Positive and negative asymmetry of intergroup contact: A dynamic approach

Lead Research Organisation: University of Oxford
Department Name: Experimental Psychology

Abstract

Although positive intergroup contact has been shown to reduce prejudice, negative contact has the potential to worsen intergroup relations. The efficacy of intergroup contact hence has policy implications for wider debates about diversity, multiculturalism, and social conflict; but doubt has recently been cast on its potential in applied settings because whilst contact can be positive (e.g., having outgroup friends) it can also be negative (e.g., being bullied by an ethnic outgroup member). Moreover, diverse settings (e.g., schools, neighbourhoods) could potentially have unintended consequences of actually increasing prejudice because they may expose people to greater frequency of both positive and negative contact (which may undermine positive effects of contact). Yet little is known about the combined effects of positive and negative contact, and the consequences for increasingly diverse nations and communities. Using a variety of different methodological approaches, and testing participants in UK, GER and NL (where in each case majority group members interact with a different set of minority groups) we test the claim that negative contact exerts stronger effects on various outcomes than positive contact. In a series of studies involving both ethnic majority and minority participants, and using different methodological paradigms (diary, experimental, longitudinal survey, and social network studies), we seek to test three hypotheses that emerge from a critical reading of the relevant literature: buffering (i.e., positive contact weakens any negative effect of negative contact), augmentation (i.e., negative contact augments the impact of positive contact, because it contrasts the positive with the negative contact), and poisoning (i.e., negative contact reduces the impact of positive contact) effects. Finally, we use the cumulative knowledge of the project to test and seek to improve a large-scale social intervention, which is premised on promoting positive contact, but that has ignored any possible role for negative contact. This research further examines how and when positive and negative contact have their joint effects. They may have their impact, for example, via category salience (when contact is negative, categories are more salient). And the effect of positive vs negative contact on category salience may be qualified by various 'psychological' and 'social' factors. Among key psychological moderators, we will explore the order in which positive and negative contact are experienced (are outcomes worse when initial contact is negative?), and the intensity and consistency of the (positive or negative; are outcomes worst when one instance of negative contact follows another?). Among more 'social' factors, we will explore the impact of social norms.

Planned Impact

The most obvious beneficiaries of the research are academics. The areas of intergroup contact and of diversity - and indeed their intersection - have become growth areas in the social sciences, particularly for the disciplines of social psychology, sociology and political science. Indeed, the UK PI (Professor Miles Hewstone) has contributed via peer-reviewed journal articles, keynote lectures, and workshops at conferences, as well as departmental seminars in universities, in each of these three disciplines, and his work is highly cited across these disciplines. This area of research really is social science, and does not belong to any one discipline. Traditionally, sociology and political science have tended to look to social psychology for experimental evidence, but more recently they also appear to have been influenced by the complementary ways social psychologists analyse survey data (e.g., exploring mediating and moderating processes). Only slightly less obvious are the stakeholders. At the time of applying for the ORA grant we obtained letters of support from the following organizations/individuals: Department of Communities and Local Government (DCLG); Equality & Diversity Forum (EDF), Runnymede Trust (RT); The Challenge (TC; one of the largest charities in the UK focused on building social contact between people of different backgrounds to promote trust, including running the intervention which will be evaluated as part of the current grant); Ted Cantle (author of the Cantle Report), and Trevor Phillips (for Chair of CRE and EHRC). A range of Government Departments and agencies now draw on this work (e.g., DCLG and DfE). They are particularly keen to learn the implications of the new analysis of both positive and negative contact for their work on diverse communities and schools; see the letter of support for the original ORA grant written by Miriam Light at DCLG: "Evidence which helps us to understand 'what works' in promoting positive relations between groups is therefore of great interest and relevance to us. Within the proposed research project, we are especially interested in understanding more about the effect of negative contact and whether/how this may vary between population groups." Stakeholders require, and will demand, timely dissemination of the new work, and its presentation in non-technical form at both policy round-tables and in briefing papers. The PI has a strong track record of doing this. Finally, there is great interest beyond government, among policy researchers in Think Tanks. The PI has been a regular presenter of his work in the past, in oral and written form, and been asked to return in the future to report on this new work (IPPR and Demos).

