hypothesis contact theory

What Is the Contact Hypothesis in Psychology?

Can getting to know members of other groups reduce prejudice?

Jacob Ammentorp Lund / Getty Images 

• Archaeology
• Ph.D., Psychology, University of California - Santa Barbara
• B.A., Psychology and Peace & Conflict Studies, University of California - Berkeley

The contact hypothesis is a theory in psychology which suggests that prejudice and conflict between groups can be reduced if members of the groups interact with each other.

Key Takeaways: Contact Hypothesis

• The contact hypothesis suggests that interpersonal contact between groups can reduce prejudice.
• According to Gordon Allport, who first proposed the theory, four conditions are necessary to reduce prejudice: equal status, common goals, cooperation, and institutional support.
• While the contact hypothesis has been studied most often in the context of racial prejudice, researchers have found that contact was able to reduce prejudice against members of a variety of marginalized groups.

Historical Background

The contact hypothesis was developed in the middle of the 20th century by researchers who were interested in understanding how conflict and prejudice could be reduced. Studies in the 1940s and 1950s , for example, found that contact with members of other groups was related to lower levels of prejudice. In one study from 1951 , researchers looked at how living in segregated or desegregated housing units was related to prejudice and found that, in New York (where housing was desegregated), white study participants reported lower prejudice than white participants in Newark (where housing was still segregated).

One of the key early theorists studying the contact hypothesis was Harvard psychologist Gordon Allport , who published the influential book The Nature of Prejudice in 1954. In his book, Allport reviewed previous research on intergroup contact and prejudice. He found that contact reduced prejudice in some instances, but it wasn’t a panacea—there were also cases where intergroup contact made prejudice and conflict worse. In order to account for this, Allport sought to figure out when contact worked to reduce prejudice successfully, and he developed four conditions that have been studied by later researchers.

Allport’s Four Conditions

According to Allport, contact between groups is most likely to reduce prejudice if the following four conditions are met:

• The members of the two groups have equal status. Allport believed that contact in which members of one group are treated as subordinate wouldn’t reduce prejudice—and could actually make things worse.
• The members of the two groups have common goals.
• The members of the two groups work cooperatively. Allport wrote , “Only the type of contact that leads people to do things together is likely to result in changed attitudes.”
• There is institutional support for the contact (for example, if group leaders or other authority figures support the contact between groups).

Evaluating the Contact Hypothesis

In the years since Allport published his original study, researchers have sought to test out empirically whether contact with other groups can reduce prejudice. In a 2006 paper, Thomas Pettigrew and Linda Tropp conducted a meta-analysis: they reviewed the results of over 500 previous studies—with approximately 250,000 research participants—and found support for the contact hypothesis. Moreover, they found that these results were not due to self-selection (i.e. people who were less prejudiced choosing to have contact with other groups, and people who were more prejudiced choosing to avoid contact), because contact had a beneficial effect even when participants hadn’t chosen whether or not to have contact with members of other groups.

While the contact hypothesis has been studied most often in the context of racial prejudice, the researchers found that contact was able to reduce prejudice against members of a variety of marginalized groups. For example, contact was able to reduce prejudice based on sexual orientation and prejudice against people with disabilities. The researchers also found that contact with members of one group not only reduced prejudice towards that particular group, but reduced prejudice towards members of other groups as well.

What about Allport’s four conditions? The researchers found a larger effect on prejudice reduction when at least one of Allport’s conditions was met. However, even in studies that didn’t meet Allport’s conditions, prejudice was still reduced—suggesting that Allport’s conditions may improve relationships between groups, but they aren’t strictly necessary.

Why Does Contact Reduce Prejudice?

Researchers have suggested that contact between groups can reduce prejudice because it reduces feelings of anxiety (people may be anxious about interacting with members of a group they have had little contact with). Contact may also reduce prejudice because it increases empathy and helps people to see things from the other group’s perspective. According to psychologist Thomas Pettigrew and his colleagues , contact with another group allows people “to sense how outgroup members feel and view the world.”

Psychologist John Dovidio and his colleagues suggested that contact may reduce prejudice because it changes how we categorize others. One effect of contact can be decategorization , which involves seeing someone as an individual, rather than as only a member of their group. Another outcome of contact can be recategorization , in which people no longer see someone as part of a group that they’re in conflict with, but rather as a member of a larger, shared group.

Another reason why contact is beneficial is because it fosters the formation of friendships across group lines.

Limitations and New Research Directions

Researchers have acknowledged that intergroup contact can backfire , especially if the situation is stressful, negative, or threatening, and the group members did not choose to have contact with the other group. In his 2019 book The Power of Human , psychology researcher Adam Waytz suggested that power dynamics may complicate intergroup contact situations, and that attempts to reconcile groups that are in conflict need to consider whether there is a power imbalance between the groups. For example, he suggested that, in situations where there is a power imbalance, interactions between group members may be more likely to be productive if the less powerful group is given the opportunity to express what their experiences have been, and if the more powerful group is encouraged to practice empathy and seeing things from the less powerful group’s perspective.

Can Contact Promote Allyship?

One especially promising possibility is that contact between groups might encourage more powerful majority group members to work as allies —that is, to work to end oppression and systematic injustices. For example, Dovidio and his colleagues suggested that “contact also provides a potentially powerful opportunity for majority-group members to foster political solidarity with the minority group.” Similarly, Tropp—one of the co-authors of the meta-analysis on contact and prejudice— tells New York Magazine’s The Cut that “there’s also the potential for contact to change the future behavior of historically advantaged groups to benefit the disadvantaged.”

While contact between groups isn’t a panacea, it’s a powerful tool to reduce conflict and prejudice—and it may even encourage members of more powerful groups to become allies who advocate for the rights of members of marginalized groups.

Sources and Additional Reading:

• Allport, G. W. The Nature of Prejudice . Oxford, England: Addison-Wesley, 1954. https://psycnet.apa.org/record/1954-07324-000
• Dovidio, John F., et al. “Reducing Intergroup Bias Through Intergroup Contact: Twenty Years of Progress and Future Directions.”  Group Processes & Intergroup Relations , vol. 20, no. 5, 2017, pp. 606-620. https://doi.org/10.1177/1368430217712052
• Pettigrew, Thomas F., et al. “Recent Advances in Intergroup Contact Theory.”  International Journal of Intercultural Relations , vol. 35 no. 3, 2011, pp. 271-280. https://doi.org/10.1016/j.ijintrel.2011.03.001
• Pettigrew, Thomas F., and Linda R. Tropp. “A Meta-Analytic Test of Intergroup Contact Theory.”  Journal of Personality and Social Psychology , vol. 90, no. 5, 2006, pp. 751-783. http://dx.doi.org/10.1037/0022-3514.90.5.751
• Singal, Jesse. “The Contact Hypothesis Offers Hope for the World.” New York Magazine: The Cut , 10 Feb. 2017. https://www.thecut.com/2017/02/the-contact-hypothesis-offers-hope-for-the-world.html
• Waytz, Adam. The Power of Human: How Our Shared Humanity Can Help Us Create a Better World . W.W. Norton, 2019.
• What Was the Robbers Cave Experiment in Psychology?
• What Is a Microaggression? Everyday Insults With Harmful Effects
• Understanding Social Identity Theory and Its Impact on Behavior
• What Is Deindividuation in Psychology? Definition and Examples
• What Is Social Loafing? Definition and Examples
• Definition of Scapegoat, Scapegoating, and Scapegoat Theory
• What Is Stereotype Threat?
• What Is the Mere Exposure Effect in Psychology?
• Implicit Bias: What It Means and How It Affects Behavior
• What Is Mindfulness in Psychology?
• What's the Difference Between Prejudice and Racism?
• What Is Groupthink? Definition and Examples
• Understanding the Big Five Personality Traits
• Definition and Examples of Social Distance in Psychology
• What Is Flirting? A Psychological Explanation
• What Is a Contact Language?

magazine issue 2 2013 / Issue 17

Justice seems not to be for all: exploring the scope of justice.

written by Aline Lima-Nunes, Cicero Roberto Pereira & Isabel Correia

That human touch that means so much: Exploring the tactile dimension of social life

written by Mandy Tjew A Sin & Sander Koole

Intergroup Contact Theory: Past, Present, and Future

• written by Jim A. C. Everett
• edited by Diana Onu

In the midst of racial segregation in the U.S.A and the ‘Jim Crow Laws’, Gordon Allport (1954) proposed one of the most important social psychological events of the 20th century, suggesting that contact between members of different groups (under certain conditions) can work to reduce prejudice and intergroup conflict . Indeed, the idea that contact between members of different groups can help to reduce prejudice and improve social relations is one that is enshrined in policy-making all over the globe. UNESCO, for example, asserts that contact between members of different groups is key to improving social relations. Furthermore, explicit policy-driven moves for greater contact have played an important role in improving social relations between races in the U.S.A, in improving relationships between Protestants and Catholics in Northern Ireland, and encouraging a more inclusive society in post-Apartheid South Africa. In the present world, it is this recognition of the benefits of contact that drives modern school exchanges and cross-group buddy schemes. In the years since Allport’s initial intergroup contact hypothesis , much research has been devoted to expanding and exploring his contact hypothesis . In this article I will review some of the vast literature on the role of contact in reducing prejudice , looking at its success, mediating factors, recent theoretical extensions of the hypothesis and directions for future research. Contact is of utmost importance in reducing prejudice and promoting a more tolerant and integrated society and as such is a prime example of the real life applications that psychology can offer the world.

