INTER AND INTRAGROUP EMOTION AND SOCIAL IDENTIFICATION: A REAL WORLD STUDY Orla T Muldoon, Niamh McNamara, Paula

Devine, Karen Trew and Lizanne Dowds

IBIS working paper no. 103

INTER AND INTRAGROUP EMOTION AND SOCIAL IDENTIFICATION: A REAL WORLD STUDY Orla T Muldoon, Niamh McNamara, Paula Devine, Karen Trew and Lizanne Dowds

Working Papers in British-Irish Studies No. 103, 2010

Institute for British-Irish Studies University College Dublin

IBIS Working Papers No. 103, 2010

the authors, 2010

ISSN 1649-0304

ABSTRACT

INTER AND INTRAGROUP EMOTION AND SOCIAL IDENTIFICATION: A REAL WORLD STUDY Emotion research has recently begun to focus on the relationship between social identities and group based emotion. The current paper reports on a large scale study of intra and intergroup emotions and their relationship to social identities in Northern Ireland. 1,179 respondents reported their explicit intragroup emotion, selfcategorised religion and strength of religious identity as well as four emotional responses to four visual primes related to them in their own homes. The valence of the prime and strength of identification were manipulate and measured respectively. Intragroup primes were found to be more strongly and significantly related to both identification and type of emotion expressed than intergroup primes. The results of the study are interpreted in terms of our understanding of appraisal and intergroup emotions theory and the practical implications for intergroup relations considered. Publication information Paper presented at the IBIS Annual Conference 2009 Protestant Traditions and the Paths to Peace: Beyond the Legacies of Plantation, Global Irish Institute, University College, Dublin, 9 June 2009. IBIS is grateful to the Department of the Taoiseach for its support in funding the conference.

BIOGRAPHICAL INFORMATION

Professor Orla Muldoon is the Foundation Chair and Head of Department at the Department of Psychology at University of Limerick. Since graduating with an undergraduate degree in psychology from Queens University Belfast in 1993 (1st Hons) and subsequently a PhD, Professor Muldoon has been engaged in teaching and research in the area of social psychology. Her overarching research interest concerns the impact of group memberships on social behaviours. Interests include how perceiving one-self as a member of a religious, socio-economic or racial group can impact on views of the self and others, how these group memberships are transmitted to and understood by children and young people and how such memberships impact on mental health. In much of her research she has used existing groups in Northern Ireland, though much of the evidence from Northern Ireland can and does converge with the wider literature on these topics. Previous appointments were as a Lecturer and then Senior Lecturer and Director of Health and Social Issues Research Cluster at Queen's University Belfast (1998-2007) and Lecturer at University of Ulster, Magee College (1996-1998). Dr Niamh McNamara graduated with a BA (Hons) in Applied Psychology from University College Cork in 2002 and was awarded a PhD in Applied Psychology from University College Cork in 2006. She worked as a researcher in the Human Factors Research Group in the Department of Applied Psychology in UCC before joining UL in January 2008. She is currently a Postdoctoral Research Fellow on the IRCHSS funded project Revitalising Communities: A Framework for Social Change. Dr Paula Devine is Research Director of ARK, based in the School of Sociology, Social Policy and Social Work, Queen's University Belfast. In addition, she is coordinator of the Northern Ireland Life and Times Survey, an annual survey recording attitudes to a range of social policy issues. Dr Karen Trew came to Belfast from London in 1961 and completed both her undergraduate and postgraduate degrees at Queens University. Her first job was with the Northern Ireland Council for Educational Research which was housed within the Psychology Department and after eight years she moved back into the Department as a lecturer. Dr Trews current research extends her long-term interest in relating academic scholarship in the area of social psychology with practice and policy especially in relation to children and young people. Dr Lizanne Dowds is an Honorary Senior Research Fellow at the University of Ulster and consultant to ARK (Access, Research Knowledge). Acknowledgements The authors are grateful for a grant awarded from the Economic and Social Research Council (UK) which funded this research (Award Number: RES-000-221938).

INTER AND INTRAGROUP EMOTION AND SOCIAL IDENTIFICATION: A REAL WORLD STUDY Orla T Muldoon, Niamh McNamara, Paula Devine, Karen Trew and Lizanne Dowds

INTRODUCTION Emotions are typically viewed in psychology as private states and research in this area is usually conducted at the level of the individual. While it is acknowledged that emotions can be social in nature, the term social typically refers to interpersonal interactions. Existing individual-level theories of emotion, in particular, appraisal theories of emotion conceptualise emotions as motivational states (Arnold, 1960). This represents a social cognition approach in emotions in terms of the individuals perception or interpretation of whether a situation/event is personally relevant. Arnold (1960) argues that we respond emotionally only when the situation is relevant to us in some way. Smith and Lazarus (1993) developed this further by stating that in addition to motivational relevance, the valence of the resulting emotion depends on the appraisal of motivational congruence (i.e. whether the situation helps or hinders the individuals concerns), and further that the specific nature of the emotion depends on secondary appraisals such as personal and social coping resources. Such coping resources can be conceptualised at both individual and group level (Haslam, 2006). Traditionally social identity approaches emphasise that people can define themselves both in personal terms and also in terms of the groups to which they belong. When a particular social identity is made salient, individuals who identify with that group see themselves as interchangeable exemplars of the group rather than as distinct individuals (Smith, 1993). Their social identity as a group member becomes an integral aspect of the self and thus the group is infused with emotional significance (Tajfel and Turner, 1979; Tajfel, 1982). In spite of the fact that identification with a group infuses that group with value and affective significance (Tajfel, 1972), emotion research has not been much concerned with the idea that people may experience emotions on behalf of others, be they in-group or outgroup members. This remains a serious research gap given the degree to which people are invested in their identities and the emotions that they can arouse. Research on group-based emotions began in earnest in the 1990s. It was postulated that people can experience emotions as group members, that these emotions are distinct from individual-level emotions, and that they can have important implications for inter-group relations. Such emotions can be directed at the in-group, outgroup, or intergroup relation. Through the process of identification, group-relevant events become self-relevant and elicit group-based emotions (Smith, 1993). This assumption lies at the core of intergroup emotions theory (Smith et al, 1993). This theory merges appraisal theories of emotion, social identity theory, and social categorisation theory in an attempt to better understand the nature of emotions that
IBIS WORKING PAPER NO. 103

