Estimating the Relationship Between Perceived Stigma and Victimization of People With Mental Illness

Abstract

Over the past two decades, we have substantially increased our understanding of violence committed by individuals with mental illness, while comparatively less is known about the victimization experiences of this population. What has been established in the literature is that individuals with mental illness are more likely to experience victimization than the general public, and certain risk factors influence the likelihood of victimization. What remains unexplored is the possibility that a person with mental illness’ perception that mental illness is stigmatized may be significantly associated with victimization experiences. Thus, the purpose of the current study is to examine whether stigma and victimization are associated, and in what direction. In other words, does perceived stigma lead to victimization? Or does victimization lead to perceived stigma? To assess these research questions, data from the Community Outcomes of Assisted Outpatient Treatment study are used, which is a longitudinal study of individuals with serious mental illness (n = 184). A variety of methods are employed to assess the association between victimization and perceived stigma including logistic and ordinary least squares regression models. Results from the logistic regression model indicate that perceived stigma is associated with an increase in the odds that a person with mental illness will experience victimization at later follow-ups. Results from the ordinary least squares regression analysis, however, show that victimization at baseline does not predict perceived stigma at later times. Implications regarding future research and clinical practice are discussed.

Keywords

perceived stigma mental health victimization violence exposure longitudinal data

Introduction

Considerable research attention has focused on violence committed by individuals with mental illness (see Elbogen & Johnson, 2009; Estroff et al., 1994; Hiday, 1997; Mulvey, 1994; Silver, 2006; Swartz et al., 1998), whereas victimization experiences among this population have received much less attention (see Khalifeh et al., 2016; Monahan et al., 2017). In response to this imbalance, a handful of investigators have called for increased attention to victimization, and have begun to explore this important domain (e.g., Hiday et al., 1999, 2002; Silver, 2002, 2005; Teasdale, 2009; Teasdale et al., 2014). What is known at this point is that individuals with mental illness are at increased risk of victimization experiences when compared with the general population (Goodman et al., 2001; Hiday et al., 1999, 2002; Silver, 2002; Teplin et al., 2005; Walsh et al., 2003). In fact, some studies have suggested that people with mental disorders experience from two (Hiday et al., 1999; Walsh et al., 2003) to four (Teplin et al., 2005) times the risk of violent victimization when compared with the general population.

In addition, people with mental disorders are subject to stigma and its consequences. Common conceptualizations of stigma include an, “attribute that is deeply discrediting” (Goffman, 1963, p. 3), a characteristic of persons that is contrary to societal norms (Stafford & Scott, 1986), or a characteristic that is devalued in the social context (Crocker, 1999). Unfortunately, people with mental illness, who sometimes display bizarre behaviors or fail to fulfill traditional societal expectations, have been extensively stereotyped and stigmatized (Link et al., 1999). In fact, people with mental illness are often portrayed in media outlets as ineffective in fulfilling societal roles, as well as being threats to community safety (Gerbner et al., 1981; Myrick & Pavelko, 2017; Owen, 2012).

Given that people with mental illness are at increased risk for victimization, and are subjected to stigma, a question that arises is whether there is a relationship between the two phenomena. Is it possible that stigma is significantly associated with the victimization experiences of people with mental illness? As Link and Phelan (2001) argue, cultural stereotypes play an important role in understanding stigma. The cultural stereotype of mental illness may lead the general public to believe that individuals with mental disorders are dangerous (Link et al., 1999). As a consequence of this perception, members of the public may desire increased social distance from individuals with mental illness (Link et al., 1999). This strong stereotype, or stigma, of people with mental illness as dangerous and violent (Pescosolido et al., 1999) may contribute to their victimization experiences. That is to say, as theorized by other scholars, it is possible that people with mental disorders may be victimized because of the stigma associated with having a mental illness (Teasdale, 2009). If this stereotype triggers fear and anxiety, individuals in the general public may preemptively defend themselves against people they perceive as dangerous, thus victimizing them, purportedly in self-defense. Alternatively, this stigma may drive away potential guardians, leaving stigmatized individuals vulnerable to victimization.

To assess empirically the role of stigma in the experience of victimization among people with mental disorders, data from the Community Outcomes of Assisted Outpatient Treatment (AOT), a longitudinal study of individuals with serious mental illness (Link et al., 2011), will be utilized. The primary goal of our study is to determine whether stigma is significantly associated with victimization experiences among people with mental illness.

Literature Review

Violent Victimization of Individuals With Mental Disorders

Despite the stereotypes of dangerousness associated with mental disorder, individuals with mental illness are actually more likely to be victims of violence than its perpetrators (Choe et al., 2008; Latalova et al., 2014; Maniglio, 2009). Although prevalence rates for violent victimization vary by type of victimization assessed, study population, and study design (Choe et al., 2008; Latalova et al., 2014), researchers have consistently reported prevalence ratios several times higher than the general population (e.g., Goodman et al., 1999; Maniglio, 2009). For example, one study comparing violent victimization rates from a subset of the National Crime Victimization Survey (NCVS) with a sample of people receiving treatment for serious mental illness found that roughly 25% of the patient sample had experienced violent victimization over the prior 6 months—a prevalence rate 11.8 times higher than that found in the NCVS comparison data (Teplin et al., 2005). The study’s reported prevalence ratios did vary by type of violent victimization; however, the patient sample experienced higher rates of victimization across all categories and subcategories of violence, including rapes and sexual assaults, robberies, and physical assaults. Recent examinations of data from the MacArthur Violence Risk Assessment Study have also found high rates of violent victimization (Monahan et al., 2017; Teasdale, 2009). In fact, researchers have reported a 43% past-year prevalence of violent victimization among this population (Monahan et al., 2017). Thus, across a range of samples and contexts, individuals with serious mental illness are significantly more likely than others to experience violent victimizations.

What still remains to be explained is why these victimizations occur. Several recent studies have attempted to fill this gap by examining the role clinical factors play in this increased victimization risk (Latalova et al., 2014; Maniglio, 2009). These studies have identified comorbid alcohol and substance use problems (Hiday et al., 1999; Langeveld et al., 2018; Marley & Buila, 2001) as correlates of victimization among people with severe mental illness. Some research has also found that younger age at first psychiatric hospitalization and a recent history of hospitalization substantially increase risk for violent victimization among both men and women (e.g., Goodman et al., 2001). Taken together, these findings suggest that certain factors, specifically those that are associated with increased disorder-related symptomatology, may play a role in increasing risk for violent victimization. This may be especially true in instances where disorder-related behavior is poorly understood by bystanders or is interpreted as cues of dangerousness (Pescosolido et al., 1999).