The Contact Hypothesis

The intergroup contact hypothesis was first proposed by Allport (1954), who suggested that positive effects of intergroup contact occur in contact situations characterized by four key conditions: equal status, intergroup cooperation , common goals, and support by social and institutional authorities (See Table 1). According to Allport, it is essential that the contact situation exhibits these factors to some degree. Indeed, these factors do appear to be important in reducing prejudice , as exemplified by the unique importance of cross-group friendships in reducing prejudice (Pettigrew, 1998). Most friends have equal status, work together to achieve shared goals, and friendship is usually absent from strict societal and institutional limitation that can particularly limit romantic relationships (e.g. laws against intermarriage) and working relationships (e.g. segregation laws, or differential statuses).

Since Allport first formulated his contact hypothesis , much work has confirmed the importance of contact in reducing prejudice . Crucially, positive contact experiences have been shown to reduce self-reported prejudice (the most common way of assessing intergroup attitudes) towards Black neighbors, the elderly, gay men, and the disabled - to name just a few (Works, 1961; Caspi, 1984; Vonofako, Hewstone, & Voci, 2007; Yuker & Hurley, 1987). Most interestingly, though, in a wide-scale meta-analysis (i.e., a statistical analysis of a number of published studies), it has been found that while contact under Allport’s conditions is especially effective at reducing prejudice , even unstructured contact reduces prejudice (Pettigrew & Tropp, 2006). What this means is that Allport’s proposed conditions should be best be seen as of a facilitating, rather than an essential, nature. This is important as it serves to show the importance of the contact hypothesis : even in situations which are not marked by Allport’s optimal conditions, levels of contact and prejudice have a negative correlation with an effect size comparable to those of the inverse relationship between condom use and sexually transmitted HIV and the relationship between passive smoking and the incidence of lung cancer at work (Al-Ramiah & Hewstone, 2011). Contact between groups, even in sub-optimal conditions, is strongly associated with reduced prejudice .

Importantly, contact does not just influence explicit self-report measures of prejudice , but also reduces prejudice as measured in a number of different ways. Explicit measures (e.g. ‘How much do you like gay men?’) are limited in that there can be a self-report bias: people often answer in a way that shows them in a good light. As such, research has examined the effects of contact on implicit measures: measures that involve investigating core psychological constructs in ways that bypass people’s willingness and ability to report their feelings and beliefs. Implicit measures have been shown to be a good complement to traditional explicit measures - particularly when there may be a strong chance of a self-report bias. In computer reaction time tasks, contact has been shown to reduce implicit associations between the participant’s own in-group and the concept ‘good’, and between an outgroup (a group the participant is not a member of) and the concept ‘bad’ (Aberson and Haag, 2007). Furthermore, positive contact is associated with reduced physiological threat responses to outgroup members (Blascovich et al., 2001), and reduced differences in the way that faces are processed in the brain, implying that contact helps to increase perceptions of similarity (Walker et al., 2008). Contact, then, has a real and tangible effect on reducing prejudice – both at the explicit and implicit level. Indeed, the role of contact in reducing prejudice is now so well documented that it justifies being referred to as intergroup contact theory (Hewstone & Swart, 2011).

How does it work?

Multiple mechanisms have been proposed to explain just how contact reduces prejudice . In particular, “four processes of change” have been proposed: learning about the out-group , changing behavior, generating affective ties, and in-group reappraisal (Pettigrew, 1998). Contact can, and does, work through both cognitive (i.e. learning about the out-group , or reappraising how one thinks about one’s own in-group ), behavioural (changing one’s behavior to open oneself to potential positive contact experiences), and affective (generating affective ties and friendships, and reducing negative emotions) means. A particularly important mediating mechanism (i.e. the mechanisms or processes by which contact achieves its effect) is that of emotions, or affect, with evidence suggesting that contact works to reduce prejudice by diminishing negative affect (anxiety / threat) and inducing positive affect such as empathy (Tausch and Hewstone, 2010). In another meta-analysis , Pettigrew and Tropp (2008) supported this by looking specifically at mediating mechanisms in contact and found that contact situations which promote positive affect and reduce negative affect are most likely to succeed in conflict reduction. Contact situations are likely to be effective at improving intergroup relations insofar as they induce positive affect, and ineffective insofar as they induce negative affect such as anxiety or threat. If we feel comfortable and not anxious, the contact situation will be much more successful.

Generalizing the effect

An important issue that I have not yet addressed, however, is how these positive experiences after contact can be extended and generalized to other members of the outgroup . While contact may reduce an individual’s prejudice towards (for example) their Muslim colleague, its practical use is strongly limited if it doesn’t also diminish prejudice towards other Muslims. Contact with each and every member of an outgroup – let alone of all out-groups to which prejudice is directed – is clearly unfeasible and so a crucial question in intergroup contact research is how the positive effect can be generalized.

A number of approaches have been developed to explain how the positive effect of contact, including making group saliency low so that people focus on individual characteristics and not group-level attributes (Brewer & Miller, 1984), making group saliency high so that the effect is best generalized to others (Johnston & Hewstone, 1992), and making an overarching common ingroup identity salient (Gaertner, Dovidio, Anastasio, Bachman, & Rust, 1993). Each of these approaches have both advantages and disadvantages, and in particular each individual approach may be most effective at different stages of an extended contact situation. To deal with this issue Pettigrew (1998) proposed a three stage model to take place over time to optimize successful contact and generalization. First is the decategorization stage (as in Brewer & Miller, 1984), where participants’ personal (and not group) identities should be emphasized to reduce anxiety and promote interpersonal liking. Secondly, the individuals’ social categories should be made salient to achieve generalization of positive affect to the outgroup as a whole (as in Johnston & Hewstone, 1992). Finally, there is the recategorization stage, where participants’ group identities are replaced with a more superordinate group: changing group identities from ‘Us vs. Them’ to a more inclusive ‘We’ (as in Gaertner et al., 1993). This stage model could provide an effective method of generalizing the positive effects of intergroup contact.

Theoretical Extensions

Even with such work on generalization, however, it may still be unrealistic to expect that group members will have sufficient opportunities to engage in positive contact with outgroup members: sometimes positive contact between group members is incredibly difficult, if not impossible. For example, at the height of the Northern Ireland conflict, positive contact between Protestants and Catholics was nigh on impossible. As such, recent work on the role of intergroup contact in reducing prejudice has moved away from the idea that contact must necessarily include direct (face-to-face) contact between group members and instead includes the notion that indirect contact (e.g. imagined contact, or knowledge of contact among others) may also have a beneficial effect.

A first example of this approach comes from Wright, Aron, McLaughlin-Volpe, and Ropp’s (1997) extended contact hypothesis . Wright et al. propose that mere knowledge that an ingroup member has a close relationship with an outgroup member can improve outgroup attitudes, and indeed this has been supported by a series of experimental and correlational studies. For example, Shiappa, Gregg, & Hewes, (2005) have offered evidence suggesting that just watching TV shows that portrayed intergroup contact was associated with lower levels of prejudice . A second example of an indirect approach to contact comes from Crisp and Turner’s (2009) imagined contact hypothesis , which suggests that actual experiences may not be necessary to improve intergroup attitudes, and that simply imagining contact with outgroup members could improve outgroup attitudes. Indeed, this has been supported in a number of studies at both an explicit and implicit level: British Muslims (Husnu & Crisp, 2010), the elderly (Abrams, Crisp, & Marques 2008), and gay men (Turner, Crisp, & Lambert, 2007).

These more recent extensions of the contact hypothesis have offered important suggestions on how to most effectively generalize the benefits of the contact situation and make use of findings from work on mediating mechanisms. It seems that direct face-to-face contact is always not necessary, and that positive outcomes can be achieved by positive presentation of intergroup-friendships in the media and even simply by imagining interacting with an outgroup member.

Issues and Directions for Future Research

Contact, then, has important positive effects on improved intergroup relations. It does have its critics, however. Notably, Dixon, Durrheim, & Tredoux (2005) argue that while contact has been important in showing how we can promote a more tolerant society, the existing literature has an unfortunate absence of work on how intergroup contact can affect societal change: changes in outgroup attitudes from contact do not necessarily accompany changes in the ideological beliefs that sustain group inequality. For example, Jackson and Crane (1986) demonstrated that positive contact with Black individuals improved Whites’ affective reactions towards Blacks but did not change their attitudes towards policy in combating inequality in housing, jobs and education. Furthermore, contact may also have the unintended effect of weakening minority members’ motivations to engage in collective action aimed at reducing the intergroup inequalities. For example, Dixon, Durrheim, & Tredoux (2007) found that the more contact Black South Africans had with White South Africans, the less they supported policies aimed at reducing racial inequalities. Positive contact may have the unintended effect of misleading members of disadvantaged groups into believing inequality will be addressed, thus leaving the status differentials intact. As such, a fruitful direction for future research would be to investigate under what conditions contact could lead to more positive intergroup relations without diminishing legitimate protest aimed at reducing inequality. One promising suggestion is to emphasize commonalities between groups while also addressing unjust group inequalities during the contact situation. Such a contact situation could result in prejudice reduction without losing sight of group inequality (Saguy, Tausch, Dovidio, & Pratto, 2009).

A second concern with contact research is that while contact has shown to be effective for more prejudiced individuals, there can be problems with getting a more prejudiced individual into the contact situation in the first place. Crisp and Turner’s imagined contact hypothesis seems to be a good first step in tackling this problem (Crisp & Turner, 2013), though it remains to be seen if, and how, such imagined contact among prejudiced individuals can translate to direct contact. Greater work on individual differences in the efficacy of contact would provide an interesting contribution to existing work.

