differences in social marketing campaigns’ most desired outcomes, preju- dicial attitudes, social distance, and support behaviors.
Data Collection and Procedures We recruited 507 participants for the study via Amazon Me- chanical Turk; they were compensated for their time. Age range of the participants was 19–75 years, with an average age of 36.47 years (SD = 12.8); 54.2% were female; 78.1% were Caucasian; 18.1% of the participants reported that they currently had a mental illness, with another 7.7% reporting having had a mental illness in the past. Geographically, participants were fairly rep- resentative of the four regions of the United States as specified by the U.S. Census Bureau, with 19.8% located in the Northeast region, 23.7% in the Midwest, 33.5% from the South, and 19.9% from the West (2.6% of the sample took the survey from outside the United States). The survey was anonymous in order to de- crease socially desirable responding. Participants responded to the battery of measures conveying stigma that were randomly presented to eliminate order effects. Then participants completed profiling measures for attitude toward people with mental illness, support intentions, social distance, and demographics.
Measure Development To assess the thoughts and feelings that comprise mental illness stigma, we conducted a thorough assessment of the literature to determine how cognitions regarding mental illness stigma were measured. We gathered measures from several different studies because not all constructs are used consistently across public stigma studies (see Table 2). For example, many stigma studies utilize vignettes to describe a person with a mental illness. They do so to reduce socially desirable responding because many studies are conducted through interviews. However, vignettes are hypothetical and abstracted from real life, and rarely do people encounter a person with a mental illness who displays a systematic set of symptoms (Link et al. 2004). Therefore, some items required modification because they did not reference a specific person from a vignette. We also modified items to create a uniform set of surveyquestions that would allow for the greatest differentiation to guide the identification of groups.
We also took care to heed the recommendations of the mental health field in ensuring that the voices of people closest to the issue would inform this work. We do not criticize exist- ing stigma reduction efforts because they recommend soliciting the input of those who are most affected by the issue. Indeed, we agree that input from these people is crucial in crafting marketing messages that do not further exacerbate stigma, as well as in creating messages that resonate with and reflect their experience. Therefore, we heeded this recommendation. After
developing the initial pool of items from a review of the literature, we asked a panel of nine experts to provide feedback on all proposed items, in terms of both content and word- ing, for comprehensiveness of coverage of stigma cogni- tions and for item refinement. For example, the wording of “When you think of people with mental illness generally,” which prefaces most items, was suggested by the expert panel. All members of the panel work in some capacity with people with mental illness, most as therapists or in other support capacities. Several are researchers as well. Others serve in an advocacy capacity. For example, one serves as an unpaid community advocate and serves on the board of several mental health–related organizations. Five individuals also have close family members with mental illness, and at least two have a mental illness themselves.
Prior to the assessment of segmentation and profiling vari- ables in the survey of the general public, we provided the fol- lowing instructions: “Now we want your honest opinions and feelings about people with mental illness generally and the topic of mental illness. Mental illnesses include depression, bipolar disorder (i.e., manic-depression) and schizophrenia to name a few.” In previous studies, we have found that people confuse mental illness with mental retardation; thus, we also included the following instruction: “Mental illness is not the same as mental retardation or other developmental disabilities.” We sought to assess people’s existing perceptions and inclinations about people with mental illness that would be triggered when using the term “mental illness.” As such, we did not overly define the term because it might potentially introduce new knowledge or understanding of the term, whereas our goal was to trigger people’s existing stereotypes of people with mental illness.
Segmentation Variables Dangerousness was assessed with two items: “When you think of people with mental illness generally, how dangerous/violent do you think they are likely to be?” (1 = “not at all dangerous/ violent,” and 9 = “very dangerous/violent”; Corrigan et al. 2003; Crisp et al. 2000; Griffiths et al. 2006). Unpredictability and difficulty were measured with the following two items: “When you think of people with mental illness generally, to what extent do you think that they are unpredictable/difficult to deal with?” (1 = “not at all unpredictable/difficult,” and 9 = “very un- predictable/difficult”; Crisp et al. 2000; Kobau et al. 2010).
Perceptions of incompetence were measured using three items.Weassessedthisconstructwithitemsthatwerepositivebe- cause so many of the items referenced the negative aspects of the stereotype of mental illness. Thus, we assessed beliefs about the intelligence, creativity, and ability to be successful at work of people with mental illness (Jorm et al. 1999; Kobau et al. 2010; Norman et al. 2008; Van ’t Veer et al. 2006). Items included “How successful do you believe they can be at work compared to others?” (1 = “not at all successful,” and 9 = “very successful”), “How intelligent do you think that they are likely to be?” (1 = “not at all intelligent,” and 9 = “very intelligent”), and “To what extent do you think that they are creative?” (1 = “not at all creative,” and 9 = “very creative”). These items were reverse-coded for analysis.
Blame comprises beliefs about cause and controllability of the mental illness (Corrigan, Kuwabara, and O’Shaughnessy
102 The Stigma of Mental Illness
2009; Corrigan et al. 2003). Cause was assessed with the item “How responsible do you think a person is for causing his or her own mental illness?” (1 = “not at all responsible,” and 9 = “completely responsible”). Controllability was assessed using two items: “To what extent do you think they are able to pull themselves together without treatment?” (1 = “not at all able,” and 9 = “completely able”) and “How much control do you think a person has over the cause of his or her own mental illness?” (1 = “no control at all,” and 9 = “complete control”).
Shame encompasses negative feelings that people with mental illness may feel for having a mental illness. For people without mental illness, this manifests as beliefs about secrecy
(Griffiths et al. 2006; Walker et al. 2008). Two items assessed secrecy: “How much do you believe that once you have been treated for a mental illness, the best thing to do is keep it a secret?” (1 = “do not believe at all,” and 9 = “believe very strongly”) and “How likely would you be to advise a close friend or relative with a mental illness that he or she should not tell anyone about his or her mental illness?” (1 = “not at all likely,” and 9 = “very likely”).
Treatment efficacy assesses beliefs in prognosis and recovery (Crisp et al. 2000; Kobau et al. 2010). Two items assessed treatment prognosis: “To what extent do you believe they can eventually recover?” and “To what extent do you believe that
Table 2. Factors and Factor Loadings
Factors and Items Factor Loading Eigenvalue
Percentage of Variance
Cronbach’s Alpha
Factor 1: Menace 7.63 28.27 .91 ... how dangerous do you think they are likely to be?a .83 ... how violent do you think they are likely to be?a .79 ... to what extent do you feel frightened by them?a .77 ... to what extent do you feel threatened by them?a .73 How much do you believe that people with mental illness should be hospitalized because they pose a risk to the general public?
.71
How much do you believe that hospitalizing people with mental illness would reduce violent crime?
.69
... to what extent do you think that they are difficult to deal with?a .54
... to what extent do you think that they are unpredictable?a .52
... to what extent do you feel anger toward them?a .44
... to what extent do you feel irritation toward them?a .43 Factor 2: Nurturance 2.78 10.3 .72 ... to what extent do you feel pity for them?a .70 ... how much concern do you feel toward them?a .70 ... to what extent do you feel the desire to help them?a .63 ... to what extent do you believe that they would improve if given treatment and support?a
.62
... to what extent do you believe they can eventually recover?a .60 Factor 3: Blame 2.3 8.51 .72 How responsible do you think a person is for causing his or her own mental illness? .75
... to what extent do you think they are able to pull themselves together without treatment?a
.73
How much control do you think a person has over the cause of his or her own mental illness?
.71
Factor 4: Shame 1.39 5.13 .79 How much do you believe that once you have been treated for a mental illness, the best thing to do is keep it a secret?b
.87
How likely would you be to advise a close friend or relative with a mental illness that he or she should not tell anyone about his or her mental illness?b
.86
Factor 5: Difference 1.3 4.8 .61 To what extent do you perceive yourself and people with mental illness to be a part of the same group?b
.68
... to what extent do you believe that they feel differently from a person without mental illness?a
.61
How similar do you think people with mental illness generally are to you?b .52 ... to what extent do you believe they feel the same way we all do at times?a,b .50 Factor 6: Incompetence 1.1 4.08 .73 ... how intelligent do you think that they are likely to be?a,b .84 ... to what extent do you think that they are creative?a,b .78 ... how successful do you believe they can be at work compared to others?a,b .53
aItems are prefaced with “When you think of people with mental illness generally ....” bItems were reverse-coded for analysis.