Researchers have also explored whether more general risk factors are related to violent victimization among people with mental illness. Recent homelessness has been implicated as a risk factor for violent victimization both in the general population (e.g., Lee & Schreck, 2005) and among this population (Crisanti et al., 2014; Goodman et al., 2001; Hiday et al., 1999; Maniglio, 2009; Teasdale, 2009). This is unsurprising, as homelessness reduces guardianship (i.e., the absence of a capable person or entity who can prevent criminal offenses against a person or property; see Cohen & Felson, 1979; Fitzpatrick et al., 1993) and is strongly correlated with other risk factors for victimization, such as substance use (e.g., Lee & Schreck, 2005). Researchers have also consistently demonstrated the significance of prior crime and violence perpetration as a significant risk factor for victimization among disordered populations (Crisanti et al., 2014; de Waal et al., 2018; Langeveld et al., 2018; Latalova et al., 2014; Maniglio, 2009; Policastro et al., 2016; Teasdale, 2009). As some research suggests, this association may be a function of conflicted social relationships with others (Silver, 2002).

Although this literature has identified a number of risk factors associated with violent victimization among individuals with serious mental illness, which may be necessary to control for in subsequent analyses of violent victimization, it is still unclear whether and in what ways stigma affects victimization, net of these other factors. We now turn to an examination of stigma and its potential for understanding victimization.

Stigma, Mental Illness, and Negative Consequences

Stigma has been conceptualized in numerous ways including a discrediting attribute (Goffman, 1963) or elements consisting of labeling or stereotyping (Link & Phelan, 2001). As stated previously, scholars have argued that cultural stereotypes play an important role (Link & Phelan, 2001), especially in the context of people with mental illness. For example, socialization within our culture creates a set of beliefs about mental illness early in life (Link et al., 1989). These beliefs, in turn, help formulate individuals’ conceptions about what it means to have a mental illness and the behaviors associated with people suffering from mental illnesses (Link et al., 1989; Link & Phelan, 2001). Unfortunately, a popular cultural stereotype of people with mental illness may include failing to fulfill societal obligations (Gerbner et al., 1981) or being viewed as a danger to others, violent, or frightening (Phelan et al., 2000). Therefore, people may avoid individuals with mental illness in the context of employment, neighborhoods, intimate partner relationships, and so on (Link & Phelan, 2001). This avoidance may influence people with mental illness to believe that they will be devalued or discriminated against, thereby affecting their interactions with others (Link et al., 1989).

Feeling devalued or discriminated against can have negative consequences for people with mental illness. In fact, several negative consequences of this process have been identified in literature. For instance, Link and colleagues (1989) found that patients who perceived that they were devalued due to their status or stigma as a patient had constricted social networks. In addition, scholars have also found that stigma negatively affects the self-esteem of a person with mental illness (Link et al., 2001; Livingston & Boyd, 2010; Verhaeghe et al., 2008). Researchers have also found that stigma negatively affects the well-being and help-seeking behaviors of a person with mental illness (Cruwys & Gunaseelan, 2016; Link et al., 1997; Reynders et al., 2014) and can contribute to suicidality (Rüsch et al., 2014).

Given that stigma increases negative perceptions of one’s self and negative interactions with others, it is possible that stigma leads to disadvantaged circumstances in general. That is, stigma may result in a person with mental illness living in a disadvantaged neighborhood, lacking gainful employment, and lacking meaningful relationships with others. Although speculative, these disadvantaged circumstances caused by stigma may influence another negative outcome: victimization among people with mental illness.

Stigma, Mental Illness, and Victimization

We speculate that stigma and victimization may correlate for several reasons. First, it is possible that the stigma of mental illness may influence a lack of capable guardianship (see Cohen & Felson, 1979). Research has shown that the general public desires social distance from people with mental illness because of the belief that this population is dangerous (Link et al., 1999) or violent (Pescosolido et al., 1999). Because of this perception, people with mental illness may have constricted social networks (Link et al., 1989) or may be involved in conflicted social relationships (Silver, 2002). If people with mental illness have constricted social networks and lack quality relationships because of the stigma of having a mental illness, it is possible that this population also lacks capable guardians who could prevent a victimization experience.

Second, the stigma of mental illness may lead to reduced well-being (Link et al., 1997) and lower self-esteem (Verhaeghe et al., 2008), which, in turn, may lead to the involvement in risky situations. It is plausible that people with mental illness may have maladaptive coping strategies such as substance abuse (Ryan et al., 2014) that may lead this population to find themselves in situations that are conducive to victimization experiences.

Alternatively, it is also possible that stigma associated with the experience of mental illness symptoms could increase offender motivation. That is, if a person with mental illness acts bizarrely or triggers fear or anxiety in people they encounter, it is possible that individuals may preemptively defend themselves against a person perceived to be a dangerous person, ultimately victimizing them to prevent their own hypothetical, expected victimization. Osgood and colleagues (1996) argue that motivation resides not in the individual rather within the situation. It is possible that situational encounters may create offender motivation (in this case, self-protection) in individuals who hold stigmatizing beliefs about people who suffer from mental illnesses.

Finally, it is also possible that the stigma of mental illness may lead to disadvantaged situations that are conducive to victimization experiences. As mentioned previously, people may avoid individuals with mental illness in the context of employment, neighborhoods, or intimate partner relationships (Link & Phelan, 2001). Given that stigma may contribute, in part, to where a person with mental illness may live, it is possible that stigma may lead to this population residing or working in socially disorganized neighborhoods. In fact, Silver (2000) found that compared with the general population, patients discharged from a public psychiatric hospital resided in more disadvantaged and less safe neighborhoods.

Current Study

Although it is possible that stigma may contribute to victimization experiences through any or all the mechanisms just identified at this time a lynch-pin finding remains unexplored—we do not know whether indicators of stigma are associated with victimization. Thus, the purpose of the current study is to examine if stigma and victimization are associated, and in what direction. In other words, does perceived stigma lead to later victimization? Or does victimization lead to later perceived stigma? To investigate these questions, we use the Community Outcomes of AOT data. There are several benefits to using the AOT data. For instance, given the longitudinal structure of the data, we are able to assess victimization at follow-up waves, while controlling for stigma and other control variables at baseline to establish temporal order. In addition, because AOT gathered data on a multitude of clinical and general risk factors, we are able to assess the relationship between stigma and victimization in the context of known risk factors for victimization for this population. An additional benefit of the longitudinal data collection is that we can assess the possibility of reciprocal causation. That is, although we have theorized that stigma could lead to victimization, it is also possible that victimization leads to a person feeling more stigmatized. Given that multiple waves of data were collected from each participant and that both stigma and victimization were assessed at each wave, we can estimate models to see if stigma at an earlier wave predicts victimization at a later wave (holding constant victimization in the earlier wave) or if victimization in an earlier wave predicts stigma in a later wave (holding constant stigma in the earlier wave).

