You are on page 1of 17

794873

research-article2018
GPI0010.1177/1368430218794873Group Processes & Intergroup RelationsMaunder et al.

G
Group Processes & P
Intergroup Relations I
Article R

Group Processes & Intergroup Relations

Modern avenues for intergroup contact: 1­–17


© The Author(s) 2018

Using E-contact and intergroup


Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1368430218794873
emotions to reduce stereotyping and
https://doi.org/10.1177/1368430218794873
journals.sagepub.com/home/gpi

social distancing against people with


schizophrenia

Rachel D. Maunder,1 Fiona A. White1, and Stefano Verrelli1

Abstract
Intergroup contact is the leading strategy for reducing the stigma associated with mental illness. For
the first time, the current study examines the effectiveness of a contemporary intergroup contact
strategy, called electronic or E-contact, to reduce stigma against people diagnosed with schizophrenia.
It also examines the mediating role of three target-relevant intergroup emotions, namely fear, anger,
and pity. In total, 133 participants engaged in E-contact with a person diagnosed with schizophrenia
(intergroup E-contact), E-contact with a person without a mental illness (intragroup E-contact), or
no contact. Compared to the intragroup E-contact and no-contact conditions, intergroup E-contact
reduced fear, anger, and stereotyping toward people with schizophrenia. Additionally, fear and anger,
but not pity, were found to be significant affective mediators of the E-contact effect. The findings
demonstrate the value of computer-mediated intergroup contact for stigma reduction, and emphasize
the importance of intergroup emotions in this domain.

Keywords
computer-mediated communication, intergroup contact, intergroup emotions, schizophrenia, stigma

Paper received 22 November 2017; revised version accepted 20 July 2018.

The stigma faced by people with mental illness is the resulting prejudice and discrimination—have
a significant barrier to the implementation of deleterious consequences for sufferers, including
effective mental health treatments (Henderson
et al., 2014; World Health Organization [WHO], 1University of Sydney, Australia
2001). In comparison to other mental illnesses,
such as depression, individuals diagnosed with Corresponding author:
Rachel D. Maunder, School of Psychology, University of
schizophrenia are perceived to be more danger- Sydney, Camperdown, Sydney, New South Wales 2006,
ous, unpredictable, and incompetent (Angermeyer Australia.
& Matschinger, 2003). These stereotypes—and Email: rachel.maunder@sydney.edu.au
2 Group Processes & Intergroup Relations 00(0)

lower well-being and reduced employment indirect contact (e.g., imagined contact; Stathi,
opportunities (Link & Phelan, 2001), delayed Tsantila, & Crisp, 2012; West, Holmes, &
treatment seeking (Clement et al., 2015), and Hewstone, 2011) with a person diagnosed with
exacerbated symptoms (Farrelly et al., 2015). schizophrenia reduced negative and avoidance-
Thus, with the aim of reducing stigma against related attitudes, emotions, and behaviors toward
people with schizophrenia, the current study the entire group.
sought to experimentally evaluate a modern Recently, White and Abu-Rayya (2012) devel-
extension of the intergroup contact hypothesis, oped a contemporary form of intergroup con-
which suggests that positive online contact tact, termed electronic or E-contact, which may
between groups can reduce stigma (White, be particularly suitable for reducing stigma against
Harvey, & Abu-Rayya, 2015). Moreover, given people with schizophrenia. E-contact constitutes
the increasing focus on affective processes to a computer-mediated form of communication
explain stigma reduction (e.g., Seger, Banerji, where members of different groups never physi-
Park, Smith, & Mackie, 2017), the study also cally meet but interact online (Amichai-
aimed to examine the mediating role of three dis- Hamburger & McKenna, 2006). This intervention
tinct intergroup emotions related to schizophre- has been conceptualized as a bridge between
nia, namely fear, anger, and pity. direct and indirect forms of contact. Similar to
direct contact, E-contact involves a dyadic inter-
Intergroup Contact: Reducing action and reciprocal communication with an
Stigma Against People With outgroup member. These situational features
allow for greater engagement of the self in the
Schizophrenia immediate contact situation compared to indirect
One of the leading strategies to reduce stigma is contact (White, Harvey, & Abu-Rayya, 2015).
intergroup contact. Here, positive interactions But, unlike direct contact, participants do not
with a member of a stigmatized group are interact face-to-face and are not required to
hypothesized to improve attitudinal and emo- occupy the same physical space as the outgroup
tional evaluations of the group as a whole member. As such, E-contact may lack some of
(Allport, 1954). While the benefits of intergroup the multisensory complexities and richness char-
contact for stigma reduction have been exten- acteristic of face-to-face contact (Harwood,
sively demonstrated, in particular toward ethnic 2010). However, the indirect nature of E-contact
and racial groups (Pettigrew & Tropp, 2006), may make it a more suitable strategy when face-
there is markedly less research that has attempted to-face interactions are impractical, particularly
to extend these findings to people with schizo- when groups are divided either physically or
phrenia. In the context of mental health stigma, psychologically.
meta-analytic evidence suggests that intergroup White and Abu-Rayya (2012) recently utilized
contact can reduce stigma against people with a structured E-contact program that required
mental illness in general, and that contact is more Muslim and Christian students from segregated
effective than other education- and protest-based high schools to collaborate on problem-based
interventions in both the short and long term activities in a synchronous and text-based chat
(Corrigan, Morris, Michaels, Rafacz, & Rüsch, room. Compared to the control condition, which
2012). However, the lack of diagnostic specificity involved an intragroup interaction (i.e., E-contact
in these studies has restricted the conclusions that between students of the same religion), the inter-
can be drawn.1 Nevertheless, research specifically group E-contact program successfully reduced
targeting schizophrenia has found that both self- intergroup bias and anxiety. While there is grow-
reported (West, Hewstone, & Lolliot, 2014) and ing empirical support for the benefits of
experimentally manipulated direct (i.e., face-to- E-contact for improving intergroup relations
face; Giacobbe, Stukas, & Farhall, 2013) and between religious and ethnic groups (e.g., White,
Maunder et al. 3