Journal of Public Policy & Marketing 103
they would improve if given treatment and support?” (1 = “do not believe at all,” and 9 = “believe very strongly”). Recovery was assessed with the item “To what extent do you believe they can eventually recover?” using the same scale.
Affective reactions assessed in stigma studies include fear, anger, and pity (Angermeyer and Matschinger 2003a, b; Corrigan et al. 2003). Items used to assess these emotions asked “To what extent do you feel [emotional reaction] towards them?” Fear was also assessed using two items asking whether participants felt “threatened/frightened by them” (1 = “do not feel threatened/frightened at all,” and 9 = “feel very threatened/frightened”). Anger was measured by two items asking whether participants felt “anger/irritation towards them” (1 = “feel no anger/irritation at all,” and 9 = “feel a great deal of anger/irritation”). Pity was measured with one item asking whether participants felt “pity towards them” (1 = “feel no pity at all,” and 9 = ”feel a great deal of pity”).
The “us-versus-them” mentality, or othering, encompasses beliefs about difference that can be captured through a sense of identification with the group and with the experience of being mentally ill (Crisp et al. 2000; Kobau et al. 2010; Rüsch et al. 2009). We assessed group identification with two items: “To what extent do you perceive yourself and people with mental illness to be a part of the same group?” (1 = “not at all,” and 9 = “very much”) and “How similar do you think people with mental illness generally are to you?” (1 = “not at all similar,” and 9 = “very similar”). Two items assessed identification with the experience of mental illness: “To what extent do you believe that they feel differently from a person without mental illness?” (1 = “does not feel differently at all,” and 9 = “feels very differently”) and “To what extent do you believe they feel the same way we all do at times?” (1 = “do not believe at all,” and 9 = “believe very strongly”).
Behavioral predispositions of willingness to coerce and to help examined how they would treat people with mental illness (Angermeyer and Matschinger 2003a, b; Corrigan et al. 2003). Willingness to help was measured with two items: “How much concern do you feel towards them?” (1 = “feel no concern at all,” and 9 = “feel a lot of concern”) and “To what extent do you feel the desire to help them?” (1 = “feel no desire at all,” and 9 = “feel a great desire”). Willingness to coerce into treatment was measured with two items: “How much do you believe that people with mental illness should be hospitalized because they pose a risk to the general public?” and “How much do you believe that hospitalizing people with mental illness would reduce violent crime?” (1 = “do not believe at all,” and 9 = “believe very strongly”).
Profiling Variables Segmentation analysis involves not only identifying variables that can be used to identify distinct segments but also profiling variables which are used to describe and inform social marketing action. We employ in this study traditional demographic in- formation measuring age, gender, marital status, race, and ed- ucational attainment. We also use the latitude and longitude data provided by Qualtrics to determine respondents’ geographic location. In addition, we examine familiarity with mental illness, attitude toward people with mental illness, support intentions, and social distance.
Familiarity with mental illness has been demonstrated repeatedly to be associated negatively with social distance (Corrigan et al. 2003). We assessed familiarity with two items asking whether participants had a close friend or family member with a mental illness and whether participants had ever worked with people with mental illness. While many studies reviewed measure attitudes toward people with mental illness by asking questions that determine whether people ascribed to stereo- typical beliefs, we use the more traditional measure to capture people’s generally favorable or unfavorable feelings about people with mental illness. In our study, three items asked “In general, when you think of people with mental illnesses, how do you feel about them as a group?” (1 = “bad/negative/ unfavorable,” and 9 = “good/positive/favorable”; a = .96). While we use as segmentation variables behavioral endorse- ments of willingness to help, we include two behavioral in- tentions that take people’s willingness to help a step further by identifying people’s intentions to support those with mental illness. These items measured willingness to donate money to organizations that fight mental illness stigma and willingness to participate in support programs for people with mental illness (1 = “not at all willing,” and 9 = “very willing”).
Social distance was measured with nine items assessing social settings of importance to people with mental illness. These items have been successfully used to test for differential responses to messaging on mental illness (Bogardus 1925; Yeh, Jewell, and Hu 2013; Yeh and Jewell 2015). Items were phrased in the first person to engender participants’ direct feelings, with responses given on a nine-point scale anchored at 1 = “strongly disagree,” and 9 = “strongly agree.”
Results
Segmentation Analysis First, we conducted an exploratory factor analysis to uncover the underlying latent constructs that comprise public stigma that would drive group differences. The focus of our study is on identifying segments of stigmatizers among the general public. Because public stigma consists of the stigmatizing thoughts and feelings of stigmatizers, by default, people with mental illness should be excluded because they are the stigmatized. Several previous studies on mental illness stigma have surveyed the general public but without assessing whether participants have a mental illness. Those studies that have asked whether par- ticipants had a mental illness have still included in the analysis participants who respond in the affirmative, using their mental illness as an indication of their familiarity with mental illness. This inclusion is problematic because the source of public stigma exists among people without mental illness. It is stig- matizers who stigmatize and who therefore are the target for stigma-reducing social marketing campaigns.
Thus, we focus this study only on people without a mental illness, excluding people with mental illness in the exploratory factor analysis as well as the cluster analysis. Therefore, our data set consisted of n = 373 participants without self-disclosed mental illness. Because the literature has demonstrated that many of the components of stigma are related, it is reasonable to assume that resulting factors may be correlated. Therefore, we used an oblique rotation to allow for correlated factors rather than to force an orthogonal solution, which likely
104 The Stigma of Mental Illness
would not reflect the relationship of the factors (Fabrigar et al. 1999; Ford, MacCallum, and Tait 1986). The factor analysis revealed six factors that accounted for 61.07% of the var- iance. Table 2 outlines the factor items, factor loading scores, and Cronbach’s alpha coefficients for each factor.
Factor 1 includes constructs that are known to be corre- lated, such as “dangerous,” “unpredictable/difficult,” “fear,” and “willingness to coerce.” Interestingly,“anger” alsoloads ontothis factor, in a departure from attribution theory, which proposes that anger is an emotional response to attributions about the cause and controllability of mental illness (Weiner, Perry, and Magnusson 1988). We label this factor “menace.” Factor 2 also encompasses two different theoretical constructs, “general concern for people with mental illness” and “treatment ef- ficacy.” This suggests that people without mental illness relate their willingness to help people with mental illness with their belief in effective treatment and likelihood of recovery. We label this factor “nurturance.” While the other four factors— “blame,” “shame,” “difference,” and “incompetence”—are consistent with stigma constructs discussed previously, the differences in the other two factors are interesting theoretically in that they suggest that within the minds of the general public, the constructs that are a part of the process of mental illness stigmatization may vary from what has been documented in the literature.
Because this research is exploratory in nature, with the goal of finding segments of stigmatizers, we utilized a two-step clus- ter analysis that removes much of the subjectivity in the de- cision process that may occur with other forms of cluster analysis (e.g., K-means). The two-step procedure determines the best cluster solution by calculating the Bayesian information criterion for each number of clusters specified and then finding the greatest change in distance between the two closest clusters in each hierarchical clustering stage (Punj and Stewart 1983; SPSS 2001). Using factor scores, a two-step cluster analysis classified respondents into five groups. The clusters identified represent distinct segments of the general public who may stigmatize those with mental illness for different reasons; they are thus heterogeneous between the groupings but homoge- neous within these groups. A multiple discriminant analysis further validated the accuracy of the five-cluster solution. Four statistically significant canonical discriminant functions were extracted, explaining a majority of the variance, with 97.1% of the original cases correctly classified by the discriminant functions.
Table 3 shows the mean scores calculated for each factor for each stigma segment and analysis of variance (ANOVA) results, which reveal significant differences among the five segments for each variable. We also conducted post hoc analysis for significant differences (see Table 3). In the ANOVA analysis, we include the people with mental illness as a distinct group in order to compare their responsesasa baseline because weexpect people with mental illness not to endorse the stigma dimensions of blame, menace, shame, difference, and incompetence but to endorse the stigma dimension of nurturance.
AscanbeseenfromTable3,caseswerenotdistributedequally across the segments. Segment 1 has relatively few cases, with only 12.6% of those participants without mental illness. We label this group “adversaries” due to their strong, negative feelings toward people with mental illness. Adversaries had the highest scores for menace endorsing, difference, and incompetence and
the lowest score on nurturance, making them the group that endorsed almost all of the stigma dimensions. Interestingly, despite their negative feelings and beliefs about people with mental illness, this group did not endorse blame.