Publications

10 25 50
publication icon
Bracegirdle C (2022) Disentangling contact and socialization effects on outgroup attitudes in diverse friendship networks. in Journal of personality and social psychology

publication icon
Bracegirdle C (2023) The socialization of perceived discrimination in ethnic minority groups. in Journal of personality and social psychology

publication icon
Friehs, M. Technological and analytical advances in intergroup contact research in under review; Journal of Social Issues

 
Description Overall, we consider the project very successful, as it produced a unique database of multiple datasets in the yet understudied but emerging field of research on the joint effects of positive and negative intergroup contact. Although our postdoctoral researcher, who started to work in this project and who was originally responsible for coordinating and producing papers in the diary survey and longitudinal survey, left the project at the end of the second year, we were able to compensate this loss in the final year by hiring a competent research associate. Even though the fluctuation of staff slightly delayed the process of publications in this project, we were able to generate already many scientific outcomes (see, for example, 'Publications'), while more promising publications are currently under review or in preparation. Moreover, we are confident that we as well as other members of the scientific community, who have access to the dataset via the online available Open Science Framework, can further produce even more findings, which are likely to lead to many important and high-ranking publications within our field. Below, we will summarize the key findings of each study in this project so far.

1. Diary Study
The diary study provided an important basis for the following studies in this project and revealed important insights regarding the frequency, intensity and location of positive and negative contact. The mean completion length of diary entries was 4.8 days (SE = 3.23). Using an open answer format, participants described a total of 3,601 positive and 531 negative contact events, a ratio in line with previous research (e.g. Graf et al., 2014; Hayward et al., 2017). A qualitative content-analysis (Mayring, 2010), conducted by the German team, was used to analyze these open questions. Two basic dimensions emerged, one concerned with where the contact happened, the other specifying the positive and negative interactions. The three most frequent places for positive and negative contact, were while going shopping or eating out (positive: 666; negative: 43), at work (positive: 568; negative: 50) and in public spaces, like on the bus or in the park (positive: 244; negative: 58). The most frequent types of contact interactions were small talk (785), communication (329) and receiving help (173) for positive contact as well as being threatened / harmed (103) and communication (38) for negative contact.

2. Longitudinal Survey
Findings of the longitudinal survey investigated, for example, the role of intensity of valenced intergroup contact as a moderator for positive and negative contact effects on intergroup attitudes. Interestingly, our results suggest that perceived intensity adds to the effects of positive contact frequency, but not to the effects of negative contact frequency, for both majority and minority group members (Schäfer et al., 2020). These findings are important for planning and designing effective contact interventions, because they suggest that it might need strong positive contact in order to match the effects of mild negative events. Further analyses regarding the dynamic PNAC effects over time (i.e., the extent to which negative contact effects threaten the efficacy of positive contact for reducing prejudice) are ongoing.