IBIS WORKING PAPERS

NO. 103, 2010

arise from group identification and membership (Smith, Seger, & Mackie, 2007: 432). Initially, the focus was on emotions of in-group members directed at outgroups although recently there have been investigations of group-based emotions directed at in-groups (Smith et al, 2007). In a recent review however, Lyer and Leach (2008) identified three general principles of the field: (1) people can experience group-level emotions that are distinct from individual emotions and have important consequences for inter-group relations; (2) particular emotions are associated with specific appraisals of the emotion object be that in-group, out-group or the inter-group relation, and finally; (3) particular emotions can predict specific action tendencies towards an in-group or out-group. THE CURRENT STUDY Conceptualisation of group-based emotions is under-developed. While there is consensus among researchers on the general principles underlying group-based emotions, Lyer and Leach (2008) caution that the literature can be characterised as conceptually confused. They contend that there is evidence of widespread inconsistency in the use of terminology and operationalisations. To this end, Lyer and Leach (2008) propose a typology that differentiates between group level emotions along two dimensions, namely the subject of the emotion (someone who sees themselves as a unique individual or someone who perceives themselves as a member of a group) and the object of the emotion (in-group or out-group). They argue that there are five types of group level emotions and it is important for researchers that they clearly define which of these they are examining. The five types are: inter-group emotions; personal emotions directed at out-groups; group-based emotions directed at in-groups; personal emotions directed at in-groups; and group-based emotions directed at individuals. The vast majority of research focuses on inter-group emotions, however, because participants are often not asked to subjectively selfcategorise as in-group members and if they are it is unclear whether the selfcategorisation was salient when the emotions were measured (Lyer & Leach, 2008). It is possible therefore that what is being measured is not inter-group emotions but individual level emotions directed at groups. When assessing inter-group emotions it is important to specify inter-group relation that includes in-group as subject and out-group as object of emotion as well as assessing self-relevance. Our study accounts for these concerns. Yzerbyt et al (2002) argue that under certain conditions it is likely that people will experience emotions even though they are not directly confronted with the trigger situation. In the majority of group-based emotion studies, people are asked to selfcategorise as an in-group member or level of identification is assessed in order to make a particular social identity salient. However, identities can be made salient through other means. In a recent study, Seger, Smith and Mackie (2009) illustrate that social identities can be subtly activated through the use of priming techniques (i.e. listening to music and viewing photographs) which in turn produce group-based emotions. Consistent with their previous study (Smith et al, 2007), they found that subtle activation produced group level emotions that were distinct from individual level emotions, were socially shared, and were related to in-group identification. Our
-2-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

study employs this methodology and extends it by varying the valence and object of the images presented. Participants in the Smith et al (2007) study were presented only with positive symbols of the group. The researchers also were only examining intra-group emotions. Considering the valence and nature of the emotion in question, Smith et al (2007) noted that the above prediction is clear for positive group emotions but the situation is more ambiguous with negative emotions. In their study, they asked participants to report their emotions as group members and as well as the strength of group identification. The results revealed strong, positive correlations between level of identification and positive emotions such as satisfaction, hopefulness, pride, and happiness. However, the correlations between level of identification and negative emotions such as fear, disgust, uneasiness, and irritation were more variable although they were generally negative and weak. Smith et al (2007) put forward two possible explanations for this pattern of results. It could be that high identifiers as suggested by Doojse et al (1998) reinterpret and reappraise group-related events to avoid negative feelings. Alternatively it could be evidence of bi-directional relationship between identification and group-based emotion where identification is not only an antecedent of emotion but can also be affected itself by emotion. In this case, negative feelings about the in-group may motivate a decrease in identification. Secondly, it is clear from the preceding discussion that self-relevance is an important determinant of group-based emotions (Lyer and Leach, 2008; Smith, 1993; Smith et al, 2007). In a group context, self-relevance is conceptualised primarily in two ways, self-categorisation as an in-group member and level of identification (Lyer and Leach, 2008). Some research suggests that simple categorical inclusion in an in-group is sufficient to experience group-level emotions, as is the premise of intergroup emotions theory (Smith, 1993; Smith et al, 2007). Others argue that the level of in-group identification (the psychological connection to a group) is an important determinant. However, there are mixed views on whether a minimal level of identification is sufficient to facilitate group-based emotions or whether a moderate or high level of identification is required (Lyer and Leach, 2008). Mackie, Silver & Smith (2004: 231) argue that the higher the level of in-group identification, the more easily, frequently and intensely group-based emotions are produced. Following on from this, it is suggested that high identifiers should experience and express group emotions to a greater extent than low identifiers (Smith et al, 2007). Our study tests this particular hypothesis that level of identification will relate to group based emotions. Overall therefore it would appear that important variables to consider are the subject of the emotion, the target of the emotion and the nature of the emotion. We conducted our study in Northern Ireland where there is a long history of group based division. The two extant groups, the can be conceptualised as two different ethnopolitical groups, with religion normally used to mark the social boundary between Catholic and Protestant. Whilst this dichotomisation can be unhelpfully simplistic (see Muldoon et al., 2007 for a discussion of these issues), in this instance it allows the assessment of relationship between identification and emotions in a real world setting, which represents an important addition of the literature in this area.