Conclusions

Contact, then, has been shown to be of utmost importance in reduction of prejudice and promotion of more positive intergroup attitudes. Such research has important implications for policy work. Work on contact highlights the importance of institutional support and advocation of more positive intergroup relations, the importance of equal status between groups, the importance of cooperation between groups and the importance of positive media presentations of intergroup friendships - to name just a few. As Hewstone and Swart (2011) argue,

“Theory-driven social psychology does matter, not just in the laboratory, but also in the school, the neighborhood, and the society at large” (Hewstone & Swart, 2011. p.380).

• Aberson, C. L., & Haag, S. C. (2007). Contact, perspective taking, and anxiety as predictors of stereotype endorsement, explicit attitudes, and implicit attitudes. Group Processes and Intergroup Relations, 10 , 179–201.
• Abrams, D. and Crisp, R.J. & Marques, S., (2008). Threat inoculation: Experienced and Imagined intergenerational contact prevent stereotype threat effects on older peoples math performance. Psychology and Aging, 23 (4), 934-939.
• Al Ramiah, A., & Hewstone, M. (2011). Intergroup difference and harmony: The role of intergroup contact. In P. Singh, P. Bain, C-H. Leong, G. Misra, and Y. Ohtsubo. (Eds.), Individual, group and cultural processes in changing societies. Progress in Asian Social Psychology (Series 8), pp. 3-22. Delhi: University Press.
• Allport, G. W. (1954). The nature of prejudice . Cambridge/Reading, MA: Addison-Wesley.
• Aronson, E., & Patnoe, S. (1997). The jigsaw classroom: Building cooperation in the classroom (Vol. 978, p. 0673993830). New York: Longman.
• Blascovich, J., Mendes, W. B., Hunter, S. B., Lickel, B., & Kowai-Bell, N. (2001). Perceiver threat in social interactions with stigmatized others. Journal of Personality and Social Psychology, 80, 253–267.
• Brewer, M. B., & Kramer, R. M. (1985). The psychology of intergroup attitudes and behavior. Annual review of psychology, 36 (1), 219-243.
• Brewer, M. B., & Miller, N. (Eds.). (1984). Groups in contact: The psychology of desegregation. Academic Press.
• Caspi, A. (1984). Contact hypothesis and inter-age attitudes: A field study of cross-age contact. Social Psychology Quarterly, 74-80.
• Chu, D., & Griffey, D. (1985). The contact theory of racial integration: The case of sport. Sociology of Sport Journal, 2 (4), 323-333.
• Cohen, E. G., & Lotan, R. A. (1995). Producing equal-status interaction in the heterogeneous classroom. American Educational Research Journal, 32 (1), 99-120.
• Crisp, R. J., & Turner, R. N. (2009). Can imagined interactions produce positive perceptions? Reducing prejudice through simulated social contact. American Psychologist, 64, 231–240.
• Crisp, R. J., & Turner, R. N. (2013). Imagined intergroup contact: Refinements, debates and clarifications. In G. Hodson & M. Hewstone (Eds.), Advances in intergroup contact. New York, NY: Psychology Press.
• Dixon, J., Durrheim, K., & Tredoux, C. (2005). Beyond the optimal contact strategy: A reality check for the contact hypothesis . American Psychologist, 60, 697–711.
• Dixon, J., Durrheim, K., & Tredoux, C. (2007). Intergroup contact and attitudes toward the principle and practice of racial equality. Psychological Science, 18, 867–872.
• Gaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust, M. C. (1993). The common ingroup identity model: Recategorization and the reduction of intergroup bias. European Review of Social Psychology, 4 (1), 1-26.
• Hewstone, M., & Swart, H. (2011). Fifty-odd years of inter-group contact: From hypothesis to integrated theory. British Journal of Social Psychology, 50 (3), 374-386.
• Husnu, S. & Crisp, R. J. (2010). Elaboration enhances the imagined contact effect. Journal of Experimental Social Psychology, 46, 943-950
• Jackman, M.R., & Crane, M. (1986). “Some of my best friends are black...”: interracial friendship and whites’ racial attitudes. Public Opinion Quarterly 50, pp. 459–86
• Johnston, L., & Hewstone, M. (1992). Cognitive models of stereotype change: 3. Subtyping and the perceived typicality of disconfirming group members. Journal of Experimental Social Psychology, 28 (4), 360-386.
• Landis D., Hope R.O., & Day H.R. (1984). Training for desegregation in the military. In N. Miller & M. B. Brewer 1984, Groups in Contact: The Psychology of Desegregation, pp. 257–78. Orlando, FL: Academic Press.
• Miller, N., & Brewer M. B., eds. (1984). Groups in Contact: The Psychology of Desegregation. Orlando, FL: Academic Press.
• Pettigrew, T. F. (1998). Intergroup contact theory. Annual review of psychology, 49 (1), 65-85.
• Pettigrew, T. F., & Tropp, L. R. (2006). A meta- analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90 (5), 751.
• Pettigrew, T. F., & Tropp, L. R. (2008). How does intergroup contact reduce prejudice ? Meta- analytic tests of three mediators. European Journal of Social Psychology, 38 (6), 922-934.
• Saguy, T., Tausch, N., Dovidio, J. F., & Pratto, F. (2009). The irony of harmony: Intergroup contact can produce false expectations for equality. Psychological Science, 20 , 114–121.
• Schiappa, E., Gregg, P. B., & Hewes, D. E. (2006). Can One TV Show Make a Difference? Will & Grace and the Parasocial Contact Hypothesis . Journal of Homosexuality, 51 (4), 15-37.
• Tausch, N., & Hewstone, M. (2010). Intergroup contact and prejudice . In J. F. Dovidio, M. Hewstone, P. Glick, & V. M. Esses (Eds.), The Sage handbook of prejudice , stereotyping, and discrimination (pp. 544–560). Newburg Park, CA: Sage.
• Turner, R. N., Crisp, R. J., & Lambert, E. (2007). Imagining intergroup contact can improve intergroup attitudes. Group Processes and Intergroup Relations, 10 , 427-441.
• Vonofakou, C., Hewstone, M., & Voci, A. (2007). Contact with outgroup friends as a predictor of meta-attitudinal strength and accessibility of attitudes towards gay men. Journal of Personality and Social Psychology, 92 , 804–820.
• Works, E. (1961).The prejudice -interaction hypothesis from the point of view of the Negro minority group. American Journal of Sociology. 67 : 47–52
• Wright, S. C., Aron, A., McLaughlin-Volpe, T., & Ropp, S. A. (1997). The extended contact effect: Knowledge of cross-group friendships and prejudice . Journal of Personality and Social Psychology, 73 , 73–90.
• Yuker, H. E., & Hurley, M. K. (1987). Contact with and attitudes toward persons with disabilities: The measurement of intergroup contact. Rehabilitation Psychology, 32 (3), 145.

From the editors

Everett (2013) presents an excellent overview of the research on Intergroup Contact Theory and how psychologists have used it to understand prejudice and conflict. As the article notes, friendship between members of different groups is one form of contact that helps dissolve inter-group conflict. Friendships are beneficial because of “self-expansion,” which is a fundamental motivational process that drives people to grow and integrate new things into their lives (Aron, Norman, & Aron, 1998). When an individual learns something or experiences something for the first time, his/her mind literally grows. When friendships are very intimate, people include aspects of their friends in their own self-concept (Aron, Aron, Tudor, & Nelson, 1991).

For example, if Scott (an American) becomes friends with Dan (a Russian), Scott might grow to appreciate Russian culture, because of their intimacy. Even the word “Russian” is now part of Scott’s own self-concept through this friendship, and Scott will have more positive feelings and attitudes toward Russians as a group. The same process happens for all kinds of other groups based race/ethnicity, religion, sexual orientation, etc.

Importantly, self-expansion and intimacy through friendship do not work like magic; psychologists can’t wave a wand and make them appear. Nor does it happen through superficial small talk (e.g., “how about this crazy weather?”). Intimacy develops through deep communication: sustained, reciprocal, escalating conversations in which two friends come to know each other in a meaningful way. A Christian person might say, “I have a Jewish co-worker” (while talking about a casual acquaintance) or a Caucasian person might say, “I give money to an organization that helps starving people in Africa” or a straight person might say, “I support same-sex marriage equality because I know someone who is gay.” All of that is good, but it’s not as effective at reducing inter-group conflict as a true friendship with someone in those other groups; superficial contact has a small effect on racism, anti-Semitism, or homophobia. A recent meta-analysis (Davies, Tropp, Aron, Pettigrew, & Wright, 2011) revealed that spending lots of time with cross-group friends and having lots of in-depth communication with those friends were the two strongest predictors of change in positive attitudes and prejudice reduction.

At In-Mind, we work in a transnational team and we think this is enriching. What about you? Have you found friendships, or even working relations, across social groups? Did this lead you to have more open or positive attitudes? Or, do you have other experiences?

article author(s)

Jim A. C. Everett

Jim Everett studied for his undergraduate at the University of Oxford, gaining a First Class degree in Psychology, Philosophy, and Physiology. He completed... more

article keywords

• discrimination
• intergroup contact hypothesis

article glossary

• intergroup conflict
• recognition
• Intergroup Contact Hypothesis
• contact hypothesis
• cooperation
• meta-analysis
• stereotypes
• stereotype threat
• field study

Intergroup Contact Theory: Recent Developments and Future Directions

• Published: 17 August 2018
• Volume 31 , pages 374–385, ( 2018 )

Cite this article

• Ashley Lytle 1  

3749 Accesses

15 Citations

Explore all metrics

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Allport, G. W. (1954). The nature of prejudice . Cambridge, MA: Addison-Wesley.