We label segment 2 “blamers” because the blame dimension of stigma sets them apart from all of the other segments, in- cluding people with mental illness. While every other group clearly did not feel that people with mental illness were re- sponsible for their illness, all rating their feelings of blame below 3, participants in segment 2 felt that people with mental illness were somewhat to blame, with a mean rating of 5.09, more than two points higher than other groups. Blamers made up 20.9% of participants included in the segment analysis.
We name segment 3 “ambivalents.” Ambivalence is the state of having mixed feelings or conflicting thoughts on an issue. This group has both positive and negative endorsements of the stigma dimensions. Positively, this segment’s mean score for nurturance shows no differences from people with mental illness or from segments 4 and 5, all of whom demonstrate concern and positive thoughts regarding treatment. However, segment 3 has the second highest score in endorsing people with mental illness as different from themselves.
We label segment 4 “shamers”; this group has the highest mean score in the shame dimension. Shamers strongly believe that people with mental illness should not share their illness with others and would advise them to keep it to themselves. These beliefs have been used as a proxy for the amount of shame someone feels a person with mental illness should have. However, some variables used to measure this construct were adapted from items used with people with mental illness. It is possible that people in this segment feel that people with mental illness should not share their illness for protective reasons, rather than that they should be ashamed of their illness, because shamers also scored very high on nurturance.
Last, we label segment 5 “allies.” This group shows no differences in their endorsement of blame, nurturance, or incompetence from people with mental illness. Of the remaining three dimensions, on which allies did differ significantly from people with mental illness, their means were different in a di- rection that indicated that they hold more favorable thoughts and feelings toward people with mental illness than people with mental illness themselves do.
Demographic Profiles of Stigma Segments For the demographic profiles, only the segments for the stig- matizers are examined. Chi-square results indicate statistically significant differences between the five segments for gender (c2 = 20.40 [d.f. = 4, N = 373], p < .001) and marital status (c2 = 20.40 [d.f. = 4, N = 373], p < .001). The adversaries segment, the segment that is most negative toward people with mental illness, is majority male (63.9%). The blamers segment is also majority male (62.8%), as is the shamers segment (56.7%), although to a lesser degree. The allies segment, which dem- onstrates the most positive thoughts and feelings toward people with mental illness, is majority female (60%), as is the am- bivalents segment (63.7%).
The survey item on marital status asked whether participants were married, divorced, widowed, separated, or never married. Because the number of participants who reported being di- vorced, separated, or widowed was so low, we dichotomized
Journal of Public Policy & Marketing 105
T ab
le 3.
S eg m en t D es cr ip ti on
s
S ti gm
a D im
en si on
s
M ea n S co re s
F -V
al u e
P eo p le
w it h M en ta l
Il ln es sa (N
= 13
4) S eg m en t 1:
A d ve rs ar ie s
(N = 47
, 12
.6 %
) S eg m en t 2:
B la m er s
(N = 78
, 20
.9 %
) S eg m en t 3:
A m b iv al en ts
(N = 91
, 24
.4 %
) S eg m en t 4:
S h am
er s
(N = 67
, 18
% )
S eg m en t 5:
A ll ie s
(N = 90
, 24
.1 %
)
M en ac e
3 .4 0 (1 .3 4 )
5 .7 5 (1 .4 4 )
5 .2 7 (1 .0 4 )
4 .8 9 (1 .0 5 )
4 .9 2 (1 .1 1 )
2 .8 7
(. 8 5 )
7 8 .7 6
1 , 2 , 3 , 4 , 5
0 , 2
0 , 1 , 3 , 5
0 , 1 , 2 , 5
0 , 1 , 5
0 , 1 , 2 , 3 , 4
N u rt u ra n ce
6 .9 0 (1 .2 5 )
4 .5 3 (1 .3 2 )
5 .7 3
(. 9 7 )
6 .8 6 (1 .0 6 )
7 .1 1
(. 7 6 )
6 .9 5 (1 .0 6 )
5 0 .9 6
1 , 2
0 , 2 , 3 , 4 , 5
0 , 1 , 3 , 4 , 5
1 , 2
1 , 2
1 , 2
B la m e
2 .5 3 (1 .3 3 )
2 .6 9 (1 .3 3 )
5 .0 9 (1 .0 2 )
2 .4 2 (1 .0 1 )
2 .9 6 (1 .0 7 )
2 .3 4
(. 8 0 )
7 1 .0 3
2 , 4
2 0 , 1 , 3 , 4 , 5
2 , 4
0 , 2 , 3 , 5
2 , 4
S h am
e 3 .3 8 (2 .2 0 )
5 .3 2 (1 .9 4 )
4 .5 0 (1 .7 8 )
2 .0 9 (1 .1 1 )
6 .3 0 (1 .4 0 )
2 .2 1 (1 .3 5 )
7 3 .9 7
1 , 2 , 3 , 4 , 5
0 , 2 , 3 , 4 , 5
0 , 1 , 3 , 4 , 5
0 , 1 , 2 , 4
0 , 1 , 2 , 3 , 5
0 , 1 , 2 , 4
D if fe re n ce
3 .9 3 (1 .2 6 )
6 .7 2 (1 .3 0 )
5 .3 0 (1 .0 2 )
5 .4 2 (1 .3 2 )
4 .9 8 (1 .2 2 )
3 .5 6 (1 .1 9 )
6 2 .3 9
1 , 2 , 3 , 4 , 5
0 , 2 , 3 , 4 , 5
0 , 1 , 5
1 , 2 , 4 , 5
0 , 1 , 3 , 5
0 , 1 , 2 , 3 , 4
In co m p et en ce
3 .3 6 (1 .3 4 )
6 .1 3 (1 .2 4 )
4 .1 8
(. 9 7 )
4 .3 9 (1 .0 2 )
4 .0 0 (1 .1 4 )
3 .0 7 (1 .1 1 )
5 4 .2 4
1 , 2 , 3 , 4
0 , 2 , 3 , 4 , 5
0 , 1 , 5
0 , 1 , 4 , 5
0 , 1 , 3 , 5
1 , 2 , 3 , 4
a P eo p le
w it h m en ta l il ln es s ar e in cl ud ed
fo r co m p ar is o n p u rp o se s. P er ce n ta g es
re p o rt ed
ar e fo r th e to ta l n u m be r o f p eo p le
w it h o u t m en ta l il ln es s.
N o te s: M ea n sc o re s d es cr ib e ra ti n g s o n a 1 – 9 sc al e, w it h st an d ar d d ev ia ti o n s in
p ar en th es es . V al u es
0 – 5 in
th e se co n d ro w fo r ea ch
d im
en si o n in d ic at e si g n if ic an t d if fe re n ce s b et w ee n th e g iv en
se g m en t an d th e
re sp ec ti v e n u m be re d se g m en ts (w
it h th e g ro up
o f p eo p le
w it h m en ta l il ln es s d es ig n at ed
as 0 ). A ll n o te d d if fe re n ce s ar e p < .0 5 . S ig ni fi ca n ce
= 0 fo r al l d im
en si o n s.
106 The Stigma of Mental Illness
marital status into “ever married” and “never married” cate- gories. Marital status results show that blamers have a higher proportion of never having been married than any other group, while adversaries and shamers have equal numbers of people having been married and never married. Ambivalents and allies, on the other hand, have higher proportions of individuals who have been married than who have never married (see Figure 1).
A one-way ANOVA finds significant differences in age between clusters as well (F(5, 501) = 2.48, p < .05). Post hoc analysis finds that this difference is driven mainly by shamers (M = 41.03, SD = 14.63), who are significantly older than blamers (M = 34.42, SD = 12.04; p < .01) and allies (M = 35.29, SD = 10.47; p < .05) and marginally significantly different in age from ambivalents (M = 37.29, SD = 13.42; p = .07). Adversaries (M = 37.64, SD = 13.30; p < .01) are not sig- nificantly different in age from any other group.
To examine differences in familiarity between the segments, we conducted a chi-square analysis for each of the familiarity items (see Figure 2). We find marginally significant differences for familiarity in terms of having a close friend or family member with a mental illness (c2 = 9.05 [d.f. = 4, N = 373], p = .06), whereas we find significant differences for familiarity in terms of having ever worked with someone with mental illness (c2 = 11.19 [d.f. = 4, N = 373], p < .05). While having a friend or family member who has a mental illness is only marginally significant, the issue of familiarity has been shown to have a strong impact on the endorsement of stigmatizing beliefs. The allies segment is proportionally more likely to have a friend or family member with mental illness (63.3%) than any of the other segments, whereas the adversaries segment is the least likely (40.4%). Similarly, allies have the highest percentage of people who have ever worked with people with mental illness (40%). We find no differences among segments for race, ed- ucational status, or geographic location.