Turner, Verrelli, Harvey, & Hanna, in press) and Finally, E-contact enables members of both
between heterosexuals and homosexuals (White, majority and minority groups to be actively
Verrelli, Maunder, & Kervinen, in press), the effi- involved in the same intergroup interaction
cacy of E-contact for reducing stigma against (White, Harvey, & Verrelli, 2015), which can be
mental illnesses such as schizophrenia remains structured to incorporate Allport’s (1954) facili-
untested (for a review of past E-contact research, tating contact conditions of equal status, coop-
see White, Harvey, & Abu-Rayya, 2015). eration, common goal, and authority support
There are several reasons to expect online (White & Abu-Rayya, 2012; see the Procedure
interactions, particularly the structured and text- and Materials section for how this can be
based E-contact program developed by White achieved). Although the adherence to these con-
and Abu-Rayya (2012), to be a practical and ditions has been shown to significantly facilitate
effective contact strategy for reducing stigma stigma reduction (Pettigrew & Tropp, 2006), they
against people with schizophrenia. Specifically, are consistently absent from contact interven-
E-contact does not require individuals to occupy tions aimed at reducing mental health stigma in
the same physical space. This characteristic may general.
be particularly important for intergroup contact
with people with schizophrenia, as this group is Reducing Stigma Through
often stereotyped as dangerous and unpredicta-
ble (Angermeyer & Matschinger, 2003, 2004).
Intergroup Emotions
Interacting with this outgroup online may shift Past research has emphasized the critical role of
the focus away from securing one’s physical intergroup emotions, in particular intergroup
safety and onto achieving the specific goals of anxiety, for stigma reduction (Pettigrew & Tropp,
the intergroup interaction. E-contact also pro- 2008). Intergroup contact with the outgroup is
vides both interaction partners with a high thought to reduce the anxiety or discomfort an
degree of psychological control over the contact individual may feel when anticipating or experi-
situation (e.g., being able to edit one’s message encing an interaction with an outgroup member,
before sending it), and can reduce negative ste- which in turn reduces stigma (Stephan & Stephan,
reotyping by ameliorating the nonverbal cues 1985). Research specific to intergroup contact
that are often used to adversely appraise the out- with people with schizophrenia has also high-
group (Amichai-Hamburger & McKenna, 2006). lighted the importance of intergroup emotions.
Together, these factors have been shown to For example, Stathi et al. (2012) found that imag-
reduce anxiety and create more intimate interac- ining contact with a person with schizophrenia
tions (White, Harvey, & Abu-Rayya, 2015), which decreased stereotyping and improved intentions
are critical for disconfirming the negative stereo- for future contact via reduced anxiety. Similarly,
types that often contribute to stigma (Stephan & West et al. (2014) demonstrated that contact with
Stephan, 1985). a person with schizophrenia improved attitudes
Moreover, the anonymity of online contact and reduced avoidance via decreased fear.
makes it more likely that individuals will disclose Nevertheless, this small body of research has
a stigmatized identity (Amichai-Hamburger & either focused entirely on negative emotions,
McKenna, 2006). This is particularly pertinent to such as fear and anxiety (Stathi et al., 2012; West
people with mental illness, as many members of et al., 2014), or has treated distinct emotions as a
this group conceal their diagnosis (Schulze & single affective response toward the outgroup
Angermeyer, 2003). Without identity disclosure, (Giacobbe et al., 2013). These approaches are
groups remain psychologically divided, which sig- inconsistent with current theorizing in intergroup
nificantly limits the extent to which the positive emotions research, which has emphasized the
impression one forms of the outgroup contact importance of differentiating between qualita-
partner will generalize to the outgroup as a whole. tively distinct group-targeted emotions (e.g., fear
4 Group Processes & Intergroup Relations 00(0)

vs. anger), in addition to considering the role of group-targeted emotions. By experimentally


positive emotions (e.g., pity) alongside negative manipulating intergroup E-contact, the current
ones (Cottrell & Neuberg, 2005). study aimed to investigate the causal relationship
The sociofunctional approach (Cottrell & between contact and emotions. Specifically, we
Neuberg, 2005) to stigma underscores the impor- sought to replicate past research (see West et al.,
tance of focusing on, and differentiating between, 2014) by examining fear toward people with schiz-
distinct intergroup emotions. According to this ophrenia as an affective mediator of the effect of
framework, it is essential to consider people’s contact on stereotyping and social distance. But
cognitive, threat-based appraisals of an outgroup more importantly, we also aimed to extend this
when predicting their specific affective reactions preliminary work by investigating the mediating
toward that outgroup (Cottrell & Neuberg, 2005). role of additional intergroup emotions, namely
For example, research has consistently demon- anger and pity.
strated that people with schizophrenia are often
evaluated as (a) dangerous and unpredictable, (b)
Overview of Aims and
incompetent and unable to follow accepted social
norms, and (c) not personally responsible for Hypotheses
their illness (Angermeyer & Matschinger, 2003, By integrating recent advances in intergroup
2004). These three distinct appraisals make it rea- contact and emotions research, the current study
sonable to expect an affective profile character- is the first to experimentally evaluate the efficacy
ized by fear, anger, and pity when examining of E-contact for reducing stereotyping and
emotions targeted at people with schizophrenia, social distance toward people with schizophre-
rather than other intergroup emotions such as nia. Consistent with the sociofunctional account
disgust, guilt, or admiration. This is because fear, of stigma (Cottrell & Neuberg, 2005), it was
anger, and pity are evoked when the outgroup is expected that positive contact with a person
perceived to (a) threaten one’s physical safety, (b) diagnosed with schizophrenia would reduce
violate social norms, and (c) not be personally stigma through group-targeted emotions specific
accountable for their current affliction, respec- to people with schizophrenia, namely fear, anger,
tively (Cottrell & Neuberg, 2005). While past and pity. Specifically, it was predicted that
research has shown that increased fear and E-contact would reduce stereotyping and social
anger as well as reduced pity play a significant distance by reducing anger and fear and increas-
role in accounting for the discrimination toward ing pity toward people with schizophrenia.
people with schizophrenia (Corrigan, Markowitz,
Watson, Rowan, & Kubiak, 2003), these distinct
emotions have not been investigated together in Method
the context of intergroup contact with a person Participants and Design
diagnosed with schizophrenia.
Further emphasizing the critical role of distinct In total, 175 undergraduate psychology students
emotions for stigma reduction, White, Abu-Rayya, were recruited through the University of Sydney’s
Bliuc, and Faulkner (2015) demonstrated that research recruitment program, in which first-year
reduced expressions of text-based anger and sad- psychology students participate in research in
ness, but not positive affect, during interreligious exchange for partial course credit. Participants
E-contact were related to reduced intergroup bias. were excluded from the analysis if they reported
However, this study only assessed participants’ a history of mental illness (n = 22), expressed
general expressions of positive and negative emo- suspicion about the study (n = 18), or failed to rec-
tions during the interaction, rather than their spe- ognize their online interaction partner had schizo-
cific affective responses toward the outgroup. It is, phrenia (n = 2). Thus, the final sample included 133
therefore, unclear if E-contact reduced bias via participants (Mage = 18.8 years, SD = 1.50), of
Maunder et al. 5

which the majority identified as female (64.7%), conversation with a fellow student, the interaction
White (49.6%), and heterosexual (94.0%). partner addressed the participant by name and
The study had a three-group between-subjects commented on the responses made by the partici-
design, with participants randomly allocated to an pant (e.g., “Good advice, [participant]!”).
intergroup E-contact, intragroup E-contact, or Participants were notified when the moderator
no-contact condition.2 Two control groups (intra- and interaction partner were typing by a message
group E-contact and no contact) were utilized to on the bottom of the screen (e.g., “Jasmine is typ-
verify that, if any change in stigma occurred, it ing…”), which appeared after a delay to account
was due to the intergroup nature of contact for time spent reading the preceding messages.
rather than contact itself and that intragroup The duration for which the moderator and inter-
contact did not raise stigma relative to intergroup action partner were ostensibly typing was also var-
contact, respectively.3 The dependent variables ied to account for differing lengths of responses.
included intergroup emotions (i.e., fear, anger, The interaction, which had a median duration
and pity), stereotyping, and social distance. of 16.85 minutes, was divided into two stages:
introduction and cooperation. In the introduction
stage, the moderator asked participants to intro-
Procedure and Materials duce themselves with their name, age, and other
All participants provided written consent. personal information such as hobbies. During this
Participants in the intergroup and intragroup stage, the interaction partner disclosed that they
E-contact conditions were informed that they had chosen to study psychology because they had
would be participating in an online, text-only chat been diagnosed with schizophrenia (intergroup
with an interaction partner under the supervision E-contact), or had an interest in mental illness
of a chat moderator. The researchers were pur- (intragroup E-contact). For example:
portedly developing this chat program to be dis-
tributed to future first-year university students. Moderator: 
So, firstly, what subjects are you
Participants believed that their interaction partner currently studying? Jasmine, you
(named Jasmine or Joseph) was another university go first.
student on campus; however, unknown to partici- Jasmine: 
ok. Hi [participant]! Nice to
pants, the responses of their interaction partner meet you :) I’m studying psych,
and chat moderator were preprogrammed. This and also government and inter-
method has been utilized successfully in other national relations, writing, and
studies (e.g., White, Turner, et al., in press; White, history
Verrelli, et al., in press) to standardize the interac-
tion across participants and remove the need for  hanks Jasmine. And what
Moderator: T
outgroup confederates. To increase the match about you [participant]?
between participants’ responses and the responses Participant: 
Hi there! Nice to meet you too!
of the virtual confederates, the preprogrammed I’m studying science so maths,
moderator and interaction partner always led and stats, chemistry, psyc and bio.
initiated the online chat. This structure encour-
aged participants to stay on track and only answer In the cooperation stage, participants then com-
the specific questions that were asked of them pleted a collaborative task with their interaction
during the short exchange. In doing so, the mod- partner that was guided by Allport’s (1954) facili-
erator designated a turn-taking style of respond- tating contact conditions. This task required the
ing by indicating when and how the participant two first-year, gender-matched university stu-
and their interaction partner should respond. dents (i.e., equal status) to work together (i.e.,
Moreover, to increase participants’ perceptions cooperation) to develop strategies to assist future
that they were experiencing a real, reciprocal first-year students transition from high school to
6 Group Processes & Intergroup Relations 00(0)