Google Scholar  

Barlow, F. K., Paolini, S., Pedersen, A., Hornsey, M. J., Radke, H. M., Harwood, J., et al. (2012). The contact caveat: Negative contact predicts increased prejudice more than positive contact predicts reduced prejudice. Personality and Social Psychology Bulletin, 38 (12), 1629–1643. https://doi.org/10.1177/0146167212457953 .

Article   PubMed   Google Scholar  

Benkler, Y., Faris, R., Roberts, H., & Zuckerman, E. (2017). Study: Breitbart-led right-wing media ecosystem altered broader media agenda. Retrieved June 2, 2018 from http://www.cjr.org/analysis/breitbart-media-trump-harvard-study.php .

Brown, R., & Hewstone, M. (2005). An integrative theory of intergroup contact. In M. P. Zanna & M. P. Zanna (Eds.), Advances in experimental social psychology (Vol. 37, pp. 255–343). San Diego, CA, US: Elsevier Academic Press. https://doi.org/10.1016/s00652601(05)37005-5 .

Chapter   Google Scholar  

Davies, K., Tropp, L. R., Aron, A., Pettigrew, T. F., & Wright, S. C. (2011). Cross-group friendships and intergroup attitudes: A meta-analytic review. Personality and Social Psychology Review, 15 (4), 332–351. https://doi.org/10.1177/1088868311411103 .

Eller, A., & Abrams, D. (2003). ‘Gringos’ in Mexico: Cross-sectional and longitudinal effects of language school-promoted contact on intergroup bias. Group Processes & Intergroup Relations, 6 (1), 55–75. https://doi.org/10.1177/1368430203006001012 .

Article   Google Scholar  

Fein, S., & Spencer, S. J. (2000). Prejudice as self-image maintenance: Affirming the self through derogating others. In C. Stangor & C. Stangor (Eds.), Stereotypes and prejudice: Essential readings (pp. 172–190). New York, NY, US: Psychology Press.

Hayward, L. E., Tropp, L. R., Hornsey, M. J., & Barlow, F. K. (2018). How negative contact and positive contact with whites predict collective action among racial and ethnic minorities. British Journal of Social Psychology, 57 (1), 1–20. https://doi.org/10.1111/bjso.12220 .

Hodson, G. (2008). Interracial prison contact: The pros for (socially dominant) cons. British Journal of Social Psychology, 47 (2), 325–351. https://doi.org/10.1348/014466607X231109 .

Hodson, G. (2011). Do ideologically intolerant people benefit from intergroup contact? Current Directions in Psychological Science, 20 (3), 154–159. https://doi.org/10.1177/0963721411409025 .

Kim, N., & Wojcieszak, M. (2018). Intergroup contact through online comments: Effects of direct and extended contact on outgroup attitudes. Computers in Human Behavior, 8163, 72. https://doi.org/10.1016/j.chb.2017.11.013 .

Lazer, D. (2015). The rise of the social algorithm: Does content curation by Facebook introduce ideological bias? Science, 348 (6239), 1090–1091. https://doi.org/10.1126/science.aab1422 .

Lemmer, G., & Wagner, U. (2015). Can we really reduce ethnic prejudice outside the lab? A meta-analysis of direct and indirect contact interventions. European Journal of Social Psychology, 45 (2), 152–168. https://doi.org/10.1002/ejsp.2079 .

Lytle, A., Dyar, C., Levy, S. R., & London, B. (2017). Essentialist beliefs: Understanding contact with and attitudes towards lesbian and gay individuals. British Journal of Social Psychology, 56 (1), 64–88. https://doi.org/10.1111/bjso.12154 .

Martens, A., Johns, M., Greenberg, J., & Schimel, J. (2006). Combating stereotype threat: The effect of self-affirmation on women’s intellectual performance. Journal of Experimental Social Psychology, 42 (2), 236–243. https://doi.org/10.1016/j.jesp.2005.04.010 .

McGregor, I., Haji, R., & Kang, S. (2008). Can ingroup affirmation relieve outgroup derogation? Journal of Experimental Social Psychology, 44 (5), 1395–1401. https://doi.org/10.1016/j.jesp.2008.06.001 .

Miles, E., & Crisp, R. J. (2014). A meta-analytic test of the imagined contact hypothesis. Group Processes & Intergroup Relations, 17 (1), 3–26. https://doi.org/10.1177/1368430213510573 .

Paolini, S., Harwood, J., & Rubin, M. (2010). Negative intergroup contact makes group memberships salient: Explaining why intergroup conflict endures. Personality and Social Psychology Bulletin, 36 (12), 1723–1738. https://doi.org/10.1177/0146167210388667 .

Pettigrew, T. F. (1998). Intergroup contact: Theory, research and new perspectives. Annual Review of Psychology, 49, 65–85. https://doi.org/10.1146/annurev.psych.49.1.65 .

Pettigrew, T. F. (2008). Future directions for intergroup contact theory and research. International Journal Of Intercultural Relations , 32 (3), 187–199. https://doi.org/10.1016/j.ijintrel.2007.12.002 .

Pettigrew, T. F., Christ, O., Wagner, U., & Stellmacher, J. (2007). Direct and indirect intergroup contact effects on prejudice: A normative interpretation. International Journal of Intercultural Relations, 31 (4), 411–425. https://doi.org/10.1016/j.ijintrel.2006.11.003 .

Pettigrew, T. F., & Tropp, L. R. (2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90 (5), 751–783. https://doi.org/10.1037/0022-3514.90.5.751 .

Pettigrew, T. F., & Tropp, L. R. (2008). How does intergroup contact reduce prejudice? Meta-analytic tests of three mediators. European Journal of Social Psychology, 38 (6), 922–934. https://doi.org/10.1002/ejsp.504 .

Reimer, N. K., Becker, J. C., Benz, A., Christ, O., Dhont, K., Klocke, U., et al. (2017). Intergroup contact and social change: Implications of negative and positive contact for collective action in advantaged and disadvantaged groups: Erratum. Personality and Social Psychology Bulletin, 43 (6), 901–902. https://doi.org/10.1177/0146167217703392 .

Rosenthal, L. (2016). Incorporating intersectionality into psychology: An opportunity to promote social justice and equity. American Psychologist, 71 (6), 474–485. https://doi.org/10.1037/a0040323 .

Rosenthal, L., & Levy, S. R. (2016). Endorsement of polyculturalism predicts increased positive intergroup contact and friendship across the beginning of college. Journal of Social Issues, 72 (3), 472–488. https://doi.org/10.1111/josi.12177 .

Saguy, T., & Kteily, N. (2014). Power, negotiations, and the anticipation of intergroup encounters. European Review of Social Psychology, 25 (1), 107–141. https://doi.org/10.1080/10463283.2014.957579 .

Saguy, T., Tausch, N., Dovidio, J. F., & Pratto, F. (2009). The irony of harmony: Intergroup contact can produce false expectations for equality. Psychological Science, 20 (1), 114–121. https://doi.org/10.1111/j.1467-9280.2008.02261.x .

Sherman, D. K., Brookfield, J., & Ortosky, L. (2017). Intergroup conflict and barriers to common ground: A self-affirmation perspective. Social and Personality Psychology Compass . https://doi.org/10.1111/spc3.12364 .

Sherman, D. K., & Hartson, K. (2011). Reconciling self-defense with self-criticism: Self affirmation theory. In M. Alicke & C. Sedikides (Eds.), The handbook of self enhancement and self-protection (pp. 128–151). New York: Guilford Press.

Turner, R. N., & Crisp, R. J. (2010). Imagining intergroup contact reduces implicit prejudice. British Journal of Social Psychology, 49 (1), 129–142. https://doi.org/10.1348/014466609X419901 .

Turner, R. N., Dhont, K., Hewstone, M., Prestwich, A., & Vonofakou, C. (2014). The role of personality factors in the reduction of intergroup anxiety and amelioration of outgroup attitudes via intergroup contact. European Journal of Personality, 28 (2), 180–192. https://doi.org/10.1002/per.1927 .

Vezzali, L., Andrighetto, L., Di Bernardo, G. A., Nadi, C., & Bergamini, G. (2017). Negative intergroup contact and support for social policies toward the minority outgroup in the aftermath of a natural disaster. The Journal of Social Psychology, 157 (4), 407–415. https://doi.org/10.1080/00224545.2016.1184126 .

Vezzali, L., Andrighetto, L., Saguy, T. (2016). When intergroup contact can backfire: The content of intergroup encounters and desire for equality (under review).

Wright, S. C., Aron, A., McLaughlin-Volpe, T., & Ropp, S. A. (1997). The extended contact effect: Knowledge of cross-group friendships and prejudice. Journal of Personality and Social Psychology, 73 (1), 73–90. https://doi.org/10.1037/0022-3514.73.1.73 .

Download references

Author information

Authors and affiliations.

College of Arts and Letters, Stevens Institute of Technology, Hoboken, NJ, USA

Ashley Lytle

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Ashley Lytle .

Ethics declarations

Ethical approval.

This article does not contain any studies with human participants or animals performed by any of the authors.

Rights and permissions

Reprints and permissions

About this article

Lytle, A. Intergroup Contact Theory: Recent Developments and Future Directions. Soc Just Res 31 , 374–385 (2018). https://doi.org/10.1007/s11211-018-0314-9

Download citation

Published : 17 August 2018

Issue Date : December 2018

DOI : https://doi.org/10.1007/s11211-018-0314-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Find a journal
• Publish with us
• Track your research

This page has been archived and is no longer being updated regularly.