Comparison of Attitude In addition to the aforementioned segments, we add to our analysis people with mental illness as a group for compara- tive purposes. A one-way ANOVA analysis yields significant differences between groups in attitude toward people with mental illness (F(5, 501) = 33.98, p < .001). Post hoc analysis finds that people with mental illness have significantly more positive attitudes (M = 6.51, SD = 1.70) than any other seg- ment, with the exception of the allies. Interestingly, allies exhibit significantly more positive attitudes toward people with mental illness (M = 7.04, SD = 1.47) than even people with mental illness themselves do. Adversaries exhibit significantly more negative attitudes toward people with mental illness (M = 3.89, SD = 1.78) than the other segments. Interestingly, the three remaining segments, ambivalents (M = 5.29, SD = 1.43), blamers (M = 5.37, SD = 1.44), and shamers (M = 5.69, SD = 1.54), do not differ from each other in terms of attitude.
Comparison of Support Intentions With regard to support intentions, a one-way ANOVA finds significant differences among segments for both willingness to donate (F(5, 501) = 88.18, p < .001) and willingness to support (F(5, 501) = 122.60, p < .001). Post hoc analyses reveal sig- nificant differences among groups (for differences in support
intentions across segments, see Figure 3). Not surprisingly, adversaries have a significantly lower willingness to donate (M = 3.09, SD = 2.08) and support (M = 3.36, SD = 2.31) than any other group. People with mental illness (M = 6.05, SD = 2.44) show no significant difference from either allies (M = 6.22, SD = 2.50) or shamers (M = 5.52, SD = 2.28) for willingness to donate. People with mental illness (M = 7.04, SD = 2.22) are different from all other groups except allies (M = 7.01, SD = 2.14) for willingness to support. Consistent with other desired outcomes, shamers and ambivalents do not differ significantly from each other for either willingness to donate (shamers: M = 5.52, SD = 2.28; ambivalents: M = 5.03, SD = 2.61) or willingness to support (shamers: M = 5.88, SD = 2.27; ambivalents: M = 5.91, SD = 2.38). For willingness to donate, blamers (M = 4.51, SD = 2.06) also do not differ from am- bivalents (M = 5.03, SD = 2.61) but do show significant dif- ferences from the other groups. For willingness to support, blamers (M = 5.22, SD = 1.85) show significant differences from the other groups, with the exception of shamers (M = 5.88, SD = 2.27), from whom they show only a marginally significant difference (p = .07).
Comparison of Social Distance We conducted an exploratory factor analysis for the social distance items, and two factors emerge, accounting for 83.45% of the variance. Items loaded onto the first factor relate to work and friendship and living next door to someone with mental illness. This factor reflects “varied” relationships (a = .93) becausethese relationshipsdonotnecessarily involve closeness between two people (although they might). Four items loaded onto the second factor, which we label as “intimate” (a = .95) because three of the four items indicate having a more intimate, personal relationship, such as dating or marrying someone with a mental illness. Whether or not participants would share a house or be roommates with someone with mental illness also loaded onto this factor, which makes sense because living in the same household with someone makes a relationship, even a friendship, more intimate due to proximity.
A one-way MANOVA finds significant differences in social distance between segments (F(10, 1,000) = 34.82; Wilks’ L = .55, partial h2 = .93), with both varied (F(5, 501) = 60.27, p < .001) and intimate social distance (F(5, 501) = 38.93, p < .001) (for differences in social distance by cluster, see Figure 3). Post hoc analysis finds that along varied social distance, all segments are significantly different from each other and from people with mentalillness,withtheexceptionof ambivalents(M=6.44,SD= 1.82) and shamers (M = 6.30, SD = 1.75), who do not show any significant differences in varied social distance from each other. Along the intimate social distance there are fewer differences among segments. Allies are no different in their willingness to engage in an intimate relationship (M = 5.58, SD = 2.22) than people with mental illness (M = 5.98, SD = 2.50), although both groups are significantly different from the other three groups. Similar to the varied results, ambivalents (M = 3.41, SD = 2.06) and shamers (M = 3.60, SD = 2.19) show no significant dif- ferences from each other. Shamers are not significantly different from blamers (M = 3.99, SD = 2.07), and blamers show only marginally significant differences (p = .08) from ambivalents.
Consistent with attitudes, adversaries have the lowest social distance scores (Mvaried = 3.80, SD = 1.83; Mintimate = 1.85,
Journal of Public Policy & Marketing 107
Figure 1. Demographics by Segment
A: Gender
B: Ever Married
63.8% 62.8%
36.3%
56.7%
40.0% 36.2% 37.2%
63.7%
43.3%
60.0%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
Adversaries Blamers Ambivalents Shamers Allies
Male Female
51.1%
64.1%
41.8%
47.8% 44.4%
48.9%
35.9%
58.2%
52.2%
55.6%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
Adversaries Blamers Ambivalents Shamers Allies
Never married Ever married
108 The Stigma of Mental Illness
Figure 2. Responses to Familiarity Items by Segment
A: Have a Friend or Family Member with Mental Illness
B: Ever Worked with People with Mental Illness
59.6% 55.1%
46.2% 50.7%
36.7% 40.4%
44.9%
53.8%
49.3%
63.3%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
Adversaries Blamers Ambivalents Shamers Allies
No Yes
70.2%
83.3%
67.0% 70.1%
60.0%
29.8%
16.7%
33.0% 29.9%
40.0%
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
Adversaries Blamers Ambivalents Shamers Allies
No Yes
Journal of Public Policy & Marketing 109
Figure 3. Support Intentions and Social Distance by Segment
A: Support Intentions
B: Social Distance
0
1
2
3
4
5
6
7
8
9
6.05
3.09
4.51 5.03
5.52
6.22
7.04
3.36
5.22
5.91 5.88
7.01
Mental Illness
People with Adversaries Blamers Ambivalents Shamers Allies
Willingness to donate Willingness to support
7.56
3.8
5.36
6.44 6.3
8.06
5.99
1.85
3.99 3.41 3.6
5.58
0
1
2
3
4
5
6
7
8
9
People with Mental Illness
Adversaries Blamers Ambivalents Shamers Allies
Close Intimate
110 The Stigma of Mental Illness
SD = 1.03) indicating a definitive unwillingness to relate with people with mental illness, whereas allies have scores (Mvaried = 8.06, SD = .92; Mintimate = 5.58, SD = 2.22) as high as or higher than those of people with mental illness (Mvaried = 7.56, SD = 1.80; Mintimate = 5.99, SD = 2.50), demonstrating their will- ingness to interact. Ambivalents (M = 6.44, SD = 1.82) and shamers (M = 6.30, SD = 1.75) are also positive along the var- ied social distance dimension, indicating support for friend- ship, work relationships, and arm’s-length living arrangements. Blamers (M = 5.36, SD = 1.58) fall right at the midpoint of our scale. Interestingly, along intimate social distance, ambivalents (M = 3.41, SD = 2.06), blamers (M = 3.99, SD = 2.07), and shamers (M = 3.60, SD = 2.19) have decidedly lower scores, falling below the midpoint of the scale, indicating an unwill- ingness to engage in close, personal relationships with people with mental illness.
Discussion and Implications We demonstrate in our study that the overall group of people without mental illness is divided into distinctly different groups according to the dimensional factors that comprise the com- ponents that drive stigma. Segmentation of the overall market of people without mental illness finds a single group of people who do not stigmatize—allies, who consistently endorse positive beliefs and feelings toward people with mental illness—and four distinct segments of stigmatizers: adversaries, who con- sistently endorse negative beliefs and feelings; ambivalents, who have mixed beliefs and feelings; blamers, who endorse the belief that people with mental illness are responsible for their illness; and shamers, the only group of people who endorse our shame items. The differences we find between the groups in their endorsements of stigmatizing thoughts and feelings can inform disparate marketing action for targeted groups.
After segmentation, marketers decide who to target with marketing action. In a traditional marketing context, factors such as the segment’s profitability and its growth potential are considered. In social marketing, the value of a segment is measured in the social change desired. Which group would yield the most social change from social marketing dollars? The size of the segment may play a role in this decision. In our sample, ambivalents and allies comprise the largest segments, with each accounting for approximately 18% of the total sample. However, factorsother thansize are relevant to targeting decisions.