university (i.e., common goal), under the supervi- about the demographic characteristics of the par-
sion of a chat moderator (i.e., authority support). ticipant’s interaction partner, such as their gender
As the participants were first-year students them- and ethnicity. To elicit any suspicion over the pre-
selves, this topic was highly relevant to their inter- programmed nature of the interaction, partici-
ests. For example: pants were also asked for any comments they had
about the interaction, their interaction partner,
Moderator: 
Now let’s discuss the transition and the chat tool.
from high school to uni in terms Before completing the dependent variables,
of workload. Obviously it’s very which were presented in a randomized order,
different. What would you tell new participants indicated their gender, age, ethnicity,
students to help them, [Participant]? sexuality, and history of mental illness. Corrigan
et al.’s (2003) 10-item Emotional Response Scale
Participant: the best advice I can have is to keep
was used to measure the extent participants
a diary with all the dates of assess-
would experience fear (four items; e.g., “I would
ments and assignments of when
feel frightened”; Cronbach’s α = .96), anger
things are due as teachers aren’t at
(three items; e.g., “I would feel irritated”;
uni to remind you - and to not
Cronbach’s α = .85), and pity (three items; e.g.,
stress but try and stay on top with
“I would feel sympathy”; Cronbach’s α = .72)
notes etc. throughout the semester
toward a person with schizophrenia (1 = strongly
Moderator: Nice one. What do you think, disagree, 7 = strongly agree). O’Reilly, Bell, and
Jasmine? Chen’s (2010) eight-item Stigmatization Scale
was used to measure the extent participants
Jasmine: 
Good advice, [Participant]! I
endorsed (1 = strongly disagree, 5 = strongly agree)
wasn’t prepared for the differ-
stereotypes about people with schizophrenia (e.g.,
ence in work between hs and uni.
“People with schizophrenia will not improve
You have to organise yourself a
after treatment”; Cronbach’s α = .71). Link’s
lot more and do things before
(1987) Social Distance Scale was used to measure
lectures so you can understand
participants’ willingness (1 = definitely willing, 4 =
the content better. No one told
definitely unwilling) to be in seven different social
me about readers and things and
relationships with a person with schizophrenia
it took me a while to work it out.
(e.g., “Share a flat with a person with schizophre-
New people coming to uni for
nia”; Cronbach’s α = .90). In the no-contact
the first time should be told
control condition, participants answered the
about readings and how to struc-
same scales with no E-contact preceding them.
ture their time better
Finally, participants were fully debriefed, and the
researcher probed for suspicion by asking par-
At the end of the interaction, participants’ per-
ticipants for any comments or questions about
ception of Allport’s (1954) facilitating conditions
the study. Participants who reported guessing the
was assessed. Participants were asked to rate (1 =
interaction was preprogrammed or who identi-
not at all, 5 = extremely) the extent to which the
fied a connection between the two studies were
interaction was positive, of equal status, cooper-
excluded.
ative, goal-oriented, and had support from
authority. Further, to ensure that participants
were aware of the interaction partner’s diagnosis, Results
participants in both E-contact conditions were
asked to identify the mental illness their partner
Preliminary Analyses
had been diagnosed with, if any. Importantly, First, to evaluate participants’ perception of the
this question was couched with others asking quality of the interaction, one-sample t tests were
Maunder et al. 7

Table 1.  Means (and standard deviations) of Allport’s (1954) facilitating contact conditions.
E-contact condition
Overall mean t p d
  Intergroup Intragroup
Positive interaction 3.96 (0.69) 3.98 (0.71) 3.97 (0.69) 13.21 < .001 1.39
Equal status 4.04 (0.86) 3.84 (0.79) 3.94 (0.83) 10.85 < .001 1.14
Cooperation 3.94 (0.84) 3.98 (0.60) 3.96 (0.73) 12.37 < .001 1.30
Goal achievement 3.85 (0.88) 3.88 (0.73) 3.87 (0.81) 10.15 < .001 1.07
Authority support 3.98 (0.98) 4.26 (0.82) 4.11 (0.91) 11.65 < .001 1.23

Note. Scores range from 1 (low perception of condition) to 5 (high perception of condition). t statistics compare the overall mean to the
midpoint of the scale (midpoint = 3).

Table 2.  Pearson’s correlations of condition and loadings of the emotions: frightened (.97), scared
dependent variables. (.96), threatened (.95), and dangerous (.76); the
1 2 3 4 5 anger factor contained high loadings of the emo-
tions: angry (.92), irritated (.78), and aggravated
1. Conditiona -   (.68); and the pity factor contained high loadings
2. Fear −.39*** -   of the emotions: sympathy (.72), concern (.69),
3. Anger −.29*** .61*** -   and pity (.67). No cross-loadings greater than .25
4. Pity −.23** .29*** .15 -   were observed.
5. Stereotyping −.33*** .55*** .50*** .16 - Finally, a Pearson’s correlation analysis was
6. Social distance −.18* .56*** .46*** .19* .56*** conducted between condition and the dependent
aCondition: 0 = control (i.e., intragroup E-contact and no variables. As displayed in Table 2, the intergroup
contact), 1 = intergroup E-contact. contact condition was significantly related to
*p < .05. **p < .01. ***p < .001. reduced fear, anger, pity, stereotyping, and social
distance compared to the two control conditions
conducted to compare participants’ ratings of (i.e., intragroup E-contact and no contact).
each of Allport’s (1954) facilitating contact con- Additionally, the majority of the outcome varia-
ditions, averaged over E-contact condition, to the bles were significantly and positively correlated,
midpoint (= 3) on the scale.4 As displayed in with the exception of pity, which did not corre-
Table 1, the results indicated that all of these con- late significantly with anger or stereotyping.
ditions were perceived to be satisfied. Moreover,
there was no difference between the intergroup
and intragroup conditions in any of these ratings
Main Analyses
(all ps > .05). First, we conducted a series of one-way
Second, to confirm that fear, anger, and pity ANOVAs on each of the dependent variables to
were distinct group-targeted emotions, the 10 determine if there was any significant difference
items were submitted to a factor analysis using between the three experimental conditions. The
principal axis factoring and promax rotation. The results revealed a significant difference between
Kaiser–Meyer–Olkin measure of sampling ade- the conditions on fear, F(2, 130) = 11.37, p <
quacy was .84 and Bartlett’s test of sphericity was .001, ηp2 = .15; anger, F(2, 130) = 6.18, p = .003,
significant, χ2 (120) = 2479.82, p < .001, indicat- ηp2 = .09; pity, F(2, 130) = 5.94, p = .003, ηp2 =
ing that the data were suitable for factor analysis. .08; and stereotyping, F(2, 130) = 8.12,
Fear, anger, and pity accounted for 50.1%, 18.3%, p < .001, ηp2 = .11. However, there was no sig-
and 10.9% of the observed variance, respectively. nificant difference on social distance, F(2, 130)
As expected, the fear factor contained high = 2.32, p = .101, ηp2 = .03.
8 Group Processes & Intergroup Relations 00(0)

Table 3.  Descriptive statistics including means (and standard deviations) for measures by condition.