All you need is contact

November 2001, Vol 32, No. 10

A longstanding line of research that aims to combat bias among conflicting groups springs from a theory called the "contact hypothesis." Developed in the 1950s by Gordon Allport, PhD, the theory holds that contact between two groups can promote tolerance and acceptance, but only under certain conditions, such as equal status among groups and common goals. Since the theory's inception, psychologists have added more and more criteria to what is required of groups in order for "contact" to work.

Recently, however, University of California, Santa Cruz research psychologist Thomas Pettigrew, PhD, has turned this research finding on its head. In a new meta-analysis of 500 studies, he finds that all that's needed for greater understanding between groups is contact, period, in all but the most hostile and threatening conditions. There is, however, a larger positive effect if some of the extra conditions are met.

His analysis turned up another unexpected finding that also runs counter to the direction of the field. The reason contact works, his analysis finds, is not purely or even mostly cognitive, but emotional.

"Your stereotypes about the other group don't necessarily change," Pettigrew explains, "but you grow to like them anyway."

Pettigrew is currently submitting his study for review; the basic findings can also be found in a chapter by him and Linda Tropp, PhD, in the book "Reducing Prejudice and Discrimination" (Erlbaum, 2000).

--T. DeANGELIS

Letters to the Editor

• Send us a letter

• Memberships

Contact Hypothesis theory explained

Contact Hypothesis Theory: this article explains the Contact Hypothesis Theory in a practical way. Next to what it is, this article also highlights the intergroup contact and prejudice, stereotypes and discrimination, the conditions and contact hypothesis examples. After reading you will understand the basics of this psychology theory. Enjoy reading!

What is the Contact Hypothesis?

The contact hypothesis is a psychology theory suggesting that prejudice and conflict between groups can be reduced by allowing members of those groups to interact with one another. This notion is also called intra-group contact. Prejudice and conflict usually arise between majority and minority group members.

The background to the contact hypothesis

Social psychologist Gordon Allport is credited with conducting the first studies on intergroup contact.

Allport is also known for this research in the field of personalities . After the Second World War, social scientists and policymakers concentrated mainly on interracial contact. Allport brought these studies together in his study of intergroup contact.

In 1954, Allport published his first hypothesis concerning intergroup contact in the journal of personality and social psychology. The main premise of his article stated that intergroup contact was one of the most effective ways to reduce prejudice between groups.

Allport claimed that contact management and interpersonal contact could produce positive effects against with stereotyping, prejudice and discrimination, leading to better and more worthwhile interaction between two or more groups.

Over the years since Allport’s original article, the hypothesis has been expanded by social scientists and used for research into reducing prejudice relating to racism, disability, women and LGBTQ + people.

Empirical and meta analytical research into intergroup contact is still ongoing today.

Intergroup contact and prejudice, stereotypes and discrimination

The term “prejudice” is used to refer to a preconceived, negative view of another person, based on perceived qualities such as political affiliation, skin colour, faith, gender, disability, religion, sexuality, language, height, education, and more.

Prejudice can also refer to an unfounded belief, or to pigeonhole people or groups. Gordon Allport, the originator of the contact hypothesis, defined prejudice as a feeling, positive or negative, prior to actual experience, that is not based on fact.

Stereotypes, as defined in the contact hypothesis, are generalisations about groups of people. Stereotypes are often based on sexual orientation, religion, race, or age. Stereotypes can be positive, but are usually negative. Either way, a stereotype is a generalisation that does not take into account differences at the individual level.

Prejudice and stereotypes concern biased views regarding others, but discrimination consists of targeted action against individuals or groups based on race, religion, gender or other identifying features. Discrimination takes many forms, from pay gaps and glass ceilings to unfair housing policies.

In recent years, more and more new legislation and regulations have been introduced, designed to tackle discrimination and prejudice reduction in, for example, the workplace. It is not however possible to eliminate discrimination through legislation. Discrimination is a complex issue relating to the justice, education and political systems in a society.

Conditions for intergroup contact to reduce prejudice

Gordon Allport claimed that prejudice and conflict between groups can be reduced by having equal status contact between groups in pursuit of common goals. This effect is even greater when contact is officially sanctioned.

This can be achieved through legislation, but also through local customs and practice. In other words, there are four conditions under which prejudice can be reduced. These are:

Equal status

Both groups taking part in the contact situations must play equal roles in the relationship. The members of each group should have similar backgrounds, qualities and other features. Differences in academic background, prosperity or experience should be kept to a minimum.

Common goals

Both groups should seek to serve a higher purpose through the relationship and working together. This is a goal which can only be achieved when the two groups join forces and work together on common initiatives.

Working together

Both groups should work together to achieve their common goals, rather than in competition.

Support from the authorities through legislation

Both groups should recognise a single authority, to support contact and collaborative interaction between groups. This contact should be helpful, considerate, and foster the right attitude towards one another.

Examples of the contact hypothesis

The effect of greater contact between members of disparate groups has been the basis of many policy decisions advocating racial integration in settings such as schools, housing, workplaces and the military.

The contact hypothesis in the desegregation of education

An example of this is a 1954 landmark court decision by the US Supreme Court. The decision brought about the desegregation of schools. In this ruling, the contact hypothesis was used to demonstrate that this would increase self-esteem among racial minorities and respect between groups in general.

Studies into the implications of this decision in subsequent years did not always yield positive results. There have been studies showing that prejudice was actually reinforced and that self-esteem did not improve among minorities. The reason for this has already been set out above.

Contact between groups in schools, for example, was not always equal, nor did it take place with social supervision. These are two essential requirements or conditions for improving relationships between disparate groups.

The contact hypothesis in developing education strategies

The contact hypothesis has also proved invaluable in developing cooperative education strategies. The best known of these is the jigsaw classroom technique. This technique involves creating a particular classroom setting where students from various racial backgrounds are brought together in pursuit of a common goal.

In practice this means that students are placed in study groups of 6. The lesson is split into six elements, and each student is assigned one part of those six. That means that each student actually represents one piece in that jigsaw.

For the lesson to succeed, students need to trust one another based on their knowledge. This increases interdependence within the group, which is necessary for improving relationships between people.

Reducing prejudice

Besides it being very important to know how prejudice arises, studies on prejudice also focus on the potential to reduce prejudice. One technique widely believed to be highly effective is training people to become more empathetic towards members of other groups.

Putting yourself in someone else’s shoes makes it easier to think about what you would do in a similar situation.

Other techniques and methods used to reduce prejudice are:

• Contact with members of other groups
• Making others aware of the inconsistencies in their beliefs and values
• Legislation and regulations which promote fair and equal treatment of people in minority groups
• Creating public support and awareness

Implication of prejudice and discrimination in the workplace

Discrimination and prejudice can lead to wellbeing issues and substantial financial loss to the organisation, along with a sharp fall in employee and company morale. According to the American Psychological Association, 61% of adults face prejudice or discrimination at some time.

For some this happens at work; others face it as part of everyday life in society. Most people are aware of the negative effects this can have on employees, but discrimination and prejudice going unchecked can also have serious consequences.

Firstly, treating people unfairly can contribute to increased stress levels. This in turn leads to more wellbeing issues for those who are personally harmed or attacked. When someone is constantly worrying about discrimination or religion, he or she is forced to think about that thing all day long. Too much stress reduces sleep quality and suppresses appetite. When this becomes the norm for someone, they are going to feel chronically ill or down.

Prejudice also has a negative effect on the company in general. Companies may even suffer financial loss as a consequence. Employees who feel ill or down because of social issues are more likely to resign. The company then incurs substantial costs training new people.

Another obvious negative outcome for organisations is employees who hate management if they feel they are not being treated fairly. This negative attitude from employees has an effect on individual employee performance and ultimately also on the performance of the organisation as a whole.

The contact hypothesis in summary

The contact hypothesis, of which the intergroup contact theory is a part, is a theory from sociology and psychology which suggests that problems such as discrimination and prejudice can be drastically reduced by having more contact with people from different social groups. This notion is also called intergroup contact. Prejudice and conflict usually arise between majority and minority group members.

The social psychologist credited for his contributions in this field is Gordon Allport. Allport brought together several studies of interracial contact after the Second World War and developed the intergroup contact theory from those. His hypothesis was published in 1954. In the decades which followed, the theory was widely used in initiatives to tackle these social problems.

Prejudice is often a negative evaluation of others based on qualities such as political affiliation, age, skin colour, height, gender or other identifying features. Stereotyping resembles prejudice, but is in fact making generalisations about groups of people.

This social failing is also based on religion, gender or other identifying features which say nothing about the group as a whole. Discrimination goes a step further than prejudice and stereotypes. Discrimination is about actually treating people in a negative way based on particular identifying features such as race or education.

Gordon Allport developed four requirements or conditions necessary for reducing prejudice through increased intergroup contact. The first is that both groups should have an equal status. The members of each group should have similar backgrounds, qualities or social status.

Differences in academic background, prosperity or experience should be kept to a minimum. The second is to have common goals. The groups should not be brought together without some purpose. As mentioned too in the example above in the jigsaw classroom section, dependence on one another is stimulating, which is a prerequisite for social equality and improved relationships. This is linked to the third condition: working together.

Now it is your turn

What do you think? Are you familiar with the explanation of the contact hypothesis? Have you ever faced prejudice or discrimination? Have you ever experienced the positive effects of contact? Had you ever heard of this theory before? Do you think eliminating discrimination and prejudice is possible? What is your view on opportunity of outcome vs opportunity of equality? Do you have any other advice or additional comments?

Share your experience and knowledge in the comments box below.