In that light, we start with the segment for which we believe stigma reduction efforts would create the least value: adver- saries. Adversaries are very negative on most of the stigma dimensions, with mean scores well below 3. They also have the lowest means for desirable outcomes, with mean scores ranging from 1.85 for intimate social distance to a high of 3.89 on attitude, indicating very negative feelings and intentions toward people with mental illness. Persuading adversaries to think differently about people with mental illness would be an uphill battle. Therefore, targeting adversaries for stigma reduction efforts is likely not worthy of resources.
However, adversaries may be the individuals who are at the greatest risk for engaging in outright discriminatory action, given their negative endorsement of the components of stigma. Although stigma reduction as a goal may not work with this segment, other messaging may be called for from a policy
perspective. Rather than attempting to make adversaries think more favorably of people with mental illness, it may be worth- while to emphasize to this group that there are consequences for discriminating on the basis of mental illness. The Americans with Disabilities Act and the Fair Housing Amendments Act protect people with mental disabilities from discrimination in em- ployment and housing. Policy and advocacy groups may find ways to target adversaries with information regarding the il- legality of discriminatory actions. They could also find ways in which these individuals, who might be considered high risk for exclusionary behavior, could be monitored or inspected. For example, identified adversaries in the workplace could be studied through exit interviews or surveys of exitingemployees.
Interestingly, adversaries are the smallest segment identified, accounting for only 9.27% of our total sample (12.6% of our sample when people with mental illness are not included). In our sample, allies outnumber adversaries almost 2–1; allies com- prise the second-largest segment (17.75% of the sample). Allies, combined with people with mental illness, make up over 44% of our total sample; thus, a significant percentage of our sam- ple scores very positively on almost every dimension of stigma. Targeting these groups is unnecessary for stigma reduction ef- forts. Just as stigma reduction among adversaries should not be a goal because it would be tough to move the needle for this segment, allies should not be a target market for stigma re- duction social marketing because they do not generally endorse stigmatizing beliefs and already possess positive views for de- sirable outcomes such as attitude and social distance.
However, allies may be targeted for other strategic goals to aid in stigma reduction efforts. Recall that allies’ scores on sup- port intentions indicate a willingness to donate and participate in support programs to fight stigma. Advocacy groups and policy makers could cultivate members of the allies segment as donors and volunteers for stigma reduction and other mental health initiatives. Campaigns and programming could also be aimed at allies to help them speak out on behalf of people with mental illness against prejudicial attitudes and discriminatory behavior. In the bullying literature, the acts of bystanders in- fluence perceptions of victims, bullies, and their perceived sense of safety in school (Gini et al. 2008). Similarly, allies could be encouraged to actively intervene when people express preju- dicial beliefs or demonstrate discriminatory behavior directly or indirectly toward people with mental illness. Even if they cannot successfully change others’ beliefs, their public show of support may be witnessed by a person with mental illness who may be hiding their illness, letting that person see that not all people hold stigmatizing beliefs. Likewise, a person who holds more stigmatizing beliefs may be positively influenced by an ally’s public display of support for people with mental illness. Another potential advocacy method could mirror the advocacy work in the LGBTQ community, which created SafeZone programs wherein allies of LGBTQ attend training and publicly identify their work area as a safe space. A similar intervention could be created for mental illness allies.
The blamers segment was named so due to their high en- dorsement of blame relative to the other segments. Blamers mean score for blame is 5.09, with all of the other groups reporting levels of blame under 3. Thus, it makes sense that policy and advocacy groups should target this group with messaging that attempts to persuade them that people with mental illness are not responsible for their illness, because no
Journal of Public Policy & Marketing 111
other group requires such convincing. However, research on programming that seeks to reduce this aspect of stigma has shown that reducing blame can have negative consequences on other stigma dimensions. For example, programs that used the biological explanations to reduce blame did not improve atti- tudes and provoked harsher behavioral intentions among those without mental illness (Angermeyer et al. 2011). This evidence guides our decision of whether to target this group and questions the wisdom of positioning the issue of mental illness around the issue of blame. Therefore, we conclude that blamers are not an appropriate target for social marketing campaign because tar- geting them may cause more harm than good.
This leaves two segments, ambivalents and shamers, as potential targets for social marketing action to reduce stigma. Interestingly, these two segments do not differ along desired outcomes, with no significant differences found in their will- ingness to donate or support, intimate relationship social dis- tance, or attitude. Both segments have particularly low scores for intimate relationship social distance reporting, at around 3.5. (An exception is varied relationships social distance, for which both segments score above 6, although this score is still more than 1.5 points lower than those of allies and people with mental illness.)
Despite their similarities on outcome measures, ambivalents and shamers do differ substantively on their endorsement of the six underlying stigma dimensions, with the exception of nur- turance. For nurturance, ambivalents and shamers show no significant differences. Shamers endorse blame (M = 2.96) and shame (M = 6.30) at significantly higher levels than ambivalents (M = 2.42 and M = 2.09, respectively). Shamers’ endorsement of blame is not high, and as was discussed earlier, the prob- lematic impact of tackling blame from past programming makes this dimension an unfavorable strategy for messaging. Shamers’ endorsement of shame, in contrast, is particularly significant because it is much higher than that of any other group, scoring fully a point higher than even adversaries. Thus, it makes sense totarget shamerswithmessaging tacklingthis stigma dimension.
Ambivalents, on the other hand, endorse difference (M = 5.42) and incompetence (M = 4.39) at higher levels than shamers (M = 4.98 and M = 4.00, respectively). Thus, mes- saging that may be more effective for ambivalents should position people with mental illness as not being different and as being competent. Larger-scale marketing campaigns that have used these strategies include a public service announcement by BringChange2Mind (2009). It shows several people who have mental illness wearing t-shirts on which their diagnosis is written; then, each person is joined by a loved one whose relationship role is shown. This campaign is intended to in- crease empathy among those without mental illness, who can relate to the loved ones portrayed. Another strategy used has been to highlight well-known, successful celebrities, historical figures, athletes, and others who have had mental illnesses to emphasize competence.
Our study illustrates how applying the marketing principle of segmentation and targeting among stigmatizers can inform the positioning of a serious public health issue, mental illness, for messaging in social marketing campaigns. It demonstrates that the general public can and should be further differentiated into groups because there is significant heterogeneity in their en- dorsement of stigmatizing thoughts and feelings about people with mental illness. Currently, public policy and advocacy groups differentiate target audiences very generally (e.g.,
SAMHSA’s “What a Difference a Friend Makes” campaign, targeting 18–24-year-olds). These campaigns also typically consult with people with mental illness and those who love or work with them to ensure that their voices are heard and their experiences authentically portrayed. Yet these are not the audiences who need changing. Our study shows that in-depth research can illuminate important differentiating factors that provide clarity for marketing action. These differences provide guidance for positioning the issue in terms of differential stig- matizing elements to target segments, which lead to disparate messaging and programming interventions to end stigma.
Interestingly, we find that demographic variables are not very informative about the segments. While gender differences are interesting, with adversaries and blamers more heavily male and ambivalents and allies more heavily female, we would not characterize any segment as having either gender predominate. Although age differences exist between the segments, the dif- ferences in age do not provide any additional insight into the targeting decision. Although shamers are generally older than the other segments, the ages within this group ranged from 20 to 65 years, and similarly broad ranges exist within all of the segments. Because targeting has been done according to age in existing campaigns, this finding supports our contention that more specific segmentation is needed to inform targeting and positioning strategy.
Policy and advocacy organizations may not conduct seg- mentation research according to psychographic profiles as we did here because of issues in implementing the targeting of these groups. That is, once segments are identified, how might organizations perform outreach to them with their social mar- keting campaigns? One major limitation of our study is that we did not collect information that would allow us to follow up with segments. We conducted the study anonymously to minimize socially desirable responding because the goal of the research was to demonstrate how segmentation could aid in the creation of more effective campaigns. But we certainly did not prove this contention because we did not follow through with these participants with social marketing communications/advertising. Policy and advocacy groups should, in practice, take care to collect data regarding these segments that might help them determine how best to reach each segment because psycho- graphic profiling makes the targeting of segments difficult. Additional survey questions could inform appropriate promo- tional channels through which to reach segments, such as ques- tions about media consumption habits and lifestyles.