E-contact
No contact
Measures Possible range Intergroup Intragroup
n = 43
n = 47 n = 43
Fear 1–7 2.24a (1.29) 3.37b (1.36) 3.38b (1.28)
Anger 1–7 1.93a (0.96) 2.60b (1.01) 2.57b (1.10)
Pity 1–7 4.85a (1.08) 5.51b (0.87) 5.09a (0.76)
Stereotyping 1–5 2.01a (0.56) 2.34b (0.42) 2.37b (0.42)
Social distance 1–4 2.21 (0.60) 2.46 (0.59) 2.40 (0.51)

Note. Higher scores = greater fear, anger, pity, stereotyping, or social distance. Means in the same row that do not share a
subscript are significantly different at p < .05. Subscripts are not shown on social distance as the nonsignificant one-way
ANOVA on this variable precluded us from conducting contrasts.

Planned orthogonal contrasts were used to stereotypes to be related to social distance (e.g.,
compare the three conditions on each intergroup West et al., 2014), a direct path from stereotyping
emotion and stereotyping. These results are dis- to social distance was also considered. Similar to
played in Table 3. We began by comparing the the analyses conducted before, the experimental
intergroup E-contact condition to the two con- conditions were included as a set of two orthogo-
trols together (i.e., intragroup E-contact and no- nally coded contrasts; the contrast coefficients
contact conditions). Consistent with hypotheses, were constructed such that the regression coeffi-
compared to the controls, intergroup E-contact cients were readily interpretable as the relevant
significantly reduced fear, t(130) = −4.77, p < effects. The contrasts compared the intergroup
.001, d = 0.87; anger, t(130) = −3.51, p < .001, d E-contact condition to the two controls (i.e., intra-
= 0.65; and stereotyping, t(130) = −4.02, p < group E-contact and no-contact conditions; .66,
.001, d = 0.70. On these dependent measures, no −.33, −.33); and the two controls to each other (0,
significant differences were observed between .5, −.5). This model (Model A; see Figure 1) was
the two control conditions (all ps > .610). estimated using the PROCESS macro for SPSS
Unexpectedly, however, compared to the con- (Hayes, 2018), with 10,000 bootstrap samples and
trols, intergroup E-contact also reduced pity, 95% percentile bootstrap confidence intervals
t(130) = −2.69, p = .008, d = 0.46. Pity was also (CI). The mediators and dependent variables were
lower in the no-contact control, t(130) = 2.16, standardized prior to estimating the model.
p = .033, d = 0.52, than in the intragroup The indirect effects for Model A are displayed
E-contact condition. To probe this further, pity in in Table 4. As predicted, both specific indirect
the intergroup E-contact condition was compared effects of intergroup E-contact (vs. the two con-
to the two controls separately using Bonferroni trols) on social distance that serially mediated
corrections. Pity was significantly lower in the through fear and stereotyping, and anger and ste-
intergroup E-contact condition than in the intra- reotyping, were significant. The specific indirect
group E-contact condition, t(130) = −3.41, p < effect of E-contact on social distance through
.001, d = 0.68; but the intergroup E-contact con- fear only, was also significant. Contrary to predic-
dition did not significantly differ from the no-con- tions, however, the specific indirect effect of
tact condition, t(130) = −1.21, p = .229, d = 0.25. intergroup E-contact (vs. the two controls) on
Next, a path model was developed to investi- social distance through pity and stereotyping was
gate the role of fear, anger, and pity as parallel not significant. All other indirect effects of inter-
mediators of the effect of E-contact on stereotyp- group E-contact (vs. the two controls) on social
ing and social distance. As past research has shown distance where only one variable (i.e., anger, pity,
Maunder et al. 9

Figure 1.  Standardized path model (SE in parentheses) for Model A representing the effect of E-contact on
stereotyping and social distance mediated by fear, anger, and pity toward people with schizophrenia.
Note. In the first stage of the model (i.e., the direct effect of condition on intergroup emotions, stereotyping, or social dis-
tance), the regression coefficients above the path represent the effect of intergroup E-contact condition versus the two con-
trol conditions (i.e., intragroup E-contact and no-contact conditions), and the regression coefficients below the path represent
the effect of intragroup E-contact versus no contact.
*p < .05. **p < .01. ***p < .001.

or stereotyping) functioned as a mediator were intergroup emotion (i.e., anger or pity) functioned
not significant. Moreover, all the indirect effects as a mediator, were significant. Furthermore, all
of the control conditions (intragroup E-contact the indirect effects of the control conditions
and no-contact conditions) on social distance (intragroup E-contact and no-contact conditions)
were also not significant. on social distance were also not significant.
Despite finding partial support for our media-
tional hypothesis, an alternate mediation model
Discussion
(Model B; see Figure 2) was tested, where the order
of the intergroup emotions and stereotyping were The present study aimed to investigate the effec-
reversed, with stereotyping and the three inter- tiveness of a contemporary intergroup contact
group emotions—still operating in parallel—seri- intervention to reduce stigma against people with
ally mediated the E-contact effect on social schizophrenia. By utilizing a preprogrammed
distance. The indirect effects for Model B are dis- outgroup member, as well as two control condi-
played in Table 5. In this revised model, only the tions, the study represents a highly controlled
specific indirect effect of intergroup E-contact (vs. experimental test of the effect of intergroup
the two controls) on social distance serially medi- E-contact on mental health stigma. Additionally,
ated through stereotyping and fear was significant. we were able to investigate the role of intergroup
Additionally, the specific indirect effects of emotions in the relationship between contact and
E-contact on social distance through stereotyping stigma. In the domain of mental health stigma,
only, and fear only, were also significant. No other researchers have previously limited their focus to
indirect effects from intergroup E-contact (vs. the negative emotions (Stathi et al., 2012; West et al.,
two controls) on social distance, where only one 2014), or have amalgamated distinct group-targeted
10 Group Processes & Intergroup Relations 00(0)

Table 4.  Indirect effects for Model A.