More information

• Amir, Y. (1969). Contact hypothesis in ethnic relations . Psychological bulletin, 71(5), 319.
• Brewer, M. B., & Miller, N. (1984). Beyond the contact hypothesis: Theoretical . Groups in contact: The psychology of desegregation, 281.
• Paluck, E. L., Green, S. A., & Green, D. P. (2019). The contact hypothesis re-evaluated . Behavioural Public Policy, 3(2), 129-158.
• Pettigrew, T. F., & Tropp, L. R. (2005). Allport’s intergroup contact hypothesis: Its history and influence . On the nature of prejudice: Fifty years after Allport, 262-277.

How to cite this article: Janse, B. (2021). Contact Hypothesis Theory . Retrieved [insert date] from Toolshero: https://www.toolshero.com/psychology/contact-hypothesis/

Original publication date: 05/25/2021 | Last update: 08/21/2023

Add a link to this page on your website: <a href=”https://www.toolshero.com/psychology/contact-hypothesis/”>Toolshero: Contact Hypothesis Theory</a>

Did you find this article interesting?

Your rating is more than welcome or share this article via Social media!

Average rating 4.2 / 5. Vote count: 13

No votes so far! Be the first to rate this post.

We are sorry that this post was not useful for you!

Let us improve this post!

Tell us how we can improve this post?

Ben Janse is a young professional working at ToolsHero as Content Manager. He is also an International Business student at Rotterdam Business School where he focusses on analyzing and developing management models. Thanks to his theoretical and practical knowledge, he knows how to distinguish main- and side issues and to make the essence of each article clearly visible.

Related ARTICLES

Victor Vroom biography and books

Friedrich Engels biography, quotes and books

Relative Deprivation Theory by Garry Runciman

Social Learning Theory by Albert Bandura

Respondents: the definition, meaning and the recruitment

Social Intelligence (SI) explained

Also interesting.

Cognitive Behavioural Therapy (CBT)

Herzberg Two Factor Theory of Motivation

Psychological Safety by Amy Edmondson

Leave a reply cancel reply.

You must be logged in to post a comment.

BOOST YOUR SKILLS

Toolshero supports people worldwide ( 10+ million visitors from 100+ countries ) to empower themselves through an easily accessible and high-quality learning platform for personal and professional development.

By making access to scientific knowledge simple and affordable, self-development becomes attainable for everyone, including you! Join our learning platform and boost your skills with Toolshero.

POPULAR TOPICS

• Change Management
• Marketing Theories
• Problem Solving Theories
• Psychology Theories

ABOUT TOOLSHERO

• Free Toolshero e-book
• Memberships & Pricing

• Search Menu
• Sign in through your institution
• Author Guidelines
• Submission Site
• Self-Archiving Policy
• Why Submit?
• About Journal of Computer-Mediated Communication
• About International Communication Association
• Editorial Board
• Advertising & Corporate Services
• Journals Career Network
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

The contact hypothesis reconsidered: interacting via internet, conditions and challenges of the contact hypothesis, practicality and contact, anxiety in contact, generalization from contact, the net advantage, ameliorating anxiety through online interaction, supersize it: generalizing from the contact to the group, getting more than just skin deep, beyond the cookie-cutter contact: tailoring the net contact to fit specific needs, conclusions, about the authors.

• < Previous

The Contact Hypothesis Reconsidered: Interacting via the Internet

• Article contents
• Figures & tables
• Supplementary Data

Yair Amichai-Hamburger, Katelyn Y. A. McKenna, The Contact Hypothesis Reconsidered: Interacting via the Internet, Journal of Computer-Mediated Communication , Volume 11, Issue 3, 1 April 2006, Pages 825–843, https://doi.org/10.1111/j.1083-6101.2006.00037.x

• Permissions Icon Permissions

One of the leading theories advocated for reducing intergroup conflict is the contact hypothesis. According to this theory, contact under certain conditions, such as equal status, cooperation towards a superordinate goal, and institutional support, will create a positive intergroup encounter, which, in turn, will bring about an improvement in intergroup relations. Despite its promise, the contact hypothesis appears to suffer from three major defects: (1) practicality—creating a contact situation involves overcoming some serious practical obstacles; (2) anxiety—the anxiety felt by the participants may cause a contact to be unsuccessful or at least not reach its potential; (3) generalization—the results of a contact, however sucessful, tend to to be limited to the context of the meeting and to the participants. The Internet has, in recent years, become an accessible and important medium of communication. The Internet creates a protected environment for users where they have more control over the communication process. This article suggests that the Internet’s unique qualities may help in the creation of positive contact between rival groups. The major benefits of using the Internet for contact are examined in this article.

The contact hypothesis has been described as one of the most successful ideas in the history of social psychology ( Brown, 2000 ). Allport (1954) presented the first widely-accepted outline of the contact hypothesis, claiming that true acquaintance lessens prejudice. In other words, knowledge, on its own, will not cause people to negate their prejudices and stereotypes about others, since they are very likely to accept only those pieces of information that fit into their preconceived schema of the world. It is through getting to know the other that people may be able to break down their stereotypes of him or her.

The Internet has, to date, been perhaps the most successful means of facilitating and enabling contact among individuals—particularly those who otherwise would not have had the opportunity, nor perhaps the inclination, to meet ( McKenna & Bargh, 2000 ). As we will argue here, the Internet is uniquely suited to the implementation of the various requirements of the contact hypothesis that are necessary for consistently producing successful outcomes. Indeed, the Internet may be the best tool yet for effectively putting the contact hypothesis into practice.

The article begins with a brief overview of the requirements necessary to create an ideal and successful contact situation and the challenges to putting these requirements into practice in traditional settings. We then discuss the ways in which the Internet can be used to meet these challenges and, in some areas, do so more successfully than can be achieved through traditional interaction contexts (e.g., in person, over the phone). We conclude with a short section providing ideas for how various aspects of the online contact environment and interaction process can be tweaked to fit specific situational needs and to improve the chances for a successful contact to occur.

Under ideal circumstances, when a member of a majority group meets with a minority group member and the experience is a positive one, an attitude change on two levels will result ( Allport, 1954 ). First, there will be an attitude change that is target-specific. That is, initial assumptions about the other that arise from the (negative) stereotypes associated with his or her group are replaced by more positive perceptions of the individual. Second, these new positive associations with the individual will become extended to that individual’s group as a whole, thus ameliorating negative attitudes toward the group. Allport delineated four key conditions for such a meeting: equal-group status within the situation; common goals; intergroup cooperation; and institutional support. Several other conditions were later added, the most important of these being voluntary participation and intimate contact ( Amir, 1969 , 1976 ).

There is strong empirical support demonstrating that, when effectively implemented, the conditions described above do indeed lead to a positive attitude change that is target-specific (e.g., Brown & Wade, 1987 ; Hewstone & Brown, 1986 ; Riordan & Ruggiero, 1980 ). The evidence is less clear regarding a global attitude change toward the group, however. A majority of studies do not find that the positive attitude toward the individual translates into a more positive attitude toward the group nor into more positive behavior toward other individual group members (see Hewstone & Brown, 1986 for a review). Scarberry, Ratcliff, Lord, Lanicek and Desforges (1997) demonstrate that a global attitude change can be consistently achieved, however, under carefully controlled conditions.

The contact hypothesis contains a long list of conditions for a sucessful contact. However, Pettigrew and Tropp (2000) , in their meta-analysis of contact studies, have found that it is not necessary that all of Allport’s (1954) conditions be present simultaneously for bias to be reduced. Mere contact can be a sufficient condition for bias reduction that is lasting and generalizes beyond the individuals to their larger group. Importantly, however, each of Allport’s conditions further enhances the bias-reducing effects of mere contact and thus the more conditions that are co-present, the more likely a successful and lasting outcome will be achieved.

Unfortunately, there are significant barriers to meeting many of the conditions and, indeed, even to arranging for a “mere contact” to take place. This, in turn, limits the number of contacts that actually take place. The major challenges are: (1) The practicality issue: Contact between rival groups according to the conditions required by the contact hypothesis might be very complicated to arrange and expensive to run. (2) Anxiety: Despite the fact that participation in a contact is voluntary, the high anxiety involved in the contact situation may hinder its success. (3) Generalization: How can a generalization be created from a specific contact with certain outgroup members to the outgroup as a whole? We turn to a more detailed discussions of these difficulties below.

Organizing a meeting among members of opposing groups raises both logistical and financial issues. Groups that are segregated and/or geographically distant from each other will be harder to bring together and any meeting will be more costly. Even when the different groups are geographically close, linking them may still prove an expensive undertaking. Joint holiday plans for Catholic and Protestant children in Northern Ireland are one such example ( Trew, 1986 ). As the undertaking becomes more expensive, the chances of it taking place decline. In addition, there may be barriers of language or of status. Language barriers may cause feelings of distance, misunderstanding, and miscommunication among the different groups. The issue of equal status is also problematic; in some cases rival groups are characterized by extreme status differences ( Pettigrew, 1971 ).

Intergroup interactions are often more anxiety-provoking than interpersonal ones and such anxiety may not be conducive to harmonious social relations ( Islam & Hewstone, 1993 ; Stephan & Stephan, 1985 ; Wilder, 1993 ). Intergroup anxiety is the result of anticipation of negative reactions during the intergroup encounter ( Stephan & Cookie, 2001 ; Stephan & Stephan, 1996 ). When an individual is anxious, he or she is more likely to use heuristics. Thus, if an intergroup contact produces significant levels of anxiety in the individual or individuals involved, he or she is more likely to apply stereotypes to the outgroup ( Bodenhausen, 1990 ; Bodenhausen & Wyer, 1985 ).