-3-

IBIS WORKING PAPERS

NO. 103, 2010

Smith et al (2007) propose that group identification correlates positively with positive emotions. Here we seek to clarify this in terms of subject-object of emotion. We hypothesize that for intra-group emotions, identification should correlate positively with positive emotions directed at the in-group. In addition, we propose that there should be a negative correlation between identification and negative emotions directed towards the in-group. Further it has been suggested that identification be partitioned into its various subcomponents and the relationship between these and groupbased emotions be investigated (Mackie et al, 2004). We therefore explore not only the relationship of overall identification with group-based emotions but also importance to identity and public evaluation of identity. Seger et al (2009) stated that different primes might produce different kinds of emotion profiles. Participants in their studies were not presented with primes that could activate a negative aspect of a social identity (p.24). The current study replicates and builds on that of Seger et al (2009). Firstly, we present our Catholic and Protestant participants resident in Northern Ireland with photographs of national flags, namely, the Union Jack and the Tricolour. We expect that when presented with a photograph that depicts a positive symbol representing the in-group that the group-based emotions that are activated should show a similar pattern to explicitly-reported intragroup emotions, i.e. there should be a positive relationship between identification and positive emotions directed towards the ingroup and a negative relationship between identification and negative emotions directed at the ingroup. Secondly we build on Seger et als (2009) study by showing participants images that depict a negative aspect of their social identity. For Protestant participants we included a photograph of the crest of the Royal Ulster Constabulary (RUC) and for Catholic participants, a photograph of a memorial commemorating the 1981 hunger strike. These images may remind ingroup members of negative events in their groups history and should impact on the group-based emotions reported. We will explore the relationship between emotion and identification and determine whether there are differences between high and low identifiers. Finally we explore the relationship between identification and intergroup emotions. We will investigate whether there are differences in emotions directed towards the out-group between high and low identifiers. We will also look at responses to images depicting the outgroup to investigate the nature of the relationship between identification and inter-group emotions. METHOD 1: Intragroup emotion and level of identification Participants The 2008 Northern Ireland Life and Times Survey involved 1,179 face-to-face interviews with adults aged 18 years or over. A systematic random sample of addresses was selected from the Land and Property Services Agency list of private addresses. This is the most up-to-date listing of private households and is made available to the Northern Ireland Statistics and Research Agency for research purposes. People living in institutions (though not in private households in such institutions) are excluded. A total of 33% of those selected did not want to participate, 8% were non
-4-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

contactable giving a final response rate of 61%. Of the final sample, 47.2% of respondents were male and 52.8% female. Design and analysis The design for this analysis was correlational. When asked, do you regard yourself as belonging to any particular religion, 85.6% (N = 1009) responded with either Protestant (47.1%, N = 555) or Catholic (38.5%, N = 454). Analysis was conducted for Protestants and Catholics separately, given that we were interested in the group based nature of emotion. Procedure The Land and Property Services Agency provides a good sampling frame of addresses, but contains no information about the number of people living at an address. Further selection stages were therefore required to convert the listing of addresses to a listing of individuals from which one person is chosen to complete the questionnaire. The interviewers listed all members of the household eligible for inclusion in the sample: that is, all persons aged 18 or over living at the address. From this listing of eligible adults, the interviewer's computer randomly selected one adult. This person, the selected respondent, was then asked to complete the survey. Results relating to this study represent one module of the survey administered using Computer Assisted Personal Interviewing (CAPI). Materials Respondents were asked to subjectively self-categorize their religious identity. Religious self-categorization was obtained by asking respondents, i.e. How would you describe your religious tradition? Of the final sample, 38.5% described themselves as Catholics, 47.1% as Protestant and 14.4% described their religion as none (10.9%) or other (3.5%) than one of the two main traditions in Northern Ireland. We then measured explicit collective self-esteem related to religious tradition using the two subscales of Luhtanen and Crockers (1992) Collective self-esteem (CSE) scale, importance of identity and public regard for identity. The reliability and validity of the scale is good and it has been widely used. The scale was adapted to ensure the CSE in relation to religious group membership was assessed so that items read for example: In general, Im glad to be a member of the religion that I belong to. Finally respondents were asked about their intragroup emotions. Four emotions, namely uneasiness, satisfaction, annoyance and hopefulness were assessed. Participants were asked to report successively on a scale of 1 (not at all) to 3 (very), how uneasy/annoyed/hopeful/satisfied they felt about being a Catholic/Protestant in Northern Ireland.

-5-

IBIS WORKING PAPERS

NO. 103, 2010

Results 1: Intragroup emotion and level of identification Table 1: Correlations of intra-group emotions with overall religious group identification and subcomponents of identification (importance and public evaluation) Identification Catholic Overall Evaluation Importance Protestant Overall Evaluation Importance
* p<0.05; **p<0.01

Uneasy

Annoyed

Satisfied

Hopeful

-.22** -.25** -.12*

-.14** -.20** -.05

.26** .24** .17**

.31** .21** .25**

-.10* -.20** .01

-.02* -.06 .02

.25** .15** .23**

.27** .20** .21**

Overall strength of identification was significantly related to all four intragroup emotions assessed for Catholic respondents (see Table 1). Identification as a Catholic was negatively correlated with uneasiness (r=-.22, p<0.01) and annoyance (r=-.14, p<0.01) and positively correlated with satisfaction (r=.26, p<0.01) and hopefulness (r=.31, p<0.01). The same pattern emerged for Protestant respondents. Identification was negatively correlated with uneasiness (r=-.01, p<0.05) and annoyance (r=.02, p<0.05) and positively correlated with satisfaction (r=.25, p<0.01) and hopefulness (r=.21, p<0.01). The importance of identity subscale assesses the importance of the group to the individuals self-concept. For Catholics, importance was significantly related to all intragroup emotions except for annoyance. Importance was negatively correlated with uneasiness (r=-.12, p<0.05) and positively correlated with satisfaction (r=.17, p<0.01) and hopefulness (r=.25, p<0.01), that is those who attached greater importance to their identity reported less uneasiness and more satisfaction and hopefulness about being Catholics in Northern Ireland. For Protestants, satisfaction (r=.23, p<0.01) and hopefulness (r=.21, p<0.01) were also positively correlated with importance, again with those reporting more satisfaction and hopefulness about being Protestants rating the importance of their religious identity more highly. However, uneasiness (r=.01, p>.05) and annoyance (r=.02, p>0.05) were not significantly related to importance to identity. Finally, the public evaluation subscale assessed perceptions of how positively other people evaluate their social group. For Catholics, public regard was significantly related to all four intra-group emotions. Public regard was negatively related to uneasiness (r=.-25, p<0.01) and annoyance (r=.-20, p<0.01) and positively related to satisfaction (r=.24, p<0.01) and hopefulness (r=.21, p<0.01). Satisfaction (r=.15,
-6-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