Method

Sampling

The Community Outcomes of AOT is a longitudinal study of 184 people with serious mental illness (Link et al., 2011; Phelan et al., 2010). Participants were selected from treatment facilities in New York and were between the ages of 18 and 65 years (Phelan et al., 2010). Two groups were utilized in the study, including a court-ordered assisted outpatient group (n = 76) and an outpatient comparison group of individuals recently discharged from a psychiatric facility (n = 108; Phelan et al., 2010). Once informed consent was obtained, interviews were conducted every 3 months for a year (Phelan et al., 2010). Approximately 40% of the sample was diagnosed with schizophrenia spectrum disorders, 32% was diagnosed with schizo-affective disorders, 7% was diagnosed with major depressive disorder, 19% was diagnosed with bipolar disorder, and 2% was diagnosed with substance-induced disorders. For a detailed account of the sampling procedures utilized in the Community Outcomes of AOT study, see Phelan and colleagues (2010), Link and colleagues (2008), or Link and colleagues (2011).

Measures

Dependent variable

Violent victimization

Following the MacArthur violence scale (see Monahan et al., 2001), there are eight categories of violent victimization, including (a) having had something thrown at the participant; (b) being pushed, grabbed, or shoved; (c) being slapped; (d) kicked, bitten, or choked; (e) being hit with a fist or beaten; (f) being physically forced to have sex against his or her will; (g) being threatened with a knife, gun, or weapon; and (h) having someone fire a gun, use a knife, or weapon on the participant (Link et al., 2011; Phelan et al., 2010). If an individual experienced one of these acts of violence after baseline, the participant was coded as a victim resulting in a dichotomous indicator of violent victimization (1) or not (0).

Independent variable

Perceived societal stigma

Perceived societal stigma was captured by asking participants a series of Likert-type questions to determine how strongly they felt devalued/discriminated against, different, or ashamed. Response options included strongly disagree, disagree, agree, or strongly agree. To create a stigma scale, an exploratory factor analysis (EFA) was conducted. EFA results identified three latent constructs. To avoid multicollinearity issues, only one of the scales identified through the EFA was utilized. Specific questions included in the scale are, “most people would accept a person who has been in a mental hospital as a close friend,” “most people believe that a person who has been hospitalized for mental illness is just as trustworthy as the average citizen,” “most people would accept a person who has fully recovered from mental illness as a teacher of young children in a public school,” “most women would be willing to marry a man who has been a patient in a mental hospital,” or “most employers will hire a person who has been hospitalized for mental illness if he or she is qualified for the job.” Since the items showed acceptable reliability (Cronbach’s α = .74), a stigma measure was created by taking the mean of the five items at baseline. The resulting scale captures the perceptions that people with mental illnesses would not be accepted by most people.

Control variables

Violent victimization at baseline

Victimization reported at baseline is measured with the same items as the dependent variable and included as a control measure. If an individual reported any type of victimization in 1 month before the baseline interview, they are coded as (1), if they reported none of the victimizations, they are coded (0). This allows us to control for the effect of prior victimization on reported stigma.

Alcohol use at baseline

It is possible that alcohol use may lead to victimization. Thus, to test for this possibility, alcohol use is included as a control variable. The frequency of alcohol use (measured as ≥ 10 days/month, 2–9 days/month, <2 days/month) is collapsed into a dichotomous indicator of any alcohol use (1) or not (0) at baseline.

Delusions at baseline

It is also possible that experiencing a delusion could lead to a victimization event (Goodman et al., 1997; Johnson et al., 2016; Teasdale, 2009; Walsh et al., 2003). To account for this possibility, delusions are included as a control variable. Delusions are assessed using the clinician-administered Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders (DSM; that is, SCID; First & Gibbon, 2004). Specifically, participants were asked questions that assessed the presence or absence of delusions within the past 3 months. To control for delusions that may influence victimization (i.e., paranoid, persecutory, delusions of control, and thought broadcasting), the following questions are utilized: “has it seemed like people were talking about you or taking special notice of you?” “what about anyone going out of their way to give you a hard time or trying to hurt you?” “did you ever feel that your actions were being controlled or directed by some other external force?” and “did you feel as if your thoughts were being broadcast out loud so that other people could actually hear what you were thinking?” If an individual reported one of these delusions at baseline, the participant is coded as experiencing delusions (1) or not (0).

Violence at baseline

Eight categories of violent behavior are utilized, including (a) throwing something at someone; (b) pushing, grabbing, or shoving someone; (c) slapping someone; (d) kicking, biting, or choking someone; (e) hitting someone with a fist or beating someone up; (f) physically forcing someone to have sex against his or her will; (g) threatening someone with a knife, gun, or weapon; and (h) firing a gun, using a knife, or weapon on someone (Link et al., 2011; Phelan et al., 2010). If an individual reported one of these experiences at baseline, the participant is coded as violent resulting in a dichotomous indicator of violent behavior (1) or not (0).

Perceived coercion

Perceived coercion is assessed through a modified version of the MacArthur Perceived Coercion scale (Cronbach’s α = .86; Gardner et al., 1993), which captures if the respondent felt free to choose to be in outpatient treatment (Link et al., 2008). Examples include statements such as “it was your idea to get mental health treatment” or “you had control over getting mental health treatment.” Response options ranged from strongly agree (0) to strongly disagree (3). Thus, the perceived coercion scale is created by taking the mean of the six items at baseline.

Number of involuntary hospitalizations

The number of involuntary hospitalizations is measured by asking how many times the respondent had been hospitalized against their will at baseline. The natural log is taken to account for the skewness of the variable, with higher numbers representing a greater number of involuntary hospitalizations.

Site

To control for the site from which the respondent was recruited from, a site control measure is included. Specifically, three hospitals participating in the study are utilized. To maintain confidentiality, we have labeled and created a series of dummy variables of these sites as (a) Site 1, (b) Site 2, and (c) Site 3. Site 2 is the excluded referent category.

Age

The age in years of the respondent is included at baseline.

Sex

A dichotomous variable of sex at baseline is included (i.e., male [1] and female [0]).

Social desirability

Social desirability is captured through a 15-question version of the Crowne–Marlowe scale (Cronbach’s α = .68; Crowne & Marlowe, 1960). Examples include “never hesitating to help someone,” “never intensely disliked anyone,” or “always willing to admit mistakes.” Thus, a social desirability measure is created by taking the sum score of the 15 items at baseline.

Race

Three dummy variables, including Black, Hispanic, and White/Other, are used to control for race at baseline with White/Other as the reference category (only 32 subjects were code as White/Other).