Wilder (1993) pointed out that when in a state of anxiety, group members are likely to ignore any disconfirming information supplied in the contact context. Under such conditions, as Wilder and Shapiro (1989) demonstrated, when a member of the outgroup behaves in a positive manner that contradicts the expectations of the other side, members of the ingroup do not alter their opinions and recall the outgroup as behaving in a manner consistent with the stereotype. In such a case, the contact between these members is unlikely to bring about any change in the group stereotype.

One of the greatest challenges to the contact hypothesis is the issue of whether or not the results of a positive contact with a member of the outgroup will be generalized further. Group saliency during the interaction appears to be of critical importance to successful generalization. However, there is much debate among researchers as to what level that salience should be. Hewstone and Brown (1986) argued that a general contact is likely to be perceived on the interpersonal level and therefore not have any impact on the intergroup level. In other words, if the individual is perceived only as an individual rather than also as a representative member of his or her group, then any attitude change will remain target-specific. They suggested that, for a positive contact to have a wider group-level impact, individual participants need to be seen as representatives of their group so that the (out)group identity is highly salient. Conversely, Brewer and Miller (1984) among others have suggested that in order for a contact to succeed, group saliency should be low.

Hamburger (1994) suggested that when the central tendency of the stereotype is the only component to be measured, a large part of the picture is ignored. He added that this component may be the most resistant to change. Thus, negative results of group generalization based solely on central tendency measures may lead to erroneous conclusions regarding the contact theory in general. The inclusion of more sensitive measurements, such as variability, will give a more accurate picture, as well as allow an investigation into the background processes. Several recent studies have demonstrated Hamburger’s suggestion that the central tendency is likely to be the more rigid component in the stereotype ( Garcia-Marques & Mackie, 1999 ; Hewstone & Hamberger, 2000 ).

Clearly, when all the necessary ingredients are present, positive and beneficial results may be obtained, but just as clearly, “getting the recipe right” to produce such an outcome may be difficult at best under traditional circumstances. Yet the major literature dealing with the contact hypothesis (for a review see Brown & Hewstone, 2005 ) fails to take into account the potential role of the Internet in helping towards the success of an intergroup contact. Below we examine the ways in which contact over the Internet may overcome the practical difficulties inherent in the creation of a face-to-face contact situation according to the conditions set out in the contact hypothesis.

Internet contact and practicality

Leveling the field.

The contact hypothesis requires that there should be equal status between the members of both groups taking part in the contact. According to McClendon (1974) , equal status increases the likelihood for perceived similarities between the groups and so enhances the likelihood for improvement in their relationship and in the reduction of stereotypes ( Pettigrew, 1971 ). Optimally, there should be both external equal status (in real life) and internal equal status (within the contact) between the people taking part in the encounter. In face-to-face encounters, even very subtle differences in manner of dress, body language, use of personal space, and the seating positions taken in the room can belie real (or perceived) status differences. As Hogg (1993) has shown, within group interactions people tend to be highly sensitive in discerning subtle cues that may be indicative of status. Online interactions have the advantage here because many, although not all, of the cues individuals typically rely on to gauge the internal and external status of others are not typically in evidence.

For instance, when contact takes place in text-based environments on the Internet, regular status symbols are not part of the interaction; “on the Internet no one knows that I am wearing a diamond necklace or have teeth missing.” This point is particularly pertinent with regard to face-to-face contact, where organizers may have gone to extraordinary lengths to ensure that all participants are of equal status only to have one arrive with a Rolex watch or a similarly inappropriate status symbol.

Even when status differences are known, electronic interaction tends to ameliorate some of the effects of status differentials. For instance, when bringing together members of two established groups, the members are likely to be well aware of the internal pecking order within their own group even if they do not have knowledge of the established hierarchy among the other group’s members. In face-to-face interactions such distinctions within the groups often quickly become apparent to all, as those who stand lower tend to speak up less often and, in ways both obvious and subtle, give deference to those with higher status within their group.

Such is not the case in electronic interactions. One aspect of electronic communications that has long been decried (e.g., Sproull & Kiesler, 1991 ) is the tendency, within organizational settings, for there to be a reduction in the usual inhibitions that typically operate when interacting with one’s superiors. In other words, existing internal status does not carry as much weight and does not affect the behavior of the group members to such an extent. Underlings are more likely to speak up, to speak “out of turn,” and to speak their mind. Thus electronic interaction makes power less of an issue during discussion which leads group members, regardless of status, to contribute more to the discussion ( Spears, Postmes, Lea, & Wolbert, 2002 ). While this can prove to be problematic within a corporate setting, it is advantageous in the present context, as the medium serves to reduce the constraining effects of status both within and between the two groups.

Connecting from afar and with the comforts of home

As noted above, organizers may face significant difficulties in arranging a meeting among individuals and groups when it comes to finding a suitable meeting place, transporting the participants involved, and compensating participants for lost time due to travel incurred. This is particularly true when the groups in question or their various members live at some distance from one another. Participation may be limited to only those members of the groups who have the financial resources and job flexibility to enable their attendance or those who live in close proximity to the meeting site, rather than to all members who have the inclination (but not the resources) to attend. Thus the size and number of contacts possible are severely restricted when such meetings are face-to-face affairs.

The advent of computer-mediated interaction has opened the doors to connection possibilities that were previously not feasible. Time differences and physical distance are no longer obstacles to bringing people “together,” at least in the developed countries of the world. Electronic meetings are neither costly to set up nor are they time consuming for the participants. All that is required is for the participants to log onto the Internet and into the virtual meeting space at the specified day and time from an office, public library, or a home computer.

Indeed, having participants engage in the contact from the privacy of their respective homes has distinct advantages. Participants are likely to feel more comfortable and less anxious in their familiar surroundings. Further, research has shown that public, as opposed to private, settings can exacerbate the activation and use of stereotypes, especially when it comes to those tied to racial prejudice (e.g., Lambert, Payne, et al., 2003 ). As Zajonc (1965) has shown, an individual’s habitual or dominant response is more likely to emerge in public settings, whereas the individual is likely to be more open and receptive to altering the habitual response when in a private sphere. Even when participants interact in quite “public” electronic venues but do so from the privacy of their homes, they tend to feel that it is a private affair (e.g., McKenna & Bargh, 2000 ; McKenna, Green, & Gleason, 2002 ). Thus, interacting electronically from home should serve to inhibit the activation of stereotypes as compared to a more public and face-to-face setting in a new environment.

Cooperation toward superordinate goals

One of the keys to a successful contact is for both sides to participate jointly in a task, the completion of which is important to both groups ( Allport, 1954 ; Miller & Harrington, 1992 ). This is especially true when cooperation between the groups will lead to successful outcomes ( Blanchard, Adelman, & Cook, 1975 ). The question then arises as to whether such an exercise may be successfully undertaken on the Internet and, if so, will the results equal or surpass those conducted in traditonal, face-to-face environments?

Today many organizations do in fact have working teams whose members are dispersed all around the world and who frequently communicate, cooperate, and complete tasks through the Internet. Successful outcomes routinely occur despite the fact that, in many cases, the team members have never met one another and are unlikely to do so. This phenomenon is known as a virtual team. This form of working is becoming increasingly common within organizations, as the benefits of including virtual teams have become more evident ( Cascio, 2000 ). For instance, employers find that telecommuting increases worker productivity and improves attendance ( Abreu, 2000 ).

To date, the evidence seems to indicate that tasks performed by virtual teams are done equally well (or equally badly) as those conducted by face-to-face work teams. Research by Dennis (1996 ; Dennis & Kinney, 1998 ) has shown that members of verbally-interacting workgroups tend to share less vital information than do members of electronic workgroups, and hence make poor decisions. Yet, members of the electronic groups also tend to make poor group decisions, despite exchanging 50% more of the vital information needed to make an optimal decision. Galegher and Kraut (1994) also found that for virtual work groups the final product was similar in overall quality to that produced by face-to-face group members.

Institutional support and willingness to participate

One of the preconditions for a contact is that participants from both sides receive institutional support ( Allport, 1954 ; Slavin, 1985 ). This is to try to ensure that the contact will have a positive influence on the wider groups represented there. This is particularly important when the differences between the groups are deep-seated or potentially explosive. In such a case, a leader might resent sending group representatives to a meeting with the outgroup and may be concerned that such a meeting would diminish his or her standing as a leader.

A related condition is that the members from both sides take part in the meeting on their own volition. If the organization has compelled its members to take part in the meeting, they are unlikely to change their stereotypes as a negative reaction to the feelings of loss of control over their freedom of association ( Stephan & Stephan, 1996 ).

Internet contact may provide a balm to some of these issues. Participating in an Internet contact may be seen as taking on less of a risk than a face-to-face contact ( Bargh & McKenna, 2004 ; McKenna & Bargh, 1998) and this may make it easier for group members to volunteer to participate and for leaders to support such a meeting.

Bridging the language barrier

One issue that arises when coordinating meetings among groups with different native tongues is that of communication. Generally, participants need to be selected who either can communicate through a common language—often one that is native to neither group (i.e., English)—or translators must be provided, which can be costly. Bringing in individuals to provide translation assistance can be problematic for other reasons, as well. For instance, there is the danger that attention will focus on the person doing the translation and not on the individual members whose opinions he or she is expressing. This can reduce the perception of the target both as an individual and as a representative member of his or her group.