Another limitation of the work is the replicability of the segments in the general public, an issue with all segmentation research. We found five distinctly different, interesting seg- ments among our sample that inform differential marketing action. More research is needed to determine whether these segments could be used to inform an actual campaign. How- ever, this study does demonstrate what should be done in any campaign making replication less of a concern. For example, SAMHSA’s “What a Difference a Friend Makes” campaign targeted 18–24-year-olds because of the vulnerability this population faces, both because of the uncertainty of their life circumstances and because mental illnesses begin very early in life (National Institute of Mental Health 2005). We are sug- gesting that this age group be further surveyed to distinguish different segments. For example, rather than a national sam- pling, further targeting could be done. Rather than assessing all
112 The Stigma of Mental Illness
18–24-year-olds in the United States, other segmentation cri- teria could be used first. For example, perhaps 18–24-year-olds who live in the states with the highest prevalence of mental illness, such as Arizona and Mississippi (Nguyen 2015), could be studied to discover specific segments within these geo- graphic areas. These findings could inform social marketing campaigns that position the issue of mental illness within the specific contexts of those states.
Regardless, this research illustrates the significance of ade- quate segmentation for determination of target markets for a significant public health problem, reducing the stigma that surrounds people with mental illness. Significant variation exists among segments regarding their endorsement of stigma thoughts and feelings, suggesting that different approaches would be calledfor insocialmarketingeffortstoreducestigma.Inadequate understanding of the market that requires change impedes the effectiveness of these campaigns, which ultimately impedes changing the attitudes of stigmatizers and perpetuates the suf- fering that stigma causes for people with mental illness. While input from those affected is necessary and absolutely essential in creating campaigns, it is not sufficient to affect change.
Because advocacy on many social issues involves people who are affected by the issue as advisors for change, inadequate segmentation, targeting, and positioning could apply to other issues as well. For example, policy and advocacy groups who deal with racial discrimination are often run by and have ad- visory boards consisting of people of color. This makes much sense; their outrage over their victimization is a necessary first step to gain traction on the issue, and their voices need to inform marketing actions to ensure their experience is authentically portrayed. However, campaigns to reduce racism must ulti- mately result in changing the attitudes, beliefs, and behaviors of those who perpetuate racism. If the climate of the 2016 U.S. presidential election around race, ethnicity, and immigration is any indication, success in changing the minds of people who possess truly racist beliefs has yet to be achieved. Thus, ef- fectively applying segmentation, targeting, and positioning among perpetrators of the problem is widely applicable to many serious and entrenched public health and social issues.
References Andreasen, Alan R. (1994), “Social Marketing: Its Definition and Do- main,” Journal of Public Policy & Marketing, 13 (Spring), 108–14.
Angermeyer, Matthias C., Michael Beck, and Herbert Matschinger (2003), “Determinants of the Public’s Preference for Social Distance from People with Schizophrenia,” Canadian Journal of Psychiatry, 48 (November), 663–68.
Angermeyer, Matthias C., and Sandra Dietrich (2006), “Public Beliefs About and Attitudes Towards People with Mental Illness: A Review of Population Studies,” Acta Psychiatrica Scandi- navica, 113 (3), 163–79.
Angermeyer, Matthias C., Anita Holzinger, Mauro G. Carta, and Georg Schomerus (2011), “Biogenetic Explanations and Public Acceptance of Mental Illness: Systematic Review of Population Studies,” British Journal of Psychiatry, 199 (November), 367–72.
Angermeyer, Matthias C., and Herbert Matschinger (1997), “Social Distance Towards the Mentally Ill: Results of Representative Surveys in the Federal Republic of Germany,” Psychological Medicine, 27 (1), 131–41.
Angermeyer, Matthias C., and Herbert Matschinger (2003a), “Public Beliefs About Schizophrenia and Depression: Similarities and Differences,” Social Psychiatry and Psychiatric Epidemi- ology, 38 (9), 526–34.
Angermeyer, Matthias C., and Herbert Matschinger (2003b), “The Stigma of Mental Illness: Effects of Labelling on Public Atti- tudes Towards People with Mental Disorder,” Acta Psychiatrica Scandinavica, 108 (4), 304–09.
Angermeyer, Matthias C., Aurélie Millier, Cécile Rémuzat, Tarek Refaı̈, and Mondher Toumi (2013), “Attitudes and Beliefs of the French Public About Schizophrenia and Major Depression: Results from a Vignette-Based Population Survey,” BMC Psy- chiatry, 13 (1), 1–12.
Anglin, Deidre M., Bruce G. Link, and Jo C. Phelan (2006), “Racial Differences in Stigmatizing Attitudes Toward People with Mental Illness,” Psychiatric Services, 57 (6), 857–62.
Black, Patrick M., Martin Duffy, John Kieran, James Mallon, and Bernadette M. Murphy (1993), “Attitudes Towards the Mentally Ill in Ireland,” Irish Journal of Psychological Medicine, 10 (2), 75–79.
Bogardus, Emory S. (1925), “Measuring Social Distance,” Journal of Applied Sociology, 9 (2), 299–308.
Borinstein, Andrew B. (1992), “Public Attitudes Toward Persons with Mental Illness,” Health Affairs (Project Hope), 11 (Fall), 186–96.
Bos, Arjan E.R., John B. Pryor, Glenn D. Reeder, and Sarah E. Stutterheim (2013), “Stigma: Advances in Theory and Research,” Basic and Applied Social Psychology, 35 (1), 1–9.
Boyd, Jennifer E., Emerald P. Katz, Bruce G. Link, and Jo C. Phelan (2010), “The Relationship of Multiple Aspects of Stigma and Personal Contact with Someone Hospitalized for Mental Illness, in a Nationally Representative Sample,” Social Psychiatry and Psychiatric Epidemiology, 45 (11), 1063–70.
BringChange2Mind (2009), “Grand Central Terminal :60,” YouTube video, https://www.youtube.com/watch?v=WUaXFlANojQ.
Clement, Sarah, Manuela Jarrett, Claire Henderson, and Graham Thornicroft (2010), “Messages to Use in Population-Level Campaigns to Reduce Mental Health-Related Stigma: Consen- sus Development Study,” Epidemiologia e Psichiatria Sociale, 19 (1), 72–79.
Cohen, Jack, and Elmer L. Struening (1962), “Opinions About Mental Illness in the Personnel of Two Large Mental Hospitals,” Journal of Abnormal and Social Psychology, 64 (5), 349–60.
Coleman, Daniel, Janet S. Walker, Junghee Lee, Barbara J. Friesen, and Peter N. Squire (2009), “Children’s Beliefs About Causes of Childhood Depression and ADHD: A Study of Stigmatization,” Psychiatric Services, 60 (7), 950–57.
Cooper, Amy E., Patrick W. Corrigan, and Amy C. Watson (2003), “Mental Illness Stigma and Care Seeking,” Journal of Nervous and Mental Disease, 191 (May), 339–41.
Corrigan, Patrick W. (2000), “Mental Health Stigma as Social Attribution: Implications for Research Methods and Attitude Change,” Clinical Psychology: Science and Practice, 7 (1), 48–67.
Corrigan, Patrick W. (2012), “Where Is the Evidence Supporting Public Service Announcements to Eliminate Mental Illness Stigma?” Psychiatric Services, 63 (January), 79–82.
Corrigan, Patrick W., Sachiko A. Kuwabara, and John O’Shaughnessy (2009), “The Public Stigma of Mental Illness and Drug Addiction,” Journal of Social Work, 9 (2), 139–47.
Journal of Public Policy & Marketing 113
Corrigan, Patrick, Fred E. Markowitz, Amy Watson, David Rowan, and Mary Ann Kubiak (2003), “An Attribution Model of Public Discrimination Towards Persons with Mental Illness,” Journal of Health and Social Behavior, 44 (2), 162–79.
Corrigan, Patrick W., and Jenessa R. Shapiro (2010), “Measuring the Impact of Programs That Challenge the Public Stigma of Mental Illness,” Clinical Psychology Review, 30 (8), 907–22.
Crisp, Arthur, Michael G. Gelder, Eileen Goddard, and Howard I. Meltzer (2005), “Stigmatisation of People with Mental Illnesses: A Follow-Up Study Within the Changing Minds Campaign of the Royal College of Psychiatrists,” World Psychiatry, 4 (2), 106–13.
Crisp, Arthur H., Michael G. Gelder, Susannah Rix, Howard I. Meltzer, and Olwen J. Rowlands (2000), “Stigmatisation of People with Mental Illnesses,” British Journal of Psychiatry, 177 (1), 4–7.