Independent variable Mediator variable(s) Dependent variable β (SE) 95% CI


Intergroup E-contact  
(vs. controlsa) →
  Fear → Stereotyping −0.28 (0.10) [−0.49, −0.11]
  Anger → Stereotyping −0.16 (0.07) [−0.31, −0.03]
  Pity → Stereotyping −0.00 (0.04) [−0.08, 0.08]
  Fear → Social distance −0.26 (0.10) [−0.49, −0.10]
  Anger → Social distance −0.06 (0.06) [−0.19, 0.04]
  Pity → Social distance −0.02 (0.04) [−0.10, 0.06]
  Stereotyping → Social distance −0.09 (0.07) [−0.24, 0.02]
  Fear → Stereotyping → Social distance −0.10 (0.04) [−0.20, −0.03]
  Anger → Stereotyping → Social distance −0.06 (0.03) [−0.13, −0.01]
  Pity → Stereotyping → Social distance 0.00 (0.02) [−0.03, 0.03]
Intragroup E-contact  
(vs. no contact) →
  Fear → Stereotyping 0.00 (0.07) [−0.14, 0.16]
  Anger → Stereotyping −0.01 (0.06) [−0.13, 0.11]
  Pity → Stereotyping −0.00 (0.04) [−0.08, 0.09]
  Fear → Social distance 0.00 (0.07) [−0.13, 0.15]
  Anger → Social distance 0.00 (0.03) [−0.07, 0.06]
  Pity → Social distance −0.01 (0.04) [−0.09, 0.06]
  Stereotyping → Social distance 0.03 (0.06) [−0.09, 0.16]
  Fear → Stereotyping → Social distance 0.00 (0.03) [−0.05, 0.06]
  Anger → Stereotyping → Social distance 0.00 (0.02) [−0.05, 0.04]
  Pity → Stereotyping → Social distance 0.00 (0.02) [−0.03, 0.03]

Note. Boldfaced effects are significant.


aControls represent the intragroup E-contact and no-contact conditions.

emotions into a single affective variable (Giacobbe (rather than increased) following intergroup
et al., 2013). By adopting a sociofunctional E-contact, but only in comparison to intragroup
approach to understanding stigma (Cottrell & E-contact. These findings are the first to demon-
Neuberg, 2005) that emphasizes the need to distin- strate the distinct roles of anger and fear in the
guish between positive and negative emotions, the context of intergroup contact with a person with
current study contributes greater depth to mental schizophrenia, and they provide the first causal
health stigma reduction research. evidence to support the benefits of E-contact as a
Consistent with expectations, intergroup contact strategy to improve several intergroup
E-contact reduced fear, anger, and stereotyping outcomes related to mental health stigma.
compared to both intragroup E-contact and no
contact. The data also suggested that stereotyping The E-Contact Effect and the
and group-targeted fear and anger played an
important role in explaining the E-contact effect
Role of Intergroup Emotions
on reduced social distance toward people with The current findings contribute to the small, but
schizophrenia. Unexpectedly, however, pity did growing, body of contact research that has specifi-
not emerge as a key affective mediator of the cally targeted the stigma toward schizophrenia
E-contact effect. Specifically, participants’ feelings (e.g., Giacobbe et al., 2013; Stathi et al., 2012; West
of pity toward people with schizophrenia decreased et al., 2014) by demonstrating for the first time the
Maunder et al. 11

Figure 2.  Standardized path model (SE in parentheses) for Model B representing the effect of E-contact on
social distance mediated by stereotyping as well as fear, anger, and pity toward people with schizophrenia.
Note. In the first stage of the model (i.e., the direct effect of condition on intergroup emotions, stereotyping, or social dis-
tance), the regression coefficients above the path represent the effect of intergroup E-contact condition versus the two con-
trol conditions (i.e., intragroup E-contact and no-contact conditions), and the regression coefficients below the path represent
the effect of intragroup E-contact vs. no contact.
*p < .05. **p < .01. ***p < .001.

effectiveness of computer-mediated interactions elicit fear and anger, respectively (Corrigan et al.,
for reducing fear, anger, and stereotyping toward 2003; Cottrell & Neuberg, 2005). Past research has
people with schizophrenia. In contrast to educa- shown that counterstereotypical information
tional approaches to reducing mental health either separate from (Ramasubramanian, 2011) or
stigma, which directly disconfirm inaccurate stere- integrated into (Dunaev, Brochu, & Markey, 2018)
otypes and replace them with factual information an intergroup interaction has beneficial effects on
(Corrigan et al., 2012), we did not explicitly target stigma in comparison to stereotypical information.
participants’ prior knowledge of people with Thus, if the interaction partner in the present
schizophrenia during the E-contact intervention. study had exhibited characteristics stereotypical of
However, by getting to know an outgroup mem- people with schizophrenia, we may not have
ber, intergroup contact is believed to similarly observed the same effects.
reduce stigma by challenging people’s beliefs about It is currently unclear whether intergroup con-
the outgroup (Allport, 1954; Amichai-Hamburger tact was solely responsible for the reduction in
& McKenna, 2006). In the current study, partici- stigma, or whether this effect could also be attrib-
pants met and worked with a person with schizo- uted to exposing participants to stereotype-discon-
phrenia who was not aggressive or unpredictable, firming information in the form of a
that is, a counterstereotypical group member. In counterstereotypical outgroup member. In this
doing so, the E-contact intervention may have regard, however, Corrigan et al. (2002) showed that
subtly disconfirmed the belief that people with contact targeting specific stereotypes about people
schizophrenia are dangerous and unable to follow with schizophrenia is more effective at reducing
social norms, characteristics which are known to stigma than educational interventions targeting the
12 Group Processes & Intergroup Relations 00(0)

Table 5.  Indirect effects for Model B.

Independent variable Mediator variable(s) Dependent variable β (SE) 95% CI


Intergroup E-contact  
(vs. controlsa) →
  Stereotyping → Fear −0.33 (0.10) [−0.53, −0.15]
  Stereotyping → Anger −0.31 (0.10) [−0.51, −0.14]
  Stereotyping → Pity −0.07 (0.07) [−0.24, 0.05]
  Stereotyping → Social distance −0.25 (0.09) [−0.45, −0.09]
  Fear → Social distance −0.16 (0.08) [−0.34, −0.04]
  Anger → Social distance −0.03 (0.03) [−0.11, 0.02]
  Pity → Social distance −0.01 (0.03) [−0.09, 0.05]
  Stereotyping → Fear → Social distance −0.11 (0.05) [−0.21, −0.03]
  Stereotyping → Anger → Social distance −0.03 (0.03) [−0.10, 0.02]
  Stereotyping → Pity → Social distance 0.00 (0.01) [−0.03, 0.01]
Intragroup E-contact  
(vs. no contact) →
  Stereotyping → Fear 0.03 (0.09) [−0.14, 0.20]
  Stereotyping → Anger 0.03 (0.08) [−0.13, 0.20]
  Stereotyping → Pity 0.01 (0.02) [−0.05, 0.06]
  Stereotyping → Social distance 0.03 (0.07) [−0.10, 0.16]
  Fear → Social distance −0.01 (0.06) [−0.13, 0.12]
  Anger Social distance −0.01 (0.03) [−0.07, 0.05]
  Pity Social distance −0.02 (0.04) [−0.09, 0.06]
  Stereotyping → Fear → Social distance 0.01 (0.03) [−0.04, 0.08]
  Stereotyping → Anger → Social distance 0.00 (0.01) [−0.02, 0.03]
  Stereotyping → Pity → Social distance 0.00 (0.00) [−0.00, 0.01]

Note. Boldfaced effects are significant.


aControls represent the intragroup E-contact and no-contact conditions.