3. School Social Network Study
The key findings from the school social network study regard the effects of friendships on outgroup attitude development. We found high levels of ethnic segregation in the school friendship networks. On average, students had approximately six ingroup friends and one outgroup friend, while this numbers of ingroup and outgroup friends remained relatively stable over the studied one-year period. By using innovative longitudinal social network analysis (developing an adaptation of longitudinal social network dynamics; code available on OSF), we examined the effects of having ingroup and outgroup friends, and the effects of those friends' attitudes, on individuals' outgroup attitudes. In accordance with prior research, we found that having more outgroup friends led to more positive outgroup attitudes over time. However, we found that it was the attitudes of ingroup friends that actually shaped individuals' outgroup attitudes. This finding has important implications: Research to date has focused primarily on the effects of outgroup friends, not ingroup friends, on individuals' outgroup attitudes. Our findings show that future research should consider the under-examined, but potentially more influential, role of ingroup friendship in outgroup attitude development.
Analyses of the network data regarding negative contact, positive contact, and bullying measures are ongoing, while the most promising additional findings to date regard the bullying data. These initial findings demonstrate that bullying seems to be more common between members of the same ethnic group than between members of different ethnic groups. This suggests low levels of ethnically motivated bullying and intergroup hostility in the schools. Furthermore, bullying ties were not stable across time, indicating that bullying did not continue throughout the school year.
The outcomes of this dataset so far comprise two publications and five conference presentations. The first publication (Bracegirdle, Reimer, van Zalk, Wölfer, & Hewstone, in prep.) is a manuscript currently ready for submission. The second publication (Friehs, Bracegirdle, Barlow, Watt, & Zúñiga, 2020) is an initial (short form) manuscript currently under review for inclusion in a special issue. Moreover, we were invited to contribute to a special issue on advances in intergroup contact research in the Journal of Social Issues, which details technological and analytical advances in intergroup contact research by providing an overview of the use of longitudinal social network analysis in examining the effects of outgroup and ingroup friendships.

4. Social Intervention Study
In order to evaluate the social intervention study, we ran a series of multilevel ordinal-regression models to examine whether, as hypothesized, participants improved in measures relevant to social integration by comparing the change before and after the intervention with the change two weeks before and immediately before the intervention. We found that participants reported, on average, more favorable intergroup attitudes, more outgroup perspective-taking, and less intergroup anxiety after participating in the intervention. Only for intergroup anxiety, however, was the intervention-based change greater than the natural change before the intervention.
Moreover, we examined to what extent the effects of the intervention were greater for participants who had experienced less positive and more negative contact before the intervention, who had more positive contact during the intervention, and/or who talked about differences or commonalities with people from other social groups during the intervention. These analyses offered important insights about the factors that made the intervention more and less successful. First, in line with the broader aims of the project, we found that, for some of the outcomes, the intervention had a greater effect for participants who had more negative and fewer positive contact experiences prior to participating. Second, we found some evidence for the importance of talking not only about commonalities, but also about differences, in intergroup interactions during the interventions.
This study resulted in three main outcomes. First, we communicated our findings and recommendations (for example, about structuring intergroup dialogue) to The Challenge Network and contributed data to the organization's annual impact report (https://the-challenge.org/impact/reports/impact-report-2017-18/). Second, we presented our research at several international academic conferences. Third, we have written a manuscript describing our findings that we will soon submit for publication (Reimer et al., in prep).
Exploitation Route Policymakers working in social integration in ethnically diverse areas - such as housing or education - should be encouraged to promote interethnic mixing, in particular given the stark levels of interethnic segregation revealed in our study. At the same time, the risk of negative contact is less than expected, and the effects of positive contact make it worthwhile. It might be important though to try making the intensity of positive contact as strong as possible, while reducing the likelihood for intense negative contact in order to end up with favorable attitudes as a consequence of interethnic mixing.
Sectors Communities and Social Services/Policy,Education,Other

 
Description see section on: 'Influence on Policy, Practice, Patients & the Public'
First Year Of Impact 2018
Sector Communities and Social Services/Policy,Education,Other
Impact Types Societal,Policy & public services

 
Description Evaluated intergroup contact intervention
Geographic Reach National 
Policy Influence Type Influenced training of practitioners or researchers
Impact Social workers benefit from our research in the form of an evaluated intergroup contact intervention (explained in more detail in 'Research Databases & Models' and 'Key Findings').
 