p<0.01) and hopefulness (r=.20, p<0.01) were also positively related to public regard for Protestants. There was a negative association between uneasiness (r=-.20, p<0.01) and public regard for this group. However there was no significant relationship between public regard and annoyance (r=-.06, p>0.05). 2: Variability in intragroup emotion and identification Participants The total sample comprised of 874 respondents, accessed via the NILT as previously outlined. Design The impact of visual primes on Protestants and Catholics intragroup emotion measured at four levels (uneasy/hopeful/annoyed/satisfied) was examined. The nature of the visual prime was manipulated and order of presentation randomised. The intragroup visual primes could be viewed as either positive for the ingroup or contentious. Overall identification with self-categorised religion is also included in this analysis, first to assess relationships via correlation and subsequently as an independent variable based on a standard median split of those scoring above the median for overall collective self-esteem as measured by Luhtanen & Crocker (1992) subscales employed. Materials A set of four visual images were selected to prime group based emotions. In all cases the images represented icons related to the conflict in Northern Ireland which of course is underpinned by identity and group based divisions (Muldoon et al., 2007). To date the group emotions literature has concentrated on the consequences of positive shared experience on emotion with less attention focused on more negative or ambiguous experiences. So in the first instance we used the Irish and British national flags to elicit emotional responses, as well as two more contentious images from the past that can be seen to have a chequered history by both groups. These images were a picture of Royal Ulster Constabulary emblem, the largely Protestant police force often targets of republican violence but have been seen as partisan by others particularly in the Catholic community. Subsequent to the Belfast Good Friday Agreement of 1998, the police force in Northern Ireland was reformed The second image was of a Hunger Strike Memorial, erected to commemorate the Irish Republicans who died in prison attempting to secure their status as prisoners of war. Others, particularly those in the Protestant community, view the Hunger Strikers as terrorists. Procedure Respondents were presented with a set of four visual images, the two relating to each group reported hereunder. Respondents were asked to indicate on a scale of 1 to 3 (where 3 was very, 2 was a bit, and 1 was not at all) how annoyed, how hopeful, how satisfied, and how uneasy each image made them feel.
-7-

IBIS WORKING PAPERS

NO. 103, 2010

Results Protestant Respondents On presentation of the Union Jack image, strength of identification was observed to be positively related to satisfaction (r=.22, p<.01) and hopefulness (r=.31, p<.01). There was no correlation between identification and annoyance (r=-.01, n.s.) or uneasiness (r=.03, n.s.). For the image of the contentious RUC crest, identification was positively related to annoyance (r=.16, p<.01), satisfaction (r=.17, p<.01) and hopefulness (r=.10, p<.05). There was no correlation between identification and uneasiness (r=.16, p<.01).
Table 2: Correlations of intra-group emotions with overall religious group identification after priming Identification on presentation of Union Jack RUC crest
*p<.05; **p<.01

Uneasy Annoyed .03 .07 -.01 .16**

Satisfied .22** .17**

Hopeful .31** .10*

A mixed MANOVA examined the effects of picture type, level of identification on intragroup emotion at four levels for Catholics and Protestants separately. The nature of the image presented (flag versus contentious) interacted with level with level of identification to affect reported in-tragroup emotion amongst Protestants, F=(4,493) 7.7, p<.001. Univariate follow up analysis of this effect indicated that this interaction was observed in relation to annoyance, F=(1, 496)8.03, p<.01, and hopefulness, F=(1,496) 17.9, p<.001, only (see figure 1). A main effect was also observed across all emotions (uneasy, F=(1, 496) 6.1, p<.05; satisfied, F=(1, 496)77.6, p<.001); annoyed, F=(1,496)35.8, p<.001) hopeful, (F=1, 496) 96.7, p<.001) related to whether the picture presented was the Union Jack or RUC picture. Level of identification also exerted a main effect on intragroup emotion reported, F=(4, 493)=7.2, p<.0.001), that was observable in relation to satisfaction, F=(1, 496)16.3, p<.001), annoyance, F=(1, 496)7.4, p<.01 and hopefulness, F=(1,496) 22.6, p<.001). In each case those with higher levels of identification with their religion reported more annoyance, less satisfaction and hopefulness than those with lower levels of identification.

-8-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

Figure 1 : Means level of emotion reported in response to in-group images for Protestant respondents

2.4 2.2 2 1.8 1.6 1.4 1.2 1 Annoyed with Union Jack Annoyed with RUC crest Hopeful with Union Jack Hopeful with RUC crest

High identifiers

Low identifiers

Catholic Respondents For the Tricolour image, strength of identification was positively associated with satisfaction (r=.18, p<.01) and hopefulness (r=.19, p<.01), that is stronger identifiers reported more satisfaction and hopefulness. Identification was not related to uneasiness (r=-.05, n.s) and annoyance (r=-.06, n.s.). For Catholics the hunger strike memorial, strength of identification was positively related to both annoyance (r=.10, p<.05) and hopefulness (r=.11, p<.05). There was no significant relationship between uneasiness and identification (r=.00, n.s.) or identification and satisfaction (r=.06, n.s.).
Table 3: Correlations of intra-group emotions with overall religious identification after priming Image Tricolour Hunger strike
*p<.05; **p<.01

Uneasy -.05 .00

Annoyed -.06 .10*

Satisfied .18** .06

Hopeful .19** .11*

We then conducted a mixed MANOVA to assess the relationship between levels of identification, the nature of the image presented and intragroup emotion. The nature of the image (neutral versus contentious) presented interacted level with level of identification to effect reported emotion, F=(4,404) 2.8, p<.05). Univariate follow up analysis of this effect indicated that this interaction was observed in relation to satisfaction, F=(1, 407)7.3, p<.01, annoyance, F=(1,407) 5.3, p<.05, and hopefulness F=(1, 407) 4.9, p<.05) (see figure 2). A main effect was also observed across all
-9-