Emerging software, however, will soon allow individuals interacting through a text-based environment to receive messages in their own language even though those messages were created in another. There are already a number of text translation tools currently available for use on the Internet. None have yet reached the point of refinement and accuracy needed for a successful exchange of ideas with all the nuances, but the translation programs are improving by leaps and bounds (e.g., Climent, Moré, Oliver, Salvatierra, Taulé, Sanchez, & Vallmanya, 2003 ; Coughlin, 2001 ). It will not be long before we can all speak “your” language and you can speak “ours.” This will allow for the removal of a (human) third party translator to obvious advantage and, because communication will take place in each party’s “own” language, feelings of similarity and kinship should be enhanced.

Thus, in terms of sheer practicality, there are some distinct advantages to conducting contact interactions over the Internet. Participants can readily take part from disparate locales and do so from a position of greater comfort and security than can be obtained in face-to-face meetings. Many of the most obvious status “give-aways” are not in evidence in text-based online interactions. When the status of a member is known, the nature of the communication medium tends to ameliorate the negative influence and effects that status can have on an interaction. Cooperative tasks can be conducted just as well online as they could were the participants to undertake them in person and, indeed, there may be greater willingness to take part in the online task. Finally, as our technology continues to evolve, better software tools are being developed that will enhance the meeting and interaction between participants in ways not possible in traditional settings. Beyond issues of practicality, there are additional ways that an online setting can prove advantageous; these are discussed below.

There is a growing body of evidence supporting the notion that the anxiety an interaction situation may provoke is significantly reduced when that interaction takes place through a text-based exchange on the Internet as compared to face-to-face. As noted earlier, inter-group interactions are often more anxiety-provoking than interpersonal ones, although those too can often elicit feelings of anxiousness. Anxiety increases the tendency to rely on stereotyping during an interaction, with lasting effects. Further, should an individual’s anxiety or nervousness be apparent to others, they tend to be liked and accepted less by those others (e.g., Leary, 1983 ).

However, many of the situational factors that can foster feelings of anxiety in social situations (e.g., having to respond on the spot, feeling under visual scrutiny) are absent in online interactions. Because participants have more control over how they present themselves and their views online (e.g., being able to edit one’s comments before presenting them), they should tend to feel more comfortable and in control of the situation. They should be better able to and to more often express themselves, to be liked more by their online interactions partners than if they interacted in person, and to develop closer, more intimate relationships through online interaction.

Research on those with chronically high levels of social anxiety demonstrates just that. For instance, a recent laboratory study examined small group interaction among socially anxious and nonanxious participants (see McKenna & Seidman, 2005 ). Seventy-five undergraduate students at New York University were pre-selected for this study based on their responses on the Interaction Anxiousness Scale ( Leary, 1983 ). Only those who scored at the high and low extremes of the scale were recruited for the study, and they were randomly assigned to interact in groups of three either face-to-face or in a specially-created Internet chat room. Immediately following the interaction, participants assessed how they felt during the interaction, as well as how accepted and included they felt by the other group members.

Consistent with their responses on the Interaction Anxiousness Scale, socially anxious individuals in the face-to-face condition reported feeling anxiety, shyness, and discomfort during the group interaction, while the opposite was true for non-anxious participants. In marked contrast, interacting online produced significantly different results. Participants reported feeling significantly less anxious, shy, and uncomfortable, and more accepted by their fellow group members than did those who interacted face-to-face—but these effects were wholly qualified by differences in levels of social anxiety. That is, the extremely extroverted participants felt equally comfortable, outgoing, and accepted interacting online and in person. For those experiencing high levels of social anxiety, however, the mode of communication proved pivotal to their feelings of comfort, shyness, and acceptance. Moreover, the self-reports of the socially anxious participants in the online condition on these measures were virtually identical to those of nonanxious participants in the face-to-face condition.

Those experiencing anxiety in social situations have also been found to take more active leadership roles in online groups than in their face-to-face counterparts. In a study by McKenna, Seidman, Buffardi, and Green (2005 ), participants were again preselected based on their interaction anxiety scores and randomly assigned to interact in groups of four (composed of two anxious and two nonanxious members) either in an Internet chat room or face-to-face. They then engaged in a decision-making task, following which they rated each of their interaction peers on measures of leadership, degree of participation in the discussion as compared to the other members, extroversion, and how much they liked the person based on their interaction. Peer ratings showed that socially anxious participants were as likely as their nonanxious counterparts to be perceived as leaders within the respective groups and to participate as actively when the interaction took place online. In the face-to-face condition, nonanxious participants received the leadership vote and were the more active participants. Socially anxious participants were viewed as more likeable and extroverted when they interacted online than in person, while their nonanxious counterparts were viewed as equally likeable and extroverted in both situations.

The largest hurdle to overcome is the tendency for the various members of the contact situation to come to feel quite close to one another and yet to view their new comrades from the outgroup as exceptions to their group rather than as normative representatives. Unless the members of the outgroup are perceived as representative members, the contact will have failed, for no changes in the perceived stereotype of the group as a whole will have taken place.

One of the advantages of online communication is that one can quite easily manipulate the degree of individual versus group saliency in a given contact situation in order to achieve a desired outcome. Spears et al. (2002) have argued that anonymous communication within groups leads to a sense of depersonalization by the group members. That is, members feel an absence of personal accountability and personal identity and thus the group-level identity becomes more important. When the group-level identity is thus heightened, Spears et al. (2002) have shown that group norms can have an even stronger effect than occurs in face-to-face interactions. The degree to which the group identity is salient, however, plays an important role in determining what the effects of anonymity will be on the development of group norms.

For instance, Spears, Lea, and Lee (1990) found that when members of online groups interacted under anonymous conditions and group salience was high, normative behavior increased in those groups as compared to electronic groups in which members were anonymous but the salience of the group was low. Whether group salience was high or low, participants who interacted under individuating conditions displayed an intermediate level of conformity to group norms.

One of the most interesting sets of studies examining the interaction between anonymity and identity-salience tested the effects of primed behavior in electronic groups. Postmes, Spears, Sakhel, and De Groot (2001) primed participants with either task-oriented or socioemotional behavior and then had them interact in electronic groups under either anonymous or identifying conditions. Members in the anonymous groups displayed behavior consistent with the respective prime they received considerably more so than did their counterparts who interacted under identifiable conditions within their groups. Normative behavior strengthened over time in the anonymous groups, with the members conforming even more strongly to the primed behavior. In contrast, when members were identifiable to other group members they actually bucked the norms and behaved more prime-inconsistently over time.

Further studies provided even stronger evidence of the effect that anonymity can have on normative behavior. In a study by Postmes et al. (2001) , only half of the participants in each group received the behavioral prime. In the anonymous groups, those participants who did not receive the prime nonetheless conformed to the task or socioemotional behavior being exhibited by their primed cohorts and did so significantly more than did the nonprimed participants in the identifiable groups. Further, those who interacted anonymously reported feeling a significantly stronger attachment to their group and to the other group members.

Importantly, in an online environment one can hit two birds with one stone; one can heighten the perception of the individual members as representative of their disparate groups while simultaneously fostering feelings of kinship and attachment to the “new group” composed of all members taking part in the exercise. There are a number of means by which the first can be achieved. For instance, one can provide all members with anonymous screen names that are evocative of the group they are representing (e.g., Pakistan 1, India 1, Pakistan 2) or, following Leah Thompson’s procedure (see Thompson & Nadler, 2002 ) one can have each member briefly introduce him- or herself at the beginning of the interaction and ask each to include a statement stressing his or her typicality as a member, and so forth. As the interaction in the online environment progresses, group norms will begin to quickly emerge ( Spears et al., 2002 ). These norms will be distinct from those that operate when members of group A are alone together and distinct from those unique to group B. Rather, these norms will emerge from the combined membership of groups A and B in the online setting, leading to heightened feelings of attachment and camaraderie among the participants. Thus, one can effectively invoke the necessary balance of a sense of both “us and them” among the participants that will allow for acceptance and generalization.

Cook (1962) suggested that the more intimate the relationship, the more favorable the attitude of the groups was likely to be. He stressed the importance of the “acquaintance potential,” or the opportunity provided by the situation for the contact participants to get to know each other. Recent research into the importance of personalized interaction (e.g., Miller, 2002 ) and intergroup friendships (e.g., Pettigrew, 1997 , 1998 ) has reawakened interest in this aspect of the contact. It is also has a particular relevance to this discussion on contact through the Internet.

Mutual self-disclosure is a critical component for the formation of close interpersonal bonds and the establishment of a sense of belonging and acceptance. Thus problematically, interactions between in-group and out-group members are usually conducted on a casual and superficial level. For example, Taylor, Dube, and Bellerose (1986) reported a study carried out at McGill University, an English University in a French-speaking area of Canada. Most of the students (76%) are English speakers. One might think that, given the relatively high percentage of English-speaking students, the French-speaking students would engage in more interaction with English-speaking students than with French-speaking students. However, the sample of French students reported that fully 50% of their social interactions were with other French-speaking students, a proportion notably higher (in fact, double) than their representation in the university. English-speaking students reported interacting with in-group members 87% of the time. Tellingly, participants reported that their interaction with in-group members was significantly more intimate than their interaction with out-group members.

One of the major advantages of Internet interactions over face-to-face interactions is the general tendency for individuals to engage in greater self-disclosure and more intimate exchanges there. Interactions online tend to become “more than skin deep” and to do so quite quickly (e.g., McKenna et al., 2002 ; Walther, 1996 ).

Spears and Lea (1994) suggest that it is the protection of anonymity often provided by the Internet that helps people openly to express the way they really think and feel. In line with this, McKenna and Bargh (1998, 1999 ) suggest that this sense of anonymity allows people to take risks in making disclosures to their Internet friends that would be unthinkable to them in a face-to-face interaction. More recently, McKenna, Buffardi, and Seidman (2005) found that, while people tend to engage in the greatest acts of self