same stereotypes. Nevertheless, future studies counterstereotypical characteristics of the out-


should address the issue of outgroup member ste- group member in the current study. It has been
reotypicality (for a recent discussion of this issue, argued that pity stems from the perception that
see McIntyre, Paolini, & Hewstone, 2016) by exam- the outgroup is afflicted by an uncontrollable situ-
ining the effectiveness of E-contact with more ste- ation (Cottrell & Neuberg, 2005). In the current
reotypical outgroup members, as well as by study, the outgroup member disclosed that they
comparing contact with counterstereotypical group were studying psychology to learn more about
members to counterstereotypical group members managing their diagnosis. This information may
that are merely described. have suggested that the outgroup member had a
In the current study and contrary to expecta- degree of control over their condition, which in
tions, pity toward people with schizophrenia was turn, may have reduced participants’ feelings of
also reduced following E-contact, although only pity (Corrigan et al., 2003). Seger et al. (2017) have
in comparison to the intragroup E-contact condi- also argued that pity is not an unambiguously pos-
tion. This finding is inconsistent with previous itive emotion because, despite facilitating helping
research that has found higher levels of contact behaviors, it implies that the target outgroup is
with people with mental illness to predict increased lower in status than the ingroup. Therefore,
pity (Corrigan et al., 2003). Nevertheless, this reducing pity is not necessarily inconsistent with
unexpected finding may also be explained by the reducing stigma against schizophrenia.
Maunder et al. 13

It should be reemphasized that pity was only Stephan, 1985). For example, evaluating a group
reduced in the intergroup contact condition in as dangerous could be a necessary precondition to
comparison to the intragroup contact condition fearing that group (Cottrell & Neuberg, 2005).
and not the no-contact condition. This finding However, numerous past studies have placed
suggests that there may be something particular emotions first in the causal order, both when spe-
about intragroup contact that increases pity rela- cifically examining the effect of contact on stereo-
tive to the two other conditions. In the present typing against schizophrenia (e.g., Stathi et al.,
study, the ingroup member expressed a desire to 2012; West et al., 2014) and when observing preju-
learn more about mental illness, which may have dice against other stigmatized groups (e.g.,
framed this group as one in need of help and Pettigrew & Tropp, 2008). Taken together, it is
understanding, thereby increasing pity in this con- likely that intergroup emotions and stereotyping
dition. Nevertheless, such a result reinforces the are cyclically related; thus, both the originally
need for both intragroup and no-contact control hypothesized and the alternate model may be
conditions to be utilized when investigating the equally important in elucidating the E-contact
effects of intergroup contact for stigma reduc- effect. One way to further investigate these find-
tion. Moreover, over and above the positive rela- ings might be to manipulate intergroup emotions
tionships between fear and anger on social and/or outgroup stereotypes experimentally,
distance, reduced pity did not independently pre- rather than measuring them cross-sectionally.
dict increased social distance. This finding may Finally, the intergroup emotions considered in
suggest that, while distinct perceived threats are the present study did not comprise an exhaustive
important for eliciting different affective responses list. Although it is intuitive to expect fear, anger,
toward people with schizophrenia, pity does not and pity to be the most relevant emotions elicited
necessarily play an important mediating role for by people with schizophrenia (Corrigan et al.,
the E-contact effect. Additional research is now 2003), researchers are encouraged to consider the
needed to further substantiate this hypothesis. mediating role of other group-targeted emotions,
Consistent with the sociofunctional approach such as admiration and contempt. One way for-
to prejudice (Cottrell & Neuberg, 2005), the cur- ward may be to examine intergroup contact with
rent findings highlight the important role of inter- people with other mental illnesses that may be
group emotions in stigma reduction. It was associated with different perceived threats and
through the group-targeted emotions of fear and affective profiles (Cottrell & Neuberg, 2005). For
anger that negative stereotypes about and social example, fear and anger may be weaker mediators
distance toward people with schizophrenia were of the E-contact effect when the outgroup mem-
reduced. However, this interpretation of the find- ber is diagnosed with major depression. Compared
ings should only be made in light of the following to people diagnosed with schizophrenia, people
consideration. An alternative mediation model, who suffer from depression are less likely to be
where stereotyping and intergroup emotions were perceived as dangerous and dependent on others
positioned as serial mediators of the E-contact (Angermeyer & Matschinger, 2003), and therefore,
effect on social distance, was also found to fit the are less likely to elicit threat perceptions related to
data equally well. In this model, compared to both fear and anger (Cottrell & Neuberg, 2005).
control conditions, intergroup E-contact was
found to reduce the negative stereotypes about Strengths, Limitations, and
people with schizophrenia, which in turn reduced
Future Directions
participants’ intergroup feelings of fear (but not
anger or pity), and which in turn reduced their In regard to the E-contact intervention, it is
desire for social distance. This alternative finding promising that a brief online interaction was suc-
is consistent with past literature that has argued cessful at reducing stigma against people with
for intergroup cognitions, such as stereotypes, to schizophrenia. Many intergroup contact inter-
causally precede intergroup emotions (Stephan & ventions aimed at reducing mental health stigma,
14 Group Processes & Intergroup Relations 00(0)

in particular contact interventions that occur One limitation that should be considered when
face-to-face, require significantly more resources interpreting the present results is the use of an
than the computer-mediated intervention used in undergraduate student sample, which typically
the current study. Thus, there is potential for limits the generalizability of the findings to the
E-contact to be a practical and cost-effective general population. However, it should also be
stigma-reduction strategy. However, to have con- noted that the sample consisted of undergraduate
fidence in such a strategy, it is imperative that psychology students, who are more likely to be
future research first determine the long-term responsible for the care of people with mental ill-
consequences of E-contact in the context of ness in the future than individuals from the gen-
mental health stigma. Previously, the beneficial eral population. In their review of the extant
effects of an eight-session E-contact program on literature, Henderson et al. (2014) found evidence
interreligious bias between Muslims and to suggest that mental health care professionals
Christians were found to endure for 12 months often report similar levels of stigma toward peo-
postintervention (White, Abu-Rayya, & Weitzel, ple with mental illness compared to the general
2014). In a similar vein, we contend that the long- population. Additionally, many individuals with
term benefits of E-contact are likely to rely on mental illness report experiencing stigma when
participants experiencing multiple interactions seeking treatment, which can discourage adher-
with the outgroup, rather than a single online ence to the treatment program (Mestdagh &
interaction. Future research is now needed to Hansen, 2014). Therefore, there is a need to
examine the longevity of the E-contact effect on reduce stigma in students enrolled in health-care-
the stigma towards mental illness. related subjects before they become responsible
There is an opportunity for E-contact to for the care and mental well-being of others.
evolve further, by incorporating video and voice Nevertheless, future studies should investigate the
communication, as well as other emerging tech- use of E-contact for stigma reduction in commu-
nologies such as virtual and augmented reality. nity samples as well. Now that we have established
Contact via these channels, in comparison to text- the basic effect of E-contact on stigma against
only communication, would heighten the richness people with schizophrenia using a tightly con-
of the contact experience as well as participants’ trolled experimental design, further investigation
self-involvement in the interaction (Harwood, is required to improve the ecological validity of
2010). In the present study, we chose to use text- this research. Brief, highly structured, text-only,
only E-contact since (a) it has been empirically and supervised online interactions with a prepro-
validated and shown to be an effective strategy for grammed outgroup member may not reflect inter-
improving outgroup attitudes, (b) the supporting actions with real people with schizophrenia
technology is practical and readily available, (c) outside of the laboratory. Although this proce-
participants have psychological control over how dure afforded the researchers a high degree of
they present themselves, (d) the outgroup mem- control over the experiment, including ensuring
ber’s nonverbal characteristics (e.g., appearance that the outgroup member disclosed their stigma-
and behavior) cannot be used to bias a partici- tized identity, it is essential that future research
pant’s evaluation of them, and (e) the real-time explore the impact of an unsupervised E-contact
chat structure of the medium engages participants interaction with real people with schizophrenia.
in the intergroup interaction. However, future
research should compare these alternative contact
Conclusion
strategies, and also explore whether text-only con-
tact could be a productive preparatory step that The present study provides new evidence to sup-
acclimatizes participants to contact with outgroup port the benefits of intergroup contact for stigma
members prior to engaging in other contact expe- reduction by investigating a modern avenue for
riences (for a similar argument, see White, Harvey, contact with people with schizophrenia. Given
& Abu-Rayya, 2015). the increasing focus on web-based interventions
Maunder et al. 15