IBIS WORKING PAPERS

NO. 103, 2010

emotions (uneasy, F=(1, 407) 109, p<.001; satisfied, F=(1, 407)145, p<.001); annoyed, F=(1,407)138, p<.001) hopeful, (F=1, 407) 64.5, p<.01) dependent on whether the tricolour or hunger strike picture was presented (see figure 2). Level of identification also exerted a main effect on intragroup emotion reported, F=(4, 404)=2.9, p<.0.05), however this was observable only in relation to satisfaction, F=(1, 407)5.6, p<.05) and hopefulness, F=(1,407) 10.5, p<.001), with higher identifiers showing higher levels of satisfaction and hopefulness.
Figure 2: Means emotion reported to in-group images for Catholic respondents by strength of identification

2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1

Annoyed with Tricolour Annoyed with HS Memorial Hopeful with Tricolour Hopeful with HS Memorial Satisfaction with Tricolour Satisfaction with HS Memorial High identifiers Low identifiers

3: Intergroup emotion and level of identification This analysis focused on strength of identification, this time relating to intergroup emotion, that is, feelings about the out-group. Design The impact of visual primes on Protestants and Catholics intergroup emotion measured at four levels (uneasy/hopeful/annoyed/satisfied) was examined. The nature of the visual prime was manipulated and order of presentation randomised. Overall identification with self-categorised religion is also included in this analysis, first to assess relationships via correlation and subsequently as an independent variable based on a standard median split. Materials The same set of four visual images were selected to prime group based emotions, the two relating to intergroup emotion reported hereunder. The images presented were the national flag associated with the outgroup (Irish Tricolour for Protestants and British Union Jack for Catholics). The second set of icons were considerably

-10-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

more contentious images in this instance, as they positioned the outgroup relative to the political conflict in Northern Ireland. Procedure Respondents were presented with a set of four visual images. Order or presentation was randomised. Respondents were asked to indicate on a scale of 1 to 3 (where 3 was very, 2 was a bit, and 1 was not at all) how annoyed, how hopeful, how satisfied, and how uneasy each image made them feel. Results: Intergroup emotion and level of identification
Table 4: Correlations of inter-group emotions in response to primes with overall religious identification after priming Identification Catholic Union Jack RUC crest .04 -.11* -.07 -.06 .05 .06 .11* .08 Uneasy Annoyed Satisfied Hopeful

Protestant Tricolour Hunger strike .10* .11** .09* .08 .03 .07 .02 .01

For Catholics, strength of identification was positively related to hopefulness (r=.11, p<.05) on presentation of the image of the Union Jack. There were no other significant associations. For the image of the RUC crest, identification was negatively related to uneasiness (r=-.11, p<.05), those with higher levels of identification reporting greater uneasiness. Again, this was the only significant association for this group. For Protestants, strength of identification was negatively related to uneasiness (r=.10, p<.05) and annoyance (r=-.09, p<.05) when presented with an image of the Irish Tricolour. There were no other significant associations. For the image of the Hunger Strike memorial, identification was negatively related to uneasiness (r=-.11, p<.01). There were no other significant associations. Our final analysis looked at the relationship between levels of in group identification (high versus low), the nature of the out group image presented (a flag versus an overtly contentious image) and intergroup emotion. These analyses were again conducted for Protestants and Catholics separately. Catholics
-11-

IBIS WORKING PAPERS

NO. 103, 2010

The nature of the image (flag versus contentious) presented did not interact with level of identification to effect reported intergroup emotion amongst Catholics, F=(4,347) 1.32, n.s. A main effect for picture presented was observed in relation to hopefulness, F=(1, 350) 16.6, p<.001). Catholics felt significantly more hopeful when viewing the image of the RUC crest (M=1.3, sd=.04) than when viewing the image of the Union Jack (M=1.22, sd=.03). Level of identification was not related to intergroup emotions reported in response to these primes. Protestants The nature of the image (flag versus contentious picture) presented did not interact with level of identification to effect reported intergroup emotion amongst Protestants, F=(4,429) .67, n.s.. A main effect for picture presented was observed in relation to uneasiness, F=(1, 432)81.42, p<.001, annoyance, F=(1,432)95.9, p<.001 and hopefulness, F=(1, 432) 9.9, p<.01). Respondents felt more hopeful when viewing the image of the Tricolour (m=1.13, sd=.02) than the Hunger Strike memorial (m=1.06, sd=.02). They also felt more uneasy when viewing the image of the Hunger Strike memorial (m=2.01, sd=.05) than the image of the Tricolour (m=1.55, sd=.04) and more annoyed when viewing the image of the Hunger Strike memorial (m=1.99, sd=.04) than when viewing the image of the Tricolour (m=1.54, sd=.04). Level of identification was not related to intergroup emotions in response to these primes.
Figure 3: Mean emotion reported to out-group images for Protestant respondents
2 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1

Uneasy Annoyed Hopeful Tricolour Hunger Strike Memorial

DISCUSSION Our findings have shown that emotions can indeed be group-based given the relationship demonstrated between identification and intragroup emotion as well as intergroup emotion in line with previous research (Kessler & Hollbach, 2005; Mackie et al, 2004; Smith et al, 2007). In terms of overall identification and intragroup emotions, the patterns that emerged for both Catholics and Protestants were as predicted. Overall identification was negatively correlated with explicitly-reported negative emotions directed at the ingroup and positively correlated with explicitly-12-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