Data Analysis

Due to the design utilized in the AOT data (i.e., interviews were conducted every 3 months for a year), we were able to lag waves to address temporal ordering. The dependent variable is measured as any victimization occurring between Month 1 and Month 12 after baseline, while all the independent variables are measured temporally prior (at baseline), including stigma and prior victimization. In addition, because logistic regression takes into account the dichotomous nature of the dependent variable (i.e., victim or not), a logistic regression model was employed (Hosmer & Lemeshow, 2000).

Because the participants were interviewed every 3 months for 1 year, there were some participants who did not report data at each wave. We chose multiple imputations to address missing data for several reasons. First, because the full sample can be retained, statistical power is increased. Second, Schafer and Graham (2002) highlight that multiple imputation is a more efficient missing data technique than case deletion because there are no units that are sacrificed. Third, the missing values are predicted from each of the participant’s previous observed values (Schafer & Graham, 2002). Multiple imputation approaches have been described extensively elsewhere (Rubin, 1987; Schafer & Graham, 2002; Sinharay et al., 2001). Briefly, the approach uses a regression-based estimate for item-level missingness, under a missing-at-random assumption. That is, we assume that the data are missing at random, conditional on the covariates included in the model, but that it is not dependent on the actual value of the missing data. The approach generates multiple data sets that contain complete data with multiple plausible values for the missing data. Considering that Graham and colleagues (2007) suggest that 40 imputed data sets can remove noise from statistical summaries, we utilized 40 imputed data sets that were pooled together to produce results using Rubin’s rules (Rubin, 1987).

Results

Sample Description

As shown in Table 1, approximately 57% reported that they had been violently victimized in 1 month prior to baseline. Over the course of the follow-up interviews, 18% reported violent victimization. The mean age of the participants was 37 years, and approximately 40% were females. Among the site locations, 38% of the participants were at Site 2, 35% at Site 3, and 26% were at Site 1. On average, the mean level of perceived stigma was 1.57 (on a scale that ranged from 0.00 to 3), indicating that the majority of the sample had moderate levels of perceived stigma. In addition, 31% of the participants had consumed alcohol in 1 month prior to baseline. The average level of social desirability was 8.49, on a scale ranging from 0 to 15, and the average level of perceived coercion was 1.37, on a scale ranging from 0 to 3. Finally, the average number of involuntary hospitalizations was 1.05. Finally, approximately 54% of the participants were Black, 29% were Hispanic, and 17% were reported as Other, which included White.

Table 1.

Sample Description (N = 184).

Variable	M	SE	Range
Dependent variable
Victimization	0.18	.029	0 to 1
Independent variable
Stigma	1.57	.040	0.00 to 3.00
Control measures
Alcohol use	0.31	.035	0 to 1
Delusions baseline	0.65	.035	0 to 1
Violence baseline	0.31	.035	0 to 1
Victimization baseline	0.57	.037	0 to 1
Perceived coercion	1.38	.045	0 to 3
Involuntary hospitalization	1.05	.068	0 to 2.71
Site 2	0.38	.036	0 to 1
Site 1	0.27	.033	0 to 1
Site 3	0.35	.035	0 to 1
Age	37.00	.819	18 to 64
Male	0.59	.037	0 to 1
Social desirability	8.49	.234	−3.01 to 21.99
Black	0.54	.037	0 to 1
Latino	0.29	.034	0 to 1
Other	0.17	.028	0 to 1

Note. Mean, standard error, and range reported from the pooled imputation model.

Bivariate and Multivariate Analysis

As shown in Table 2, only three measures were significantly correlated with violent victimization in the bivariate analysis. Specifically, the variables, perceived stigma, prior victimization at baseline, and delusions were significantly correlated with violent victimization in 1 year after baseline, in the bivariate analysis.

Table 2.

Results From Logistic Regression Predicting Victimization (N = 184).

Variables	Bivariate		Multivariate
Variables	b	SE	B	SE	Partial Odds Ratios
Perceived stigma	1.064**	.41	1.137*	.50	3.12
Delusions	1.186*	.52	1.198*	.57	3.31
Alcohol use	0.085	.44	−0.400	.53	.67
Violence	0.986*	.42	0.711	.49	2.04
Victimization baseline	1.572**	.51	1.332*	.57	3.79
Perceived coercion	0.135	.32	−0.210	.39	0.81
Involuntary Hospitalization	−0.169	.31	−0.134	.39	0.87
Site 1^a	0.204	.45	−0.098	.59	0.91
Site 3^a	−0.604	.45	−0.491	.57	0.61
Age	0.018	.02	0.016	.02	1.02
Male	−0.024	.41	−0.401	.51	0.67
Social desirability	0.010	.07	−0.015	.08	0.98
Blackb	−0.396	.40	−0.315	.59	0.73
Latinob	0.151	.43	−0.225	.67	0.80

Site 2 is the excluded reference category.

Other race is the excluded reference category.

p < .05. **p < .01.

In the multivariate analysis of Table 2, the same three variables also predicted violent victimization, controlling for the other variables in the model. As hypothesized, for individuals with higher perceptions of stigma, the odds of being victimized significantly increased. That is, for every one-point increase in perceived stigma, the odds of being victimized were approximately three times higher. To illustrate, as shown in Figure 1, people who are one to two standard deviations below the mean of perceived stigma have a 10% to 20% predicted probability of experiencing violent victimization. In contrast, among people who are one to two standard deviations above the mean of perceived stigma, the predicted probability of violent victimization doubles. That is, the predicted probability of experiencing violent victimization for people one standard deviation above the mean level of stigma is approximately 40%, while the predicted probability of experiencing victimization is above 50% for people who are two standard deviations above the mean level of stigma. As can be seen, there is a clear pattern of victimization risk increasing as perceived stigma increases.

Figure 1.

Predicted probability of victimization.

As shown in Table 2, in addition to perceived stigma, for people who have been victimized at baseline, the odds of being victimized at the follow-up waves significantly increase. Specifically, baseline victims had 3.79 times the odds of being victimized during the 1-year follow-up, compared with baseline nonvictims. This is unsurprising, given the literature showing that victimization can also lead to an increased risk of subsequent victimization, known as recurring victimization (see Fisher et al., 2010; Tseloni & Pease, 2003; Turanovic & Pratt, 2014). Finally, among people who were experiencing delusions, the odds of being victimized significantly increased by 3.31 times during the 1-year follow-up, compared with people who were not experiencing delusions.

Surprisingly, however, victimization at baseline does not predict perceived stigma at the following wave. As shown in Table 3, victimization at baseline and perceived stigma at the following wave are not significantly associated in the multivariate analysis. This suggests that some mechanism in feeling stigmatized influences violent victimization of individuals with mental disorders, but being victimized does not predict perceived stigma in later waves (i.e., follow-up period).

Table 3.

Results From Ordinary Least Squares Regression Predicting Stigma at 3 Months (N = 184).