for promoting mental health literacy (Corrigan specific to people with schizophrenia.
et al., 2012), and the increase in the use of the 2. Originally, we also manipulated the duration of
Internet for communication (White, Harvey, & the E-contact interaction. However, E-contact
Abu-Rayya, 2015), we encourage researchers to duration did not significantly moderate the
E-contact effect on any of the dependent vari-
examine the contemporary role of computer-
ables. Therefore, for the sake of brevity, we have
mediated communication, such as E-contact, for
omitted a discussion of this independent variable.
reducing mental health stigma. Additionally, we 3. A reviewer suggested that the supervised nature of
are hopeful that these findings will encourage the interaction may have encouraged participants
researchers to further investigate the role of dis- to adhere to social norms of respect and polite-
tinct, negative and positive, group-targeted emo- ness during the interaction and when reporting
tions in the context of mental health stigma. their attitudes towards people with schizophrenia.
However, our two control conditions allow us to
Acknowledgements rule out this possibility. First, participants in the no-
contact control condition did not experience super-
We would like to thank Stephen Rogers from the School
vision in an interaction, and yet reported equivalent
of Psychology at The University of Sydney for program-
levels of fear, anger, and stereotyping as partici-
ming the E-contact tool. We would also like to express
pants in the intragroup contact condition who did
our gratitude to two anonymous reviewers for their help-
experience supervision. Second, participants in the
ful comments on a previous version of this manuscript.
intergroup and intragroup E-contact conditions
experienced identical interactions, which is further
Funding evidenced by their equivalent ratings of Allport’s
This research received no specific grant from any fund- (1954) facilitating conditions. Thus, the signifi-
ing agency in the public, commercial, or not-for-profit cant difference between the intergroup E-contact
sectors. condition and the two control conditions on the
dependent variables cannot be attributed to the
supervised nature of the interaction.
Notes 4. In their meta-analysis, Pettigrew and Tropp (2006)
1. There exists a large body of research that has found a stronger effect of intergroup contact on
focused on the effects of intergroup contact stigma when Allport’s (1954) facilitating contact
with “a person with a mental illness” on atti- conditions were satisfied. To explore whether
tudes and behaviors toward “people with a men- perceptions of Allport’s conditions moderated
tal illness” (for a recent meta-analytic review, the effect of intergroup E-contact on stigma in
see Corrigan et al., 2012). Other research has our study, we conducted separate moderated hier-
focused on the effects of intergroup contact archical regression analyses with fear, anger, pity,
with “a person with a mental illness” on atti- stereotyping, and social distance as the depend-
tudes and behaviors toward “people with schizo- ent variables. Prior to the analysis, we created a
phrenia” (Angermeyer & Matschinger, 2004). single index of Allport’s facilitating conditions
However, given the significant differences in by averaging participants’ ratings of perceived
stereotypes about different diagnostic catego- positivity, equal status, cooperation, common
ries (Angermeyer & Matschinger, 2003), contact goal, and authority support. E-contact condi-
measured in relation to the superordinate cat- tion and the index of Allport’s conditions were
egory of “a person with a mental illness” may be entered in the first step, followed by the two-way
inappropriate for understanding the effects of interaction. The results demonstrated that, over
intergroup contact on stereotyping toward a spe- and above the effect of E-contact, increased per-
cific diagnosis. This lack of attention to specificity ceptions of Allport’s facilitating conditions were
has also made it difficult to evaluate the mediat- negatively related to fear (β = −0.20, p = .039),
ing role of different intergroup emotions, which anger (β = −0.55, p = .006), stereotyping (β =
are hypothesized to vary depending on the target −0.26, p = .013), and social distance (β = −0.30,
group (Cottrell & Neuberg, 2005). Therefore, we p = .014), but positively related to pity (β = 0.48,
have restricted our review to empirical studies p = .019). Interestingly, the interaction between
that have examined both contact and outcomes E-contact condition and perceptions of Allport’s
16 Group Processes & Intergroup Relations 00(0)

conditions did not significantly explain additional review of quantitative and qualitative studies.
variance across all dependent variables (all ΔR2s < Psychological Medicine, 45, 11–27. doi:10.1017/
.01 and ps > .25). Together, these results suggest S0033291714000129
that perceptions of Allport’s conditions may not Corrigan, P. W., Markowitz, F. E., Watson, A., Rowan,
necessarily facilitate the E-contact effect, despite D., & Kubiak, M. A. (2003). An attributional
their unique relationship with improved intergroup model of public discrimination towards persons
outcomes. Nevertheless, it is possible that these with mental illness. Journal of Health and Social
findings are related to the outgroup examined in Behavior, 44, 162–179. doi:10.2307/1519806
the current study. Past research has found that few Corrigan, P. W., Morris, S. B., Michaels, P. J., Rafacz,
people report a history of contact with people J. D., & Rüsch, N. (2012). Challenging the pub-
with schizophrenia (e.g., West et al., 2014). Under lic stigma of mental illness: A meta-analysis of
these circumstances, merely meeting a member of outcome studies. Psychiatric Services, 63, 963–973.
the outgroup may be sufficient for stigma reduc- doi:10.1176/appi.ps.201100529
tion without the need to perceive the interaction Corrigan, P. W., Rowan, D., Green, A., Lundin, R.,
as collaborative and goal-oriented. Consequently, River, P., Uphoff-Wasowski, K., . . . Kubiak, M.
it is possible that Allport’s facilitating conditions A. (2002). Challenging two mental illness stigmas:
will only be useful in situations when past con- Personal responsibility and dangerousness. Schizo-
tact is more frequent or when relations between phrenia Bulletin, 28, 293–309. doi:10.1093/oxford-
the groups are marred by a history of intergroup journals.schbul.a006939
conflict. Cottrell, C. A., & Neuberg, S. L. (2005). Different
emotional reactions to different groups: A soci-
ORCID iDs ofunctional threat-based approach to “prejudice.”
Journal of Personality and Social Psychology, 88, 770–
Rachel D. Maunder https://orcid.org/0000-0002-
789. doi:10.1037/0022-3514.88.5.770
8436-691X
Dunaev, J. L., Brochu, P. M., & Markey, C. H. (2018).
Stefano Verrelli https://orcid.org/0000-0002-
Imagine that! The effect of counterstereotypic
4268-7061
imagined intergroup contact on weight bias. Health
Psychology, 37, 81–88. doi:10.1037/hea0000545
References Farrelly, S., Jeffery, D., Rüsch, N., Williams, P.,
Allport, G. W. (1954). The nature of prejudice. Reading, Thornicroft, G., & Clement, S. (2015). The link
MA: Addison Wesley-Brown. between mental health-related discrimination
Amichai-Hamburger, Y., & McKenna, K. Y. A. (2006). and suicidality: Service user perspectives. Psy-
The contact hypothesis reconsidered: Interact- chological Medicine, 45, 2013–2022. doi:10.1017/
ing via the Internet. Journal of Computer-Mediated S0033291714003158
Communication, 11, 825–843. doi:10.1111/j.1083- Giacobbe, M. R., Stukas, A. A., & Farhall, J. (2013).
6101.2006.00037.x The effects of imagined versus actual contact with
Angermeyer, M. C., & Matschinger, H. (2003). Pub- a person with a diagnosis of schizophrenia. Basic
lic beliefs about schizophrenia and depression: and Applied Social Psychology, 35, 265–271. doi:10.1
Similarities and differences. Social Psychiatry and 080/01973533.2013.785403
Psychiatric Epidemiology, 38, 526–534. doi:10.1007/ Harwood, J. (2010). The contact space: A novel frame-
s00127-003-0676-6 work for intergroup contact research. Journal of
Angermeyer, M. C., & Matschinger, H. (2004). The Language and Social Psychology, 29, 147–177. doi:10.
stereotype of schizophrenia and its impact on 1177/0261927X09359520
discrimination against people with schizophrenia: Hayes, A. F. (2018). Introduction to mediation, modera-
Results from a representative survey in Germany. tion, and conditional process analysis: A regression-based
Schizophrenia Bulletin, 30, 1049–1061. doi:10.1093/ approach (2nd ed.). New York, NY: Guilford Press.
oxfordjournals.schbul.a007120 Henderson, C., Noblett, J., Parke, H., Clement, S., Caf-
Clement, S., Schauman, O., Graham, T., Maggioni, F., frey, A., Gale-Grant, O., . . . Thornicroft, G. (2014).
Evans-Lacko, S., Bezborodovs, N., . . . Thorni- Mental health-related stigma in health care and
croft, G. (2015). What is the impact of mental mental health-care settings. The Lancet: Psychiatry,
health-related stigma on help-seeking? A systematic 1, 467–482. doi:10.1016/S2215-0366(14)00023-6
Maunder et al. 17