Devine, Patricia G., E.A. Plant, and Kristen Harrison (1999), “The Problem of ‘Us’ Versus ‘Them’ and AIDS Stigma,” American Behavioral Scientist, 42 (7), 1212–28.
Durand, Richard M., David L. Klemmack, and Lucinda Lee Roff (1982), “Citizen Preferences for the Allocation of Federal Tax Dollars: A Segmentation Approach,” Journal of Marketing & Public Policy, 1 (Spring), 169–80.
Eisenberg, Daniel, Marilyn F. Downs, Ezra Golberstein, and Kara Zivin (2009), “Stigma and Help-Seeking for Mental Health Among College Students,” Medical Care Research and Review, 66 (5), 522–41.
Evans-Lacko, Sara, Claire Henderson, and Graham Thornicroft (2013), “Public Knowledge, Attitudes, and Behaviour Regarding People with Mental Illness in England 2009–2012,” British Journal of Psychiatry, 202 (S55), S51–57.
Evans-Lacko, Sara, Estelle Malcolm, Keon West, Diana Rose, Jillian London, Nicolas Rüsch, et al. (2013), “Influence of Time to Change’s Social Marketing Interventions on Stigma in England 2009–2011,” British Journal of Psychiatry, 202 (S55), S77–88.
Fabrigar, Leandre R., Duane T. Wegener, Robert C. MacCallum, and Erin J. Strahan (1999), “Evaluating the Use of Exploratory Factor Analysis in Psychological Research,” Psychological Methods, 4 (3), 272–99.
Ford, Kevin J., Robert MacCallum, and Marianne Tait (1986), “The Application of Exploratory Factor Analysis in Applied Psy- chology: A Critical Review and Analysis,” Personnel Psychol- ogy, 39 (2), 291–314.
Gaebel, Wolfgang, Harald Zäske, Anja E. Baumann, Joachim Klosterkötter, Wolfgang Maier, Petra Decker, et al. (2008), “Evaluation of the German WPA ‘Program Against Stigma and Discrimination Because of Schizophrenia—Open the Doors’: Results from Representative Telephone Surveys Before and After Three Years of Antistigma Interventions,” Schizophrenia Re- search, 98 (1), 184–93.
Gini, Gianluca, Tiziana Pozzole, Francesco Borghi, and Lara Franzoni (2008), “The Role of Bystanders in Students’ Perception of Bullying and Sense of Safety,” Journal of School Psychology, 46 (6), 617–38.
Goffman, Erving (1963), Stigma: Notes on the Management of Spoiled Identity. New York: Simon & Schuster.
Griffiths, Kathleen M., Yoshibumi Nakane, Helen Christensen, Kumiko Yoshioka, Anthony F. Jorm, and Hideyuki Nakane (2006), “Stigma in Response to Mental Disorders: A Comparison of Australia and Japan,” BMC Psychiatry, 6 (1), 1–10.
Gross, James J., and Ricardo F. Muñoz (1995), “Emotion Regu- lation and Mental Health,” Clinical Psychology: Science and Practice, 2 (2), 151–64.
Hatzenbuehler, Mark L., Jo C. Phelan, and Bruce G. Link (2013), “Stigma as a Fundamental Cause of Population Health In- equalities,” American Journal of Public Health, 103 (5), 813–21.
Hinshaw, Stephen P., and Andrea Stier (2008), “Stigma as Related to Mental Disorders,” Annual Review of Clinical Psychology, 4 (April), 367–93.
Horton, Richard (2007), “Launching a New Movement for Mental Health,” Lancet, 370 (9590), 806.
Hugo, Charmaine J., Dorothy E.L. Boshoff, Annelene Traut, Nompumelelo Zungu-Dirwayi,and DanJ.Stein(2003),“Community Attitudes Toward and Knowledge of Mental Illness in South Africa,” Social Psychiatry and Psychiatric Epidemiology, 38 (12), 715–19.
Jones, Edward E., Amerigo Farina, Albert H. Hastorf, Hazel Markus, Dale T. Miller, and Robert A. Scott (1984), Social Stigma: The Psychology of Marked Relationships. New York: W.H. Freeman & Company.
Jorm, Anthony F., and Kathleen M. Griffiths (2008), “The Public’s Stigmatizing Attitudes Towards People with Mental Disorders: How Important Are Biomedical Conceptualizations?” Acta Psychiatrica Scandinavica, 118 (4), 315–21.
Jorm, Anthony F., Ailsa E. Korten, Patricia Jacomb, Helen Christensen, and Scott Henderson (1999), “Attitudes Towards People with a Mental Disorder: A Survey of the Australian Public and Health Professionals,” Australian and New Zealand Journal of Psychiatry, 33 (1), 77–83.
Kermode, Michelle, Kathryn J. Bowen, Shobha Arole, Soumitra Pathare, and Anthony F. Jorm (2009), “Attitudes to People with Mental Disorders: A Mental Health Literacy Survey in a Rural Area of Maharashtra, India,” Social Psychiatry and Psychiatric Epidemiology, 44 (12), 1087–96.
Kessler, Ronald C., Steven Heeringa, Matthew D. Lakoma, Maria Petukhova, Agnes E. Rupp, Michael Schoenbaum, et al. (2008), “Individual and Societal Effects of Mental Disorders on Earnings in the United States: Results from the National Comorbidity Survey Replication,” American Journal of Psychiatry, 165 (June), 703–11.
Kobau, Rosemarie, Colleen DiIorio, Daniel Chapman, and Paolo Delvecchio (2010), “Attitudes About Mental Illness and Its Treatment: Validation of a Generic Scale for Public Health Sur- veillance of Mental Illness Associated Stigma,” Community Men- tal Health Journal, 46 (2), 164–76.
Link, Bruce G., and Jo C. Phelan (2001), “Conceptualizing Stigma,” Annual Review of Sociology, 27, 363–85.
Link, Bruce G., Elmer L. Struening, Michael Rahav, Jo C. Phelan, and Larry Nuttbrock (1997), “On Stigma and Its Consequences: Evidence from a Longitudinal Study of Men with Dual Diagnoses of Mental Illness and Substance Abuse,” Journal of Health and Substance Abuse, 38 (2), 177–90.
Link, Bruce G., Lawrence H. Yang, Jo C. Phelan, and Pamela Y. Collins (2004), “Measuring Mental Illness Stigma,” Schizophrenia Bulletin, 30 (3), 511–41.
Martin, Jack K., Bernice A. Pescosolido, Sigrun Olafsdottir, and Jane D. McLeod (2007), “The Construction of Fear: Americans’ Preferences for Social Distance from Children and Adolescents with Mental Health Problems,” Journal of Health and Social Behavior, 48 (1), 50–67.
Martin, Jack K., Bernice A. Pescosolido, and Steven A. Tuch (2000), “Of Fear and Loathing: The Role of ‘Disturbing Be- havior,’ Labels, and Causal Attributions in Shaping Public At- titudes Toward People with Mental Illness,” Journal of Health and Social Behavior, 41 (2), 208–23.
114 The Stigma of Mental Illness
McGlashan, Thomas H. (1999), “Duration of Untreated Psychosis in First-Episode Schizophrenia: Marker or Determinant of Course?” Biological Psychiatry, 46 (October), 899–907.
Mehta, Nisha, Aliya Kassam, Morven Leese, Georgia Butler, and Graham Thornicroft (2009), “Public Attitudes Towards People with Mental Illness in England and Scotland, 1994–2003,” British Journal of Psychiatry, 194 (3), 278–84.
Meltzer, Herbert Y., Larry Alphs, Alfredo Carlo Altamura, Al Green, Ravi Anand, Alberto Bertoldi, et al. (2003), “Clozapine Treatment for Suicidality in Schizophrenia,” Archives of General Psychiatry, 60 (1), 82–91.
Mulligan, Kate (2005), “Federal Agency Escalates Battle Against Stigma,” Psychiatric News, Vol. 40 (November 4), http://psychnews. psychiatryonline.org/doi/10.1176/pn.40.21.0009.
Murphy, M.K., N.A. Black, D.L. Lamping, C.M. McKee, C.F. Sanderson, J. Askham, and T. Marteau (1998), “Consensus Development Methods, and Their Use in Clinical Guideline Development,” Health Technology Assessment, 2 (3).
National Institute of Mental Health (2005), “Mental Illness Exacts Heavy Toll, Beginning in Youth,” press release (June 6), https:// www.eurekalert.org/pub_releases/2005-06/niom-mie060105.php.