Variables	b	SE
Victimization baseline	0.085	.08
Delusions	−0.129	.08
Alcohol use	−0.033	.09
Violence	0.167	.09
Stigma baseline	0.387***	.08
Perceived coercion	0.064	.07
Involuntary hospitalization	0.106	.01
Site 1^a	0.100	.10
Site 3^a	0.221**	.09
Age	0.004	.00
Male	0.015	.08
Social desirability	0.000	.02
Black^b	−0.026	.12
Latino^b	−0.072	.14

Site 2 is the excluded reference category.

Other race is the excluded reference category.

p < .05. **p < .01.

Discussion

There is relatively little known about if and how a person with mental illness’s own perception of stigma may affect victimization experiences. Although we have speculated about some potential mechanisms that may explain why stigma and victimization correlate, it is first necessary to determine if perceived stigma is related to victimization and in what direction. Therefore, the purpose of the current study was to fill this gap by investigating that possibility.

Results from this study suggest that perceived stigma and victimization are significantly associated and that perceived stigma actually increases the odds of experiencing a victimization event threefold. In other words, findings from our study suggest that the stigma attached to mental illness (i.e., being dangerous, violent, undesirable) is a statistically significant predictor of victimization experiences.

Results from this study also demonstrated that victimization at baseline did not predict perceived stigma at the follow-up wave (see Table 3). Although it could be assumed that a victimization event would further exacerbate one’s perception of stigma, the current study’s results did not support this assumption. It may be that (because most people are victimized by nonstrangers) the victimization feels personal (for reasons other than mental illness–related processes). Thus, it does not alter general perceptions of how most people feel about the disordered population (as captured by our stigma measure). Rather, they are able to explain away the victimization event using other explanations (conflicted social relationships; see Silver, 2002). Consequently, stigma remains unchanged. In contrast, those who experience more stigma are at an increased risk of later victimization.

Importantly, the current study controlled for a variety of variables that are significantly correlated with victimization. We did find that prior victimization events and experiencing delusions increased the odds of experiencing a victimization event. Given that prior victimization (Teasdale et al., 2014) and experiencing delusions (Goodman et al., 1997; Johnson et al., 2016; Teasdale, 2009; Walsh et al., 2003) are established predictors of victimization among people with mental illness, this is unsurprising. What is surprising, however, is that despite controlling for these variables, the relationship between perceived stigma and victimization remained unchanged.

Following from this finding, a natural next line of inquiry is understanding why such a relationship exists. Some contributing mechanisms that may be important to examine in future research include if and how disadvantaged situations, reduced guardianship, and reduced well-being contribute to the relationship between stigma and victimization because these may serve as potential mediating mechanisms.

Future pursuits should also investigate where stigma belief structures are learned. We suspect that people with mental disorders learn these beliefs from the stigmatizing attitudes of those around them. Small group social learning may be the culprit in creating these belief structures, and those who help create these stigmatizing beliefs may be the same individuals who are victimizing them. Future research should take a social networks approach to the study of stigma and victimization to test this assertion. It is also possible that individuals who perceive themselves to be stigmatized alter their activity patterns in ways that make them vulnerable to victimization. One such activity pattern may be engaging in risky behaviors such as alcohol use. Although we did not find a connection between alcohol use and victimization in this sample, it is fairly commonly connected to victimization in other research (Mustaine & Tewksbury, 1998), and those who are more stigmatized may turn to alcohol as a coping mechanism, setting themselves up for increased victimization risk. It is possible that the relatively low rates of alcohol use among this sample may be due to the outpatient treatment the individuals were receiving.

Limitations

As with any study, there are several limitations that should be addressed. First, because it was impossible to assign participants randomly to varying stigma levels, it is possible that we did not measure a third variable that could affect both stigma and victimization. We do address this possibility, however, by controlling for possible confounding variables that have been shown to influence victimization, including prior victimization events. Ultimately, like all survey research, we estimate an association, one that might be spurious due to some unmeasured factor. Thus, we caution that these results may not be causal. We are, however, bolstered in our expectation that this relationship is reliable, in that a confounder would simultaneously need to cause the perception of stigma in the victim and also the victimizing behavior of a perpetrator. It is unclear what that third variable could be that would cause these two processes in separate individuals. Moreover, the stability of the association, despite the controls we did include, suggests that it is robust to model specifications.

Our measure of stigma includes five items examining the participants’ perceptions of stigma. Although these items were useful for the purpose of the current study, future research should examine expanded measures of perceived stigma. Relatedly, our measure of stigma captures one’s perceived stigma but does not capture the perceptions of individuals around the participants (i.e., the stigma attached to individuals with mental disorders by others). Although we suspect that these beliefs are learned from those in small groups surrounding the respondent, it is possible that they were learned elsewhere, from the media for example. It would be useful for future studies to incorporate a social networks approach with such measures to estimate the contribution to victimization of these stigmatizing beliefs from others in the individual’s social network.

The measure of violent victimization consisted of eight behaviors ranging from having something thrown at the participant to having someone use a weapon against the respondent. Consequently, this measurement lumped in several categories that varied in the severity of victimization. Thus, by dichotomizing violent victimization, the current study could underestimate the impact of stigma on victimization severity. However, the low rate of involvement in victimization prevents studying variation in severity with the current data.

In addition, alcohol use was measured by collapsing frequency of use into a dichotomous measure of any alcohol use. While this measure captures usage, it fails to assess frequency, duration, and severity of use. It may be useful for future research to parse out the extent to which alcohol contributes to victimization by incorporating measures that extend beyond simply engaging in drinking alcohol and, instead, incorporating other measures that establish frequency, duration, and severity of use. The low rate of involvement in alcohol usage prevented us from examining the severity of use within the current data. Similarly, delusions were measured by assessing their presence, but not their intensity or frequency. It would be useful to future research to include measures of psychotic symptomatology that captures intensity and frequency to examine if and how such measures contribute to victimization or perceptions of stigma.

Conclusion

Despite these limitations, this study has important implications for clinical practice and informing literature. Indeed, this was one of the first studies to begin to answer Teasdale’s decade-old call to assess the role of stigma in victimization among individuals with mental disorders (Teasdale, 2009). Although the current study’s results suggest that stigma plays a role in the victimization experience among this population, there are still several unanswered questions. For instance, what are the mediators of this association that may help us understand why stigma leads to victimization experiences among individuals with mental disorders? For example, future research may want to investigate if and how nonviolent social cues or negative emotions may function as mediators in the relationship between stigma and victimization. In other words, it is possible that nonviolent social cues may be a byproduct of feeling stigmatized, which could then lead to a victimization experience. Similarly, negative emotions may also be another consequence of stigma, ultimately leading to victimization. For example, it is possible that perceptions of stigma can produce negative emotions. Negative emotions, in turn, may influence a person to behave in ways that are provocative and increase their involvement with conflicted social relationships (Silver, 2002), which may ultimately lead to a victimization event. Future research should explore these potential mediators and possibilities.