Link, B. G. (1987). Understanding labelling effects in Stephan, W. G., & Stephan, C. W. (1985). Inter-
the area of mental disorders: An assessment of the group anxiety. Journal of Social Issues, 41, 157–175.
effects of expectations of rejection. American Soci- doi:10.1111/j.1540-4560.1985.tb01134.x
ological Review, 52, 96–112. doi:10.1086/228672 West, K., Hewstone, M., & Lolliot, S. (2014). Inter-
Link, B. G., & Phelan, J. C. (2001). Conceptualizing group contact and prejudice against people with
stigma. Annual Review of Sociology, 27, 363–385. schizophrenia. The Journal of Social Psychology, 154,
doi:10.1146/annurev.soc.27.1.363 217–232. doi:10.1080/00224545.2014.888327
McIntyre, K., Paolini, S., & Hewstone, M. (2016). West, K., Holmes, E., & Hewstone, M. (2011). Enhancing
Changing people’s views of outgroups through imagined contact to reduce prejudice against people
individual-to-group generalisation: Meta-analytic with schizophrenia. Group Processes & Intergroup Rela-
reviews and theoretical considerations. European tions, 14, 407–428. doi:10.1177/1368430210387805
Review of Social Psychology, 27, 63–115. doi:10.1080/ White, F. A., & Abu-Rayya, H. M. (2012). A dual
10463283.2016.1201893 identity-electronic contact (DIEC) experiment
Mestdagh, A., & Hansen, B. (2014). Stigma in patients promoting short- and long-term intergroup har-
with schizophrenia receiving community mental mony. Journal of Experimental Social Psychology, 48,
health care: A review of qualitative studies. Social 597–608. doi:10.1016/j.jesp.2012.01.007
Psychiatry and Psychiatric Epidemiology, 49, 79–87. White, F. A., Abu-Rayya, H. M., Bliuc, A.-M., & Faulkner,
doi:10.1007/s00127-013-0729-4 N. (2015). Emotion expression and intergroup bias
O’Reilly, C. L., Bell, S., & Chen, T. F. (2010). Con- reduction between Muslims and Christians: Long-
sumer-led mental health education for pharmacy term Internet contact. Computers in Human Behavior,
students. American Journal of Pharmaceutical Educa- 53, 435–442. doi:10.1016/j.chb.2015.04.074
tion, 74, 167. doi:10.5688/aj7409167 White, F. A., Abu-Rayya, H. M., & Weitzel, C. (2014).
Pettigrew, T. F., & Tropp, L. R. (2006). A meta-ana- Achieving twelve-months of intergroup bias
lytic test of intergroup contact theory. Journal reduction: The dual identity-electronic contact
of Personality and Social Psychology, 90, 751–783. (DIEC) experiment. International Journal of Inter-
doi:10.1037/0022-3514.90.5.751 cultural Relations, 38, 158–163. doi:10.1016/j.ijin-
Pettigrew, T. F., & Tropp, L. R. (2008). How does inter- trel.2013.08.002
group contact reduce prejudice? Meta-analytic White, F. A., Harvey, L. J., & Abu-Rayya, H. M. (2015).
tests of three mediators. European Journal of Social Improving intergroup relations in the Internet
Psychology, 38, 922–934. doi:10.1002/ejsp.504 age: A critical review. Review of General Psychology,
Ramasubramanian, S. (2011). The impact of stereotypi- 19, 129–139. doi:10.1037/gpr0000036
cal versus counterstereotypical media exemplars White, F. A., Harvey, L. J., & Verrelli, S. (2015). Includ-
on racial attitudes, causal attributions, and sup- ing both voices: A new bidirectional framework
port for affirmative action. Communication Research, for understanding and improving intergroup
38, 497–516. doi:10.1177/0093650210384854 relations. Australian Psychologist, 50, 421–433.
Schulze, B., & Angermeyer, M. C. (2003). Subjective doi:10.1111/ap.12108
experiences of stigma. A focus group study of White, F. A., Turner, R. N., Verrelli, S., Harvey, L. J.,
schizophrenia patients, their relatives and mental & Hanna, J. R. (in press). Improving intergroup
health professionals. Social Science and Medicine, 56, relations between Catholics and Protestants in
299–312. doi:10.1016/S0277-9536(02)00028-X Northern Ireland via E-contact. European Journal
Seger, C. R., Banerji, I., Park, S. H., Smith, E. R., & of Social Psychology. doi:10.1002/ejsp.2515
Mackie, D. M. (2017). Specific emotions as medi- White, F. A., Verrelli, S., Maunder, R. D., & Kervinen,
ators of the effect of intergroup contact on preju- A. (in press). Using electronic contact to reduce
dice: Findings across multiple participant and homonegative attitudes, emotions, and behavioral
target groups. Cognition and Emotion, 31, 923–936. intentions among heterosexual men and women:
doi:10.1080/02699931.2016.1182893 A contemporary extension of the contact hypoth-
Stathi, S., Tsantila, K., & Crisp, R. J. (2012). Imagin- esis. The Journal of Sex Research. doi:10.1080/00224
ing intergroup contact can combat mental health 499.2018.1491943
stigma by reducing anxiety, avoidance, and nega- World Health Organization (WHO). (2001). World
tive stereotyping. The Journal of Social Psychology, 152, health report 2001 – Mental health: New understanding,
746–757. doi:10.1080/00224545.2012.697080 new hope. Geneva, Switzerland: Author.

You might also like