reported positive emotions directed at the ingroup. This supports the hypothesis that increases in levels of identification are associated with increases in positive feelings about the ingroup and decreases in negative feelings about the ingroup, for explicitly reported group emotions. This differs somewhat from previous research that has produced mixed results concerning the relationship between identification and negative emotions (Smith et al, 2007). However while this relationship is particularly clear in terms of explicitly-reported intragroup emotions, the same cannot be said for our findings in relation to subtly primed group emotions. In examining positive symbols of the ingroup (i.e. national flags) we found that, in line with previous research, positive emotions were positively related to identification but found no significant relationships between negative emotions and identification. This finding was true for both Catholics and Protestants. This finding may well reflect group members work to avoid feeling negative (as suggested by Doosje et al, 2006), given that flags are not unambiguously positive in divided societies such as Northern Ireland. In considering intragroup emotion relative to the dimensions of identification measured, a more complex picture emerges. Mackie et al (2004) suggested that some aspects of identification might be more central to specific emotional reactions toevents than others. They suggested that one of these was centrality or importance of identity. Our findings were mixed concerning the relationship between importance and intragroup emotion. Importance was negatively related to uneasiness and positively related to positive emotions for Catholics. (There was no significant relationship between annoyance and importance to identity although it was in the expected direction). The pattern of relationships was very different for Protestants. Importance was positively related to positive emotions directed at the in-group (a similar finding to Catholic participants), however identity importance was not related to negative emotions. However, our data does suggest that public regard may be an important subcomponent to consider. For Catholics, public regard was negatively correlated with negative intragroup emotions and positively correlated with positive intragroup emotions. For the most part, these findings were replicated in the Protestant sample; however there was no significant correlation between feelings of annoyance and public regard (although results were in the expected direction). It is likely that believing that others hold the group in high esteem boosts positive feelings about group and depresses negative feelings. It is also possible, given the hypothesised bi-directional nature of the relationship between identification and emotion (see Kessler & Hollbach, 2005), that identification with the group could increase if members believe that others value the group. As hypothesised by Seger et al (2009), we established a relationship between emotions expressed and identification when identity was subtly primed using group relevant images. However, our results call into question the appropriateness of taking a positive/negative divide when talking about group-based emotions. It has been suggested that it is important to look at the type of emotion evoked rather than focusing on a simple positive/negative dimension of affect (Smith, 1993). Clearly, the different primes were activating different intragroup emotions, as postulated by Seger et al (2009). Further to this our evidence is that different primes were also activating different intergroup emotions. For Protestant respondents the differences in emotion directed at the out group relative to the two primes used was particularly marked,
-13-

IBIS WORKING PAPERS

NO. 103, 2010

respondents reporting significantly more uneasiness, annoyance and less hopefulness in response to the contentious Hunger Strike memorial than the Irish tricolour. In effect these findings extend the current literature by illustrating that specific primes can have particular effects that go beyond the simplistic prediction that positive images will prime positive emotion and negative images negative emotion. This would be in line with appraisal theories of emotion which suggest that different appraisals are linked to particular emotions. We hypothesized that there are differences in emotions directed towards the ingroup between high and low identifiers. The extent of identification with the religious group interacted with the valence of the image and the level of identification effect emotions directed at the ingroup for both Catholics and Protestants. Interestingly these emotions were not related in a straightforward way. For instance high Protestant identifiers reported both more positive emotion (hopefulness) and negative emotion (annoyance) when confronted with the image of an RUC emblem than low identifiers. Similarly high Catholic identifiers reported more annoyance and more hope in response to the Hunger Strike Memorial picture than low identifiers. These findings suggest that group members can feel conflicting emotions simultaneously when looking at particular images. Both contentious images represent organisations that could be considered part of the past. For Protestant high identifiers, feelings of annoyance may be related to the chequered history of the organisation and feelings of hopefulness given that it has been reformed. It could be argued that feelings of satisfaction may arise as recognition of the service provided by the force for their community. This may also be true for Catholics. Feelings of annoyance could be generated by symbol of past aggression and possible feelings of hopefulness as the image may serve as a reminder that the conflict has ended and the group has moved beyond that stage in its history. Particular images seem to affect specific emotions and again maybe this is why we should not focus on a single dimension of affect. Whilst these hypothesized differences in intragroup emotion were evident, differences in intergroup emotion related to level identification was not as readily apparent. Whilst high identifiers felt more positively about the in group, there was no evidence that they differed systematically in their emotional response to outgroup images than low identifiers. In the Protestant sample however greater identification was associated with higher overall intensity of negative emotion felt towards the out group as would be predicted by Mackie et al (2004). For some time now, researchers in the area of prejudice have argued for the need to distinguish between in group favouritism and out group derogation. This important distinction has led researchers to understanding that discrimination against the out group is not necessarily a consequence of in group liking. The current findings would appear to resonate with this pattern, in that identification with the ingroupas measured by esteem in this instancewas more strongly related to intragroup emotion rather than intergroup emotion. Theoretically this is also consistent with the origins of intergroup emotion theory which suggests that emotions reflect self-relevant motivational concerns and congruence. As the inclusion of the self in the ingroup is the basic premise of social identity approaches, it is logical and consistent with cur-14-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