Title Diary Study 
Description The main aim of the diary study was to obtain data about the frequency, intensity and location of positive and negative contact experiences and to, furthermore, investigate the effects of personality variables (e.g., RWA and SDO), perceived norms and contextual predictors on these daily positive and negative contact experiences. The implementation of the diary study was kept closely to the original plan: the 13-day diary study was conducted by an external survey company (Ipsos MORI). Respondents were N = 780 majority (White British) and N = 605 minority (British Asians) respondents, purposively sampled from ethnically mixed neighborhoods (Lower Super Output Areas, LSOA). The core of this diary study was a daily report of the frequency of overall, positive, and negative intergroup contact, with open answers to provide information about qualitative aspects of the most memorable positive and negative contact experience for each day. As planned, we also run a pre-test and a post-test survey. Regarding the latter, we asked half of the sample 2 weeks and half of the sample 6 weeks after the diary study to recall positive and negative contact experiences. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact please see section on key findings 
 
Title Longitudinal Survey 
Description The two-wave longitudinal survey is, to our knowledge, the first longitudinal dataset including positive and negative contact experiences as well as the theoretically most important moderators and mediators in this respect. The main focus of this study was to test the longitudinal effects of and interactions between positive and negative intergroup contact. The survey was conducted in line with the proposed planning by an external survey company (Ipsos MORI), which used a face-to-face random location quota approach. Respondents were sampled from diverse areas with a mix of White British and Asian British (of Indian, Pakistani and Bangladeshi heritage) residents as well as low, medium, and high levels of deprivation. In line with our estimated number of participants, White British (n = 1520) and Asian British participants (n = 1474) from 290 neighborhoods were recruited and longitudinally followed. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact see section on key findings 
 
Title New data analysis technique 
Description We developed a novel technique in longitudinal social network analysis (RSiena) that enables the examination of separate predictor and outcome variables depending on the group membership of individuals and their ties. The respective research article applying this new technique will soon be published (Bracegirdle et al., in prep.; see 'Publications'). We will make the code open access (available on OSF), so that future research can use this new technique. 
Type Of Material Data analysis technique 
Year Produced 2019 
Provided To Others? Yes  
Impact This technique facilitates future research on intergroup relations using longitudinal network analysis. 
 
Title School Social Network Study 
Description The main aim of the school social network study was to obtain social network measures of contact between adolescents in schools, and to examine the co-development of adolescents' social networks and outgroup attitudes over time. The implementation of the school social network study was kept closely to the original plan: After planning the study, designing the social network survey, and recruiting schools in year 1, we have collected five waves of data over the academic year 2017-2018 in year 2 (Wave 1: 10/2017, Wave 2: 11/2017, Wave 3: 02/2018, Wave 4: 03/2018, & Wave 5: 05/2018), and were (and still are) analyzing the data, presenting the findings at conferences, and writing papers in the third year of the grant. Respondents were N = 1328 students (aged 11-14) nested in six year-group networks in two schools in North West England. As this study focused on relations between White British and Asian British individuals, the final analytic sample included N = 1170 students (n = 341 White British & n = 829 Asian British). Students completed identical surveys at each of the five waves, which contained four social network measures (friendship, bullying, positive contact, & negative contact), and a range of self-reported measures (e.g., outgroup attitudes). 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact see section on key findings 
 
Title Social Intervention Study 
Description We worked with The Challenge Network to evaluate the effectiveness of the National Citizen Service programme as a means for improving outcomes relevant to social integration among 15- to 17-year-olds from diverse ethnic and socio-economic backgrounds. This provided us with an opportunity to evaluate an intervention that was intended to facilitate positive intergroup contact experiences while operating at a large scale (reaching, for example, 45,000 adolescents in the year of our study), allowing us to estimate how effect intergroup contact is in effecting positive change in the real-life settings. After evaluating two pilot studies and designing the questionnaires in coordination with the partner organization in year 1, we surveyed 2,099 participants (n = 579 Asian, n = 317 Black, n = 945 White, n = 258 other) of the National Citizen Service programme. To evaluate the intervention (in contrast to the natural development in key constructs), we surveyed participants two weeks before the intervention, right before the intervention, and three to four weeks after the intervention. 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact see section on key findings 
 