National Institute of Mental Health (2008), “The Numbers Count: Mental Disorders in America,” http://www.lb7.uscourts.gov/ documents/12-cv-1072url2.pdf.
Ng, Sonya L., Judith L. Martin, and Sarah E. Romans (1995), “A Community’s Attitudes Towards the Mentally Ill,” New Zealand Medical Journal, 108 (1013), 505–08.
Nguyen, Theresa (2015), “Parity or Disparity: The State of Mental Health in America 2015,” report, Mental Health America, https:// www.mentalhealthamerica.net/sites/default/files/Parity%20or% 20Disparity%202015%20Report.pdf.
Norman, Ross M., Richard M. Sorrentino, Deborah Windell, and Rahul Manchanda (2008), “The Role of Perceived Norms in the Stigmatization of Mental Illness,” Social Psychiatry and Psy- chiatric Epidemiology, 43 (11), 851–59.
Ojanen, Markku (1992), “Attitudes Towards Mental Patients,” International Journal of Social Psychiatry, 38 (2), 120–30.
Parcesepe, Angela M., and Leopoldo J. Cabassa (2013), “Public Stigma of Mental Illness in the United States: A Systematic Literature Review,” Administration and Policy in Mental Health, 40 (5), 384–99.
Perry, Brea L., Bernice A. Pescosolido, Jack K. Martin, Jane D. McLeod, and Peter S. Jensen (2007), “Comparison of Public At- tributions, Attitudes, and Stigma in Regard to Depression Among Children and Adults,” Psychiatric Services, 58 (5), 632–35.
Pescosolido, Bernice A., Danielle L. Fettes, Jack K. Martin, John Monahan, and Jane D. McLeod (2007), “Perceived Dangerous- ness of Children with Mental Health Problems and Support for Coerced Treatment,” Psychiatric Services, 58 (5), 619–25.
Pescosolido, Bernice A., Jack K. Martin, Bruce G. Link, Saeko Kikuzawa, Giovani Burgos, Ralph Swindle, et al. (1996), Americans’ View of Mental Health and Illness at Century’s End: Continuity and Change. Bloomington: Indiana University.
Pescosolido, Bernice A., Jack K. Martin, J. Scott Long, Tait R. Medina, Jo C. Phelan, and Bruce G. Link (2010), “‘A Disease Like Any Other?’ A Decade of Change in Public Reactions to Schizophrenia, Depression, and Alcohol Dependence,” American Journal of Psychiatry, 167 (11), 1321–30.
Pescosolido, Bernice A., Tait R. Medina, Jack K. Martin, and J. Scott Long (2013), “The ‘Backbone’ of Stigma: Identifying the
Global Core of Public Prejudice Associated with Mental Illness,” American Journal of Public Health, 103 (5), 853–60.
Pescosolido, Bernice A., John Monahan, Bruce G. Link, Ann Stueve, and Saeko Kikuzawa (1999), “The Public’s View of the Competence, Dangerousness, and Need for Legal Coercion of Persons with Mental Health Problems,” American Journal of Public Health, 89 (9), 1339–45.
Phelan, Julie E., and Susan A. Basow (2007), “College Students’ Attitudes Toward Mental Illness: An Examination of the Stigma Process,” Journal of Applied Social Psychology, 37 (12), 2877–902.
Pilling, Bruce K., Lawrence A. Crosby, and Pam Scholder Ellen (1991), “Using Benefit Segmentation to Influence Environmental Legislation: A Bottle Bill Application,” Journal of Public Policy & Marketing, 10 (Fall), 28–46.
Punj, Girish, and David W. Stewart (1983), “Cluster Analysis in Marketing Research: Review and Suggestions for Application,” Journal of Marketing Research, 20 (May), 134–48.
Rice, Dorothy P., Sander Kelman, and Leonard S. Miller (1991), “Estimates of Economic Costs of Alcohol and Drug Abuse and Mental Illness, 1985 and 1988,” Public Health Reports, 106 (May/June), 280–92.
Rüsch, Nicholas, Patrick Corrigan, Abigail Wassel, Patrick Michaels, M. Olschewski, S. Wilkniss, et al. (2009), “Ingroup Perception and Responses to Stigma Among Persons with Mental Illness,” Acta Psychiatrica Scandinavica, 120 (4), 320–28.
SAMHSA (2010), “Dude Got Back | What a Difference a Friend Makes PSA,” YouTube video, https://www.youtube.com/watch? v=XNAEdcAY-bw.
Silton, Nava R., Kevin J. Flannelly, Glen Milstein, and Margaret L. Vaaler (2011), “Stigma in America: Has Anything Changed? Impact of Perceptions of Mental Illness and Dangerousness on the Desire for Social Distance: 1996 and 2006,” Journal of Nervous and Mental Disease, 199 (6), 361–66.
SPSS (2001), “The SPSS TwoStep Cluster Component,” technical report, http://www.spss.ch/upload/1122644952_The%20SPSS% 20TwoStep%20Cluster%20Component.pdf.
Star, Shirley A. (1952), “What the Public Thinks About Mental Health and Mental Illness,” paper presented at the annual meeting of the National Association for Mental Health (November 19), http://www. norc.org/PDFs/publications/StarS_What_Public_Thinks_1952.pdf.
Star, Shirley A. (1955), “The Public’s Ideas about Mental Illness,” paper presented at the annual meeting of the National Association for Mental Health, Indianapolis (November 5), http://www.norc. org/PDFs/publications/StarS_Publics_Ideas_1955.pdf.
Suicide Prevention Resource Center (2007), “What a Difference a Friend Makes,” (accessed April 12, 2016), http://www.sprc.org/ library_resources/items/what-difference-friend-makes-national- mental-health-anti-stigma-campaign.
Taylor, S. Martin, and Michael J. Dear (1981), “Scaling Community Attitudes Toward the Mentally Ill,” Schizophrenia Bulletin, 7 (2), 225–40.
U.S. Department of Health and Human Services (1999), “Mental Health: A Report of the Surgeon General,” https://profiles.nlm. nih.gov/ps/access/NNBBHS.pdf.
Van ’t Veer, Job T., Herro F. Kraan, Stans H.C. Drosseart, and Jacqueline M. Modde (2006), “Determinants That Shape Public Attitudes Towards the Mentally Ill,” Social Psychiatry and Psychiatric Epidemiology, 41 (4), 310–17.
Journal of Public Policy & Marketing 115
Vaughan, Gerard, and Chris Hansen (2004), “‘Like Minds, Like Mine’: A New Zealand Project to Counter the Stigma and Dis- crimination Associated with Mental Illness,” Australasian Psy- chiatry, 12 (2), 113–17.
Wahl, Otto F. (1992), “Mass Media Images of Mental Illness: A Review of the Literature,” Journal of Community Psychology, 20 (4), 343–52.
Walker, Janet S., Daniel Coleman, Junghee Lee, Peter Noonan Squire, and Barbara J. Friesen (2008), “Children’s Stigmatization of Childhood Depression and ADHD: Magnitude and De- mographic Variation in a National Sample,” Journal of the American Academy of Child and Adolescent Psychiatry, 47 (8), 912–20.
Weiner, Bernard, Raymond P. Perry, and Jamie Magnusson (1988), “An Attributional Analysis of Reactions to Stigmas,” Journal of Personality and Social Psychology, 55 (November), 738–48.
Wolff, Geoffrey, Soumitra Pathare, Tom Craig, and Julian Leff (1996), “Community Attitudes to Mental Illness,” British Journal of Psychiatry, 168 (2), 183–90.
World Health Organization (2013), “The European Mental Health Action Plan 2013–2020,” http://www.euro.who.int/__data/assets/ pdf_file/0020/280604/WHO-Europe-Mental-Health-Acion-Plan- 2013-2020.pdf.
Yeh, Marie A., and Robert D. Jewell (2015), “The Myth/Fact Message Frame and Persuasion in Advertising: Enhancing At- titudes Toward the Mentally Ill,” Journal of Advertising, 44 (2), 161–72.
Yeh, Marie A., Robert D. Jewell, and Michael Y. Hu (2013), “Stereotype Processing’s Effect on the Impact of the Myth/Fact Message Format and the Role of Personal Relevance,” Psy- chology and Marketing, 30 (January), 36–45.
116 The Stigma of Mental Illness
Copyright of Journal of Public Policy & Marketing is the property of American Marketing Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.