rent theory that the ingroup is elicit more intense emotions and be more strongly linked to identification than to the outgroup. A second implication of our findings is the importance of conceptual clarity in the measurement of intergroup emotion. As outlined by Lyer and Leach (2008) the confabulation of intragroup, intergroup emotions as well as those relating to the relations between groups has hampered progress in this area. Our findings suggest that the distinction between intra and intergroup emotions at a minimum is an important one. Further it suggests that emotion relating to the self as a group member (i.e. as the subject of the emotion) is readily primed using which may be a more naturalistic method of eliciting emotional responses. For sure our findings add to the literature in this area by outlining the relations between the factors in an ecologically valid context. Whilst ecologically valid research is always laudable, it does without question come with its own difficulties. Our two extant groups Protestants and Catholics did not always evidence the same pattern of findings and whilst we attempted to chose equivalent symbols for priming, the reality is that the groups are not equivalent. Each group has a range of national, political and cultural baggage that is not acknowledged in the design of this study and is necessarily simplified by the Protestant/Catholic dichotomy. The different patterns of relationships without question can be said to be related to the political climate in Northern Ireland. Indeed the content of the traditional identities are fundamentally related to the changing status of the groups within Ireland and the UK and vis--vis each other. Whilst our thoughts on this issue are necessarily speculative, certainly the increased perceptions of threat and loss of esteem perceived by the Protestant community in Northern Ireland since the Belfast Good Friday agreement (MacGinty & DuToit, 2007) would appear to be reflected in the patterns of findings apparent amongst Protestants which can be characterised as more negative than that evident in Catholic respondents. The impetus for examining emotions in an inter-group context is the hypothesised link between emotion and action tendencies. The intention to act is based not only on what we think about the out-group but is also based on feelings towards the outgroup and/or inter-group relations. Of course, this intention to act depends on the emotion that is evoked. Some emotions, such as anger at the out group, are particularly associated with action tendencies (Yzerbyt et al, 2003); observers can feel angry towards out-group if the victims are member of the ingroup and the perpetrator a member of the out-group. Clearly findings such as this have particular resonance in Northern Ireland, where violence and victimization is often based in intergroup dynamics (Cairns and Darby, 1998). However our findings suggest that the strongest relationships are between identification and intragroup emotions rather than inter group emotions. This has two important implications. First, if group emotions are as relevant to community relations as is hypothesized, the intragroup concerns would appear to be of paramount importance in promoting good relations, even in situations like Northern Ireland where groups can be viewed as negatively interdependent (Muldoon, McLaughlin & Trew, 2007). Second, focusing on feelings towards the out group to maximise good relations contexts may represent a proverbial red herring.
-15-

IBIS WORKING PAPERS

NO. 103, 2010

The current study was necessarily limited. As our study was undertaken in peoples homes, it was necessary for safety reasons not to arouse or articulate intense emotions. Second though we were interested in how particular emotions resulted from appraisal of specific primes, the emotions we assessed were limited to positive and negative affects. Though our intergroup analysis allowed for the examination of how a contentious images of the outgroup can elicit negative emotions as per the predictions of appraisal theories, the inclusion of a wider array of such highly relevant emotions in the Northern Ireland context such as shame or guilt was deemed to be too sensitive and potentially risky even in these times of relative peace. at this time. Future research could also usefully extend our work to assess the relationship between different emotions and specific action tendencies which was beyond the scope of the current study. In conclusion, our findings demonstrating the emotions attached to religious identities in Northern Ireland represents an important addition to the available research in this field. Our findings have moved the literature to demonstrate these effects in real world groups as opposed to lab settings. Further we have added to literature in terms of demonstrating that these emotions can be manipulated by varying the primes presented and that group based emotion is not always unambiguously positive. Finally we demonstrate the value of conceptual clarity in the operationalisation of group emotion and have started to unpick the complexity of the specific relations between particular emotions, subtle priming and identification in context. REFERENCES
Arnold, MB (1960) Emotion and personality: Vol.1 Psychological aspects. New York: Columbia University Press Cairns, E & Darby, J (1998) The conflict in Northern Ireland: Causes, Consequences and Controls American Psychologist, 53 (7): 754-760 Crisp, RJ, Stone, CH, Hall, NR (2006) Recategorization and subgroup identification: Predicting and preventing threats from common ingroups Personality and Social Psychology Bulletin, 230-243 Doojse, B, Branscombe, NR, Spears, R, & Manstead, ASR (1998) Guilt by association: When ones group has a negative history Journal of Personality and Social Psychology, 75: 872-886 Haslam, SA (2004) Psychology in Organisations: The Social Identity Approach. London: Sage Lyer, A & Leach, CW (2008) Emotion in inter-group relations European Review of Social Psychology, 19: 86-125 Luhtanen, R, & Crocker, J (1992) A collective self-esteem scale: Self-evaluation of one's social identity Personality and Social Psychology Bulletin, 18: 302-318 Gaertner, SL, Dovidio, JF, Anastasio, PA, Bachman, BA, Rust, MC (1993) The Common Ingroup Identity Model: Recategorization and the Reduction of Intergroup Bias European Review of Social Psychology, 4 (1): 1-26

-16-

Muldoon, McNamara, Devine, Trew and Dowds / Group emotion and identification

Kessler, T & Hollbach, S (2005) Group-based emotions as determinants of ingroup identification Journal of Experimental Social Psychology 41(6): pp 677-685 Mackie, DM, Silver, LA, & Smith, ER (2004) Intergroup emotions: Emotion as an intergroup phenomenon in LZ Tiedens & CW Leach, eds, The social life of emotions. New York: Cambridge University Press: pp 227-245 MacGinty, R & Du Toit, P A Disparity of Esteem: Relative Group Status in Northern Ireland after the Belfast Agreement Political Psychology 1: 13-31 Muldoon, OT, McLaughlin, K & Trew, K (2007) Adolescents Perceptions of National Identification and Socialisation: a grounded analysis British Journal of Developmental Psychology, 579-594 Muldoon, OT, Trew, K, Todd, J, McLaughlin, K & Rougier, N (2007) The nature and meaning of identity in Northern Ireland after the Good Friday Agreement Political Psychology, 28 (1): 89-103 Pettigrew, TF & Meertens, RW (1995) Subtle and blatant prejudice in Western Europe European Journal of Social Psychology, 25: 57-75 Seger, CR, Smith, ER & Mackie, DM (2009) Subtle activation of a social categorisation triggers group-level emotions Journal of Experimental Social Psychology, 45 (3): 460-467 Smith, CA & Lazarus, RS (1993) Appraisal components, core relational themes, and the emotions Cognition and Emotion 7 (3): 233-269 Smith, ER (1993). Social identity and social emotions: Toward new conceptualisations of prejudice in DM Mackie and DL Hamilton, eds, Affect, cognition and stereotyping: Interactive processes in group perception). San Diego, CA: Academic Press: pp 297315 Tajfel, H (1982) Social psychology of intergroup relations Annual Review of Psychology, 33: 1-39 Tajfel, H & Turner, JC (1979) An integrative theory of intergroup conflict in WG Austin & S Worchel, eds, The social psychology of intergroup relations. Chicago, IL:NelsonHall: pp 7-24 Turner, J.C ( 1982) Towards a cognitive redefinition of the social group in H Tajfel, ed. Social identity and intergroup relations. Cambridge: Cambridge University Press Wenzel, M, Mummendey, A, Weber, U & Waldzuz (2003) The Ingroup as Pars Pro Toto: Projection From the Ingroup Onto the Inclusive Category as a Precursor to Social Discrimination Personality and Social Psychology Bulletin 29 (4): 461-473 Yzerbyt, V, Dumont, M, Wigdoldus, D & Gordijin, E (2003) I feel for us: The impact of categorization and identification on emotions and action tendencies British Journal of Social Psychology 42(4): 533-549