Title Summary 
Description In the course of this project, we have conducted several studies that allowed us to collect a variety of unique data (see Table 1 for an overview). All data of this project were made available on the Open Science Framework Platform (see links below). Moreover, during our first longitudinal analyses, we also started to develop a novel data analysis technique, which similarly will be made available to the scientific community via the Open Science Framework Platform. Links to the OSF sites are as follows: Longitudinal Survey: https://osf.io/9vgwu/?view_only=0f316d4aae7d43a497e6641a0e83e123; Diary Study: https://osf.io/fj2m7/?view_only=8a54bc1e53854855838d587fa585d612; School Social Network Study: https://osf.io/4ea9w/?view_only=d3d3c9cbefb8418ca8b493ba5495628d 
Type Of Material Database/Collection of data 
Year Produced 2019 
Provided To Others? Yes  
Impact see details in following entries 
 
Description Joint research project with a German and Dutch in the ORA framework 
Organisation University of Hagen
Country Germany 
Sector Academic/University 
PI Contribution The three institutions collaborated within the ORA framework on this grant. Each with a specific focus and conducted studies, which however informed each other and guided the overall research.
Collaborator Contribution see above
Impact joined papers see section on publications
Start Year 2016
 
Description Joint research project with a German and Dutch in the ORA framework 
Organisation Utrecht University
Country Netherlands 
Sector Academic/University 
PI Contribution The three institutions collaborated within the ORA framework on this grant. Each with a specific focus and conducted studies, which however informed each other and guided the overall research.
Collaborator Contribution see above
Impact joined papers see section on publications
Start Year 2016
 
Description The Challenge Network (charity no. 1129239; one of the largest charities in the UK focused on building social contact between people of different backgrounds to promote trust) 
Organisation The Challenge Network
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution We provided evidence-based recommendations (e.g., about structuring intergroup dialogue) and empirical data to the Challenge Network for their annual impact report.
Collaborator Contribution The Challenge Network provided us with access to their sample with which they were conducting a social contact intervention.
Impact Reimer et al. (in prep.: see section on publications
Start Year 2017
 
Title Social Intervention 
Description In the course of this project and in cooperation with our partner The Challenge, we have evaluated, and found support for, a large-scale intervention. 
Type Preventative Intervention - Behavioural risk modification
Year Development Stage Completed 2018
Development Status On hold
Impact The intervention produced findings that allowed us to make recommendations on how to improve intergroup relations and foster social integration in the United Kingdom. 
 
Description Challenge Network: annual impact report 
Form Of Engagement Activity A magazine, newsletter or online publication
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact In the course of the conducted social intervention study in our project (explained in more detail in 'Research Databases & Models' and 'Key Findings') we provided evidence-based recommendations (e.g., about structuring intergroup dialogue) and empirical data to the Challenge Network for their annual impact report.
Year(s) Of Engagement Activity 2018
URL https://the-challenge.org/impact/reports/impact-report-2017-18
 
Description Participation in an open-days event at Oxford University 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Schools
Results and Impact In June 2018, we organised a two-day outreach event at the Oxford of University for a total of 30 GCSE Psychology students from the schools, which participated in the School Social Network Study. The students had sessions on Psychology and were introduced to the university life and the Oxford admissions process. Accommodation in an Oxford College, meals, and visits to three additional Oxford Colleges were provided.
The outreach event specifically enabled students from underrepresented backgrounds at Oxford University to visit and get a better understanding of this institution (primarily Asian British and low-income backgrounds). The visit was a great success, as evidenced by the following quote from the school's head teacher:
"I just wanted to say a big thank you for facilitating such a fantastic trip to the University for our Year 10 Psychology students. I've caught up with many of them this morning and without exception they all said that they were made to feel very welcome and really enjoyed the experience. A significant number have even stated that they would very much like to attend the University as undergraduates in the near future!".
Year(s) Of Engagement Activity 2018