In terms of clinical practice, clinicians may want to be particularly sensitive to an individual’s perceptions of the stigma associated with his or her mental illness. Clinicians may want to instruct their clients on target-hardening strategies to adjust their perceptions of stigma, which, in turn, may reduce their victimization risk. We suggest nurturing social relationships with individuals who may serve as guardians is one such target-hardening strategy and is consistent with life skills training programs.

Finally, this study has important implications in informing the stigma literature regarding people with mental illness. A number of harmful outcomes related to stigma have already been identified in literature in this population. Results from this study suggest that victimization is another such harmful outcome, further illustrating the importance of addressing stigma-related processes. Furthermore, this study represents one of the initial efforts at understanding mental illness as a form of diversity in connection to stigma and victimization. Although diversity is often understood in terms of issues related to gender, race, class, and sexual orientation, mental health is also an important way in which individuals stigmatize or engage in othering (Goffman, 1963). In other words, mental illness is an important signifier of difference (stigma). Focusing on stigma, as we have done in the current study, is consistent with literatures on other traditionally focused studies of diversity such as sexual orientation, gender, and race. Perhaps by addressing stigma-related processes among this population through clinical practice and future research, an indirect benefit will be reducing a number of harmful outcomes, including victimization.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Michelle N. Harris

Author Biographies

Michelle N. Harris is a doctoral candidate at Georgia State University in the Department of Criminal Justice and Criminology. She holds a BA in psychology and criminal justice from the University of Arkansas and an MS in criminal justice and criminology from Georgia State University. Her research interests include mental health as it relates to criminological processes and victimology. Her recent work appears in Deviant Behavior, Journal of Interpersonal Violence, European Journal of Criminology, and Journal of Criminal Justice.

Miranda L. Baumann is a doctoral candidate in the Department of Criminal Justice and Criminology at Georgia State University. She graduated from Georgia State University with a BA in political science, a BS in criminal justice, and an MS in criminal justice and criminology. Her research focuses on violence committed by and against people with serious mental illness, the relationship between substance use and violence, and gun violence. Her work has been published in outlets including the Journal of Interpersonal Violence, the International Journal of Law and Psychiatry, and Pearson Education. She has been teaching at Georgia State University for 3 years, during which time she has taught statistical analyses in criminal justice, criminological theory, race in the criminal justice system, and American criminal courts.

Brent Teasdale is a professor and the Department Chair in Criminal Justice Sciences. He received his PhD in sociology from the Pennsylvania State University. His scholarship focuses on advanced quantitative methods, violence by and against people with mental disorders, and substance abuse prevention. His published work has appeared in outlets such as Criminal Justice & Behavior, Journal of Interpersonal Violence, Journal of Quantitative Criminology, Justice Quarterly, Prevention Science, and Social Problems. He has been recognized in several publications ranking the best criminologists in the United States (Copes et al., 2012; Walters, 2015; Khey, 2017).

Bruce G. Link earned his BA from Earlham College and MS and PhD degrees from Columbia University. His interests are centered on topics in psychiatric and social epidemiology. He has written on the connection between socioeconomic status and health, homelessness, violence, stigma, and discrimination. Currently he is conducting research aimed at understanding health disparities by race/ethnicity and socioeconomic status, the consequences of social stigma for people with mental illnesses, and the connection between mental illnesses and violent behaviors.

References

Choe

J. Y.

Teplin

L. A.

Abram

K. M.

(2008). Perpetration of violence, violent victimization, and severe mental illness: Balancing public health concerns. Psychiatric Services, 59(2), 153–164. https://doi.org/10.1176/ps.2008.59.2.153

Cohen

L. E.

Felson

(1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44, 588–608. https://doi.org/10.2307/2094589

Crisanti

A. S.

Frueh

B. C.

Archambeau

Steffen

J. J.

Wolff

(2014). Prevalence and correlates of criminal victimization among new admissions to outpatient mental health services in Hawaii. Community Mental Health Journal, 50(3), 296–304. https://doi.org/10.1007/s10597-013-9688-1

Crocker

(1999). Social stigma and self-esteem: Situational construction of self-worth. Journal of Experimental Social Psychology, 35(1), 89–107. https://doi.org/10.1006/jesp.1998.1369

Crowne

D. P.

Marlowe

(1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24(4), 349–354. https://doi.org/10.1037/h0047358

Cruwys

Gunaseelan

(2016). “Depression is who I am”: Mental illness identity, stigma and wellbeing. Journal of Affective Disorders, 189, 36–42 https://doi.org/10.1016/j.jad.2015.09.012

de Waal

M. M.

Christ

Dekker

J. J.

Kikkert

M. J.

Lommerse

N. M.

van den Brink

Goudriaan

A. E.

(2018). Factors associated with victimization in dual diagnosis patients. Journal of Substance Abuse Treatment, 84, 68–77. https://doi.org/10.1016/j.jsat.2017.11.001

Elbogen

E. B.

Johnson

S. C.

(2009). The intricate link between violence and mental disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 66(2), 152–161. https://doi.org/10.1001/archgenpsychiatry.2008.537

Estroff

S. E.

Zimmer

Lachicotte

W. S.

Benoit

(1994). The influence of social networks and social support on violence by persons with serious mental illness. Psychiatric Services, 45(7), 669–679. https://doi.org/10.1176/ps.45.7.669

10.

First

M. B.

Gibbon

(2004). The structured clinical interview for DSM-IV Axis I Disorders (SCID-I) and the structured clinical interview for DSM-IV Axis II Disorders (SCID-II). In Hilsenroth

M. J.

Segal

D. L.

(Eds.), Comprehensive handbook of psychological assessment (pp. 134–152). John Wiley.

11.

Fisher

B. S.

Daigle

L. E.

Cullen

F. T.

(2010). What distinguishes single sexual victims from recurring ones? The role of lifestyle-routine activities and first-incident characteristics. Justice Quarterly, 27, 102–129. https://doi.org/10.1080/07418820902763061

12.

Fitzpatrick

K. M.

La Gory

M. E.

Ritchey

F. J.

(1993). Criminal victimization among the homeless. Justice Quarterly, 10(3), 353–368. https://doi.org/10.1080/07418829300091881

13.

Gardner

Hoge

S. K.

Bennett

Roth

L. H.

Lidz

C. W.

Monahan

Mulvey

E. P.

(1993). Two scales for measuring patients’ perceptions for coercion during mental hospital admission. Behavioral Sciences & the Law, 11(3), 307–321. https://doi.org/10.1002/bsl.2370110308

14.