-17-

IBIS WORKING PAPERS The Institute for British-Irish Studies was established in 1999 as an interdisciplinary resource. Its aims are: to promote and conduct academic research in the area of relations between the two major traditions on the island or Ireland, and between Ireland and Great Britain, from an inter-disciplinary social science perspective. to promote and encourage collaboration with academic bodies and with individual researchers elsewhere who share an interest in the exploration of relations between different national, ethnic or racial groups. to promote contact with policy makers and opinion formers outside the university sector, and to ensure a free flow of ideas between the academic and the non-academic worlds. The Institutes working paper series is based on work in progress. The Institute accepts no responsibility for views offered. When later versions of papers are scheduled for publication elsewhere, this is indicated on the back of the title page. Please address queries to: Managing Editor IBIS Working Papers School of Politics and International Relations University College Dublin Belfield, Dublin 4, Ireland tel:+353+1+716 8670 fax:+353+1+716 1171 http://www.ucd.ie/ibis

WORKING PAPERS IN BRITISH-IRISH STUDIES

2006
57. Stephen Roper, Cross-border and local cooperation on the island of Ireland: an economic perspective 58. Jennifer Todd, A puzzle concerning borders and identities: towards a typology of attitudes to the Irish border 59. Andy Pollak, Educational cooperation on the island of Ireland: a thousand flowers and a hundred heartaches 60. Cormac Grda and Brendan M Walsh, Did (and does) the Irish border matter? 61. Liam ODowd, Cathal McCall and Ivo Damkat, Sustaining Cross-border cooperation: a cross-sectoral case study approach 62. Kevin Howard, Diasporas and ambiguous homelands: a perspective on the Irish border 63. Eoin Magennis, Patricia Clarke and Joseph Shiels, Funding support for cross-border and North-South cooperation on the island of Ireland, 1982-2005: an overview 64. Alessia Cividin, Irish cross-border cooperation: the case of the Northwest region 65. Elizabeth Meehan, Borders and employment opportunities and barriers 66. Kevin Howard, Nationalist myths: revisiting Heslingas The Irish border as a cultural divide 67. KJ Rankin, Theoretical concepts of partition and the partitioning of Ireland 68. Brian Caoindealbhin, Citizenship and borders: Irish nationality law and Northern Ireland 69. Katy Hayward, Contention, competition and crime: newspapers portrayal of borders in the north-west of Ireland 70. Kevin Howard, Continuity and change in a partitioned civil society: Whyte revisited 71. Kevin Howard, Territorial politics and Irish cycling 72. John Bradley, An island or island economies? Ireland after the Belfast Agreement 73. John Bradley, Industrial development in Ireland North and South: case studies of the textile and information technologies sectors 74. James Anderson, Irish border communities: questioning the effects of state borders and ethno-national identities 75. Michael Kennedy, The realms of practical politics: North-South cooperation on the Erne hydroelectric scheme, 1942 to 1957 76. Hastings Donnan, Fuzzy frontiers: the rural interface in South Armagh 77. Brendan OLeary, Analysing partition: definition, classification and explanation 78. Brendan OLeary, Debating partition: justifications and critiques 79. KJ Rankin, The provenance and dissolution of the Irish Boundary Commission 2007

80. Sean Farren, Sunningdale: An Agreement too soon? 81. Catherine ODonnell, The Sunningdale Communiqu, 1973, and bipartisanship in the Republic of Ireland 82. PJ McLoughlin, Dublin is just a Sunningdale away? The SDLP, the Irish Government and the Sunningdale Agreement 2009 83. Elizabeth Meehan, From conflict to consensus: the legacy of the Good Friday Agreement.The British-Irish and European contexts 84. Sir George Quigley; Tony Kennedy; Martin Mansergh, The impact of devolution on everyday life: 1999-2009 85. Aidan Gough and Eoin Magennis, The impact of devolution on everyday life, 19992009: the case of cross-border commerce 86. Gladys Ganiel and Claire Mitchell, Everyday evangelicals: life in a religious subculture after the Belfast Agreement 87. Neil Jarman and John Bell, Routine divisions: segregation and daily life in Northern Ireland 88. Elizabeth Meehan and Fiona Mackay, A "new politics" of participation? 89. Siobhan Byrne, Women and the transition from conflict in Northern Ireland: lessons for peace-building in Israel/Palestine 90. Niamh Hardiman and Muiris MacCarthaigh, Breaking with or building on the past? Reforming Irish public administration: 1958-2008 91. Frank Barry, Agricultural interests and Irish trade policy over the last half-century: A tale told without recourse to heroes 92. Graham Brownlow, Fabricating Economic Development 93. Bryan Fanning, From developmental Ireland to migration nation: immigration and shifting rules of belonging in the Republic of Ireland 94. Thomas P Murray, The curious case of socio-economic rights 95. Peter Murray, Educational developmentalists divided? Patrick Cannon, Patrick Hillery and the economics of education in the early 1960s 96. Tom Garvin, Dublin opinions: Dublin newspapers and the crisis of the fifties 97. Martin Mansergh; Jeffrey Donaldson; John Bradley, North-South relations after the boom: the impact of the credit crunch on mutual relations and understandings 2010 98. Tadhg hAnnrachin, The Church of Ireland and the native Irish population in plantation Ulster 99. Adrian Guelke, The flexibility of Northern Ireland Unionists and Afrikaner Nationalists in comparative perspective 100. Anthony D Buckley, Amity and enmity: variety in Ulster protestant culture