Gerbner

Gross

Morgan

Signorielli

(1981). Health and medicine on television. New England Journal of Medicine, 305(15), 901–904. https://doi.org/10.1056/NEJM198110083051530

15.

Goffman

(1963). Stigma: Notes on the management of spoiled identity. Simon & Schuster.

16.

Goodman

L. A.

Rosenberg

S. D.

Mueser

K. T.

Drake

R. E.

(1997). Physical and sexual assault history in women with serious mental illness: Prevalence, correlates, treatment, and future research directions. Schizophrenia Bulletin, 23(4), 685–696. https://doi.org/10.1093/schbul/23.4.685

17.

Goodman

L. A.

Salyers

M. P.

Mueser

K. T.

Rosenberg

S. D.

Swartz

Essock

S. M.

, . Swanson

(2001). Recent victimization in women and men with severe mental illness: Prevalence and correlates. Journal of Traumatic Stress, 14(4), 615–632.

18.

Goodman

L. A.

Thompson

K. M.

Weinfurt

Corl

Acker

Mueser

K. T.

Rosenberg

S. D.

(1999). Reliability of reports of violent victimization and posttraumatic stress disorder among men and women with serious mental illness. Journal of Traumatic Stress, 12(4), 587–599. https://doi.org/10.1023/A:1024708916143

19.

Graham

J. W.

Olchowski

A. E.

Gilreath

T. D.

(2007). How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prevention Science, 8(3), 206–213. https://doi.org/10.1007/s11121-007-0070-9

20.

Hiday

V. A.

(1997). Understanding the connection between mental illness and violence. International Journal of Law and Psychiatry, 20(4), 399–417. https://doi.org/10.1016/S0160-2527(97)00028-9

21.

Hiday

V. A.

Swartz

M. S.

Swanson

J. W.

Borum

Wagner

H. R.

(1999). Criminal victimization of persons with severe mental illness. Psychiatric Services, 50(1), 62–68. https://doi.org/10.1176/ps.50.1.62

22.

Hiday

V. A.

Swartz

M. S.

Swanson

J. W.

Borum

Wagner

H. R.

(2002). Impact of outpatient commitment on victimization of people with severe mental illness. American Journal of Psychiatry, 159(8), 1403–1411. https://doi.org/10.1176/appi.ajp.159.8.1403

23.

Hosmer

D. W.

Lemeshow

(2000). Applied logistic regression. John Wiley.

24.

Johnson

K. L.

Desmarais

S. L.

Tueller

S. J.

Grimm

K. J.

Swartz

M. S.

Van Dorn

R. A.

(2016). A longitudinal analysis of the overlap between violence and victimization among adults with mental illnesses. Psychiatry Research, 246, 203–210. https://doi.org/10.1016/j.psychres.2016.09.039

25.

Khalifeh

Oram

Osborn

Howard

L. M.

Johnson

(2016). Recent physical and sexual violence against adults with severe mental illness: A systematic review and meta-analysis. International Review of Psychiatry, 28(5), 433–451. https://doi.org/10.1080/09540261.2016.1223608

26.

Langeveld

Bjørkly

Evensen

Joa

Johannessen

J. O.

Larsen

T. K.

, … Friis

(2018). A 10-year follow-up study of violent victimization in first episode psychosis: Risk and protective factors. Psychiatry Research, 259, 545–549. https://doi.org/10.1016/j.psychres.2017.11.013

27.

Latalova

Kamaradova

Prasko

(2014). Violent victimization of adult patients with severe mental illness: A systematic review. Neuropsychiatric Disease and Treatment, 10, 1925–1939.

28.

Lee

B. A.

Schreck

C. J.

(2005). Danger on the streets: Marginality and victimization among homeless people. American Behavioral Scientist, 48(8), 1055–1081. https://doi.org/10.1177/0002764204274200

29.

Link

B. G.

Castille

D. M.

Stuber

(2008). Stigma and coercion in the context of outpatient treatment for people with mental illnesses. Social Science & Medicine, 67(3), 409–419. https://doi.org/10.1016/j.socscimed.2008.03.015

30.

Link

B. G.

Cullen

F. T.

Struening

Shrout

P. E.

Dohrenwend

B. P.

(1989). A modified labeling theory approach to mental disorders: An empirical assessment. American Sociological Review, 54(3), 400–423. https://doi.org/10.2307/2095613

31.

Link

B. G.

Epperson

M. W.

Perron

B. E.

Castille

D. M.

Yang

L. H.

(2011). Arrest outcomes associated with outpatient commitment in New York State. Psychiatric Services, 62(5), 504–508. https://doi.org/10.1176/ps.62.5.pss6205_0504

32.

Link

B. G.

Phelan

J. C.

(2001). Conceptualizing stigma. Annual Review of Sociology, 27(1), 363–385. https://doi.org/10.1146/annurev.soc.27.1.363

33.

Link

B. G.

Phelan

J. C.

Bresnahan

Stueve

Pescosolido

B. A.

(1999). Public conceptions of mental illness: Labels, causes, dangerousness, and social distance. American Journal of Public Health, 89(9), 1328–1333. https://doi.org/10.2105/AJPH.89.9.1328

34.

Link

B. G.

Struening

E. L.

Neese-Todd

Asmussen

Phelan

J. C.

(2001). Stigma as a barrier to recovery: The consequences of stigma for the self-esteem of people with mental illnesses. Psychiatric Services, 52(12), 1621–1626. https://doi.org/10.1176/appi.ps.52.12.1621

35.

Link

B. G.

Struening

E. L.

Rahav

Phelan

J. C.

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 Social Behavior, 38(2), 177–190. https://doi.org/10.2307/2955424

36.

Livingston

J. D.

Boyd

J. E.

(2010). Correlates and consequences of internalized stigma for people living with mental illness: A systematic review and meta-analysis. Social Science & Medicine, 71(12), 2150–2161. https://doi.org/10.1016/j.socscimed.2010.09.030

37.

Maniglio

(2009). Severe mental illness and criminal victimization: A systematic review. Acta Psychiatrica Scandinavica, 119(3), 180–191. https://doi.org/10.1111/j.1600-0447.2008.01300.x

38.

Marley

J. A.

Buila

(2001). Crimes against people with mental illness: Types, perpetrators, and influencing factors. Social Work, 46(2), 115–124. https://doi.org/10.1093/sw/46.2.115

39.

Monahan

Steadman

H. J.

Silver

Appelbaum

P. S.

Robbins

P. C.

Mulvey

E. P.

, . Banks

(2001). Rethinking risk assessment: The macarthur study of mental disorder and violence. Oxford University Press.

40.

Monahan

Vesselinov

Robbins

P. C.

Appelbaum

P. S.

(2017). Violence to others, violent self-victimization, and violent victimization by others among persons with a mental illness. Psychiatric Services, 68(5), 516–519. https://doi.org/10.1176/appi.ps.201600135

41.

Mulvey

E. P.

(1994). Assessing the evidence of a link between mental illness and violence. Psychiatric Services, 45(7), 663–668. https://doi.org/10.1176/ps.45.7.663

42.

Mustaine

E. E.

Tewksbury

(1998). Specifying the role of alcohol in predatory victimization. Deviant Behavior, 19(2), 173–199. https://doi.org/10.1080/01639625.1998.9968082

43.

Myrick

J. G.

Pavelko

R. L.

(2017). Examining differences in audience recall and reaction between mediated portrayals of mental illness as trivializing versus stigmatizing. Journal of Health Communication, 22(11), 876–884. https://doi.org/10.1080/10810730.2017.1367338

44.

Osgood

D. W.

Wilson

J. K.

O’Malley

P. M.

Bachman

J. G.

Johnston

L. D.

(1996). Routine activities and individual deviant behavior. American Sociological Review, 61(4), 635–655. https://doi.org/10.2307/2096397

45.

Owen

P. R.

(2012). Portrayals of schizophrenia by entertainment media: A content analysis of contemporary movies. Psychiatric Services, 63(7), 655–659. https://doi.org/10.1176/appi.ps.201100371

46.

Pescosolido

B. A.

Monahan

Link

B. G.

Stueve

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), 1339–1345. https://doi.org/10.2105/AJPH.89.9.1339

47.

Phelan

J. C.

Link

B. G.

Stueve

Pescosolido

B. A.

(2000). Public conceptions of mental illness in 1950 and 1996: What is mental illness and is it to be feared? Journal of Health and Social Behavior, 41(2), 188–207. https://doi.org/10.2307/2676305

48.

Phelan

J. C.

Sinkewicz

Castille

D. M.

Huz

Link

B. G.

(2010). Effectiveness and outcomes of assisted outpatient treatment in New York State. Psychiatric Services, 61(2), 137–143. https://doi.org/10.1176/ps.2010.61.2.137

49.

Policastro

Teasdale

Daigle

L. E.

(2016). The recurring victimization of individuals with mental illness: A comparison of trajectories for two racial groups. Journal of Quantitative Criminology, 32(4), 675–693. https://doi.org/10.1007/s10940-015-9271-8

50.

Reynders

Kerkhof

A. J. F. M.

Molenberghs

Van Audenhove

(2014). Attitudes and stigma in relation to help-seeking intentions for psychological problems in low and high suicide rate regions. Social Psychiatry & Psychiatric Epidemiology, 49(2), 231–239. https://doi.org/10.1007/s00127-013-0745-4

51.

Rubin

D. B.

(1987). Multiple imputation for nonresponse in surveys. John Wiley.

52.

Rüsch

Zlati

Black

Thornicroft

(2014). Does the stigma of mental illness contribute to suicidality? The British Journal of Psychiatry, 205(4), 257–259. https://doi.org/10.1192/bjp.bp.114.145755

53.

Ryan

Rapley

Dziurawiec

(2014). The meaning of coping for psychiatric patients. Qualitative Health Research, 24(8), 1068–1079. https://doi.org/10.1177/1049732314541011

54.

Schafer

J. L.

Graham

J. W.

(2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177. https://doi.org/10.1037/1082-989X.7.2.147

55.

Silver

(2000). Extending social disorganization theory: A multilevel approach to the study of violence among persons with mental illnesses. Criminology, 38(4), 1043–1074. https://doi.org/10.1111/j.1745-9125.2000.tb01414.x

56.

Silver

(2002). Mental disorder and violent victimization: The mediating role of involvement in conflicted social relationships. Criminology, 40(1), 191–212. https://doi.org/10.1111/j.1745-9125.2002.tb00954.x

57.

Silver

(2006). Understanding the relationship between mental disorder and violence: The need for a criminological perspective. Law and Human Behavior, 30(6), 685–706. https://doi.org/10.1007/s10979-006-9018-z

58.

Silver

Arseneault

Langley

Caspi

Moffitt

T. E.

(2005). Mental disorder and violent victimization in a total birth cohort. American Journal of Public Health, 95(11), 2015–2021. https://doi.org/10.2105/AJPH.2003.021436

59.

Sinharay

Stern

H. S.

Russell

(2001). The use of multiple imputation for the analysis of missing data. Psychological Methods, 6(4), 317–329. https://doi.org/10.1037/1082-989X.6.4.317

60.

Stafford

M. C.

Scott

R. R.

(1986). Stigma, deviance, and social control. In S. C. Ainlay, L. M. Becker, & L. M. Coleman (Eds.), The dilemma of difference (pp. 77–91). Springer.

61.

Swartz

M. S.

Swanson

J. W.

Hiday

V. A.

Borum

Wagner

H. R.

Burns

B. J.

(1998). Violence and severe mental illness: The effects of substance abuse and nonadherence to medication. American Journal of Psychiatry, 155(2), 226–231.

62.

Teasdale

(2009). Mental disorder and violent victimization. Criminal Justice and Behavior, 36(5), 513–535. https://doi.org/10.1177/0093854809331793

63.

Teasdale

Daigle

L. E.

Ballard

(2014). Trajectories of recurring victimization among people with major mental disorders. Journal of Interpersonal Violence, 29(6), 987–1005. https://doi.org/10.1177/0886260513506054

64.

Teplin

L. A.

McClelland

G. M.

Abram

K. M.

Weiner

D. A.

(2005). Crime victimization in adults with severe mental illness: Comparison with the National Crime Victimization Survey. Archives of General Psychiatry, 62(8), 911–921. https://doi.org/10.1001/archpsyc.62.8.911

65.

Tseloni

Pease

(2003). Repeat personal victimization: “Boosts” or “flags”? British Journal of Criminology, 43, 196–212. https://doi.org/10.1093/bjc/43.1.196

66.

Turanovic

J. J.

Pratt

T. C.

(2014). “Can’t stop, won’t stop”: Self-control, risky lifestyles, and repeat victimization. Journal of Quantitative Criminology, 30(1), 29–56. https://doi.org/10.1007/s10940-012-9188-4

67.

Verhaeghe

Bracke

Bruynooghe

(2008). Stigmatization and self-esteem of persons in recovery from mental illness: The role of peer support. International Journal of Social Psychiatry, 54(3), 206–218. https://doi.org/10.1177/0020764008090422

68.

Walsh

Moran

Scott

McKenzie

Burns

Creed

, . Fahy

(2003). Prevalence of violent victimisation in severe mental illness. The British Journal of Psychiatry, 183(3), 233–238. https://doi.org/10.1192/bjp.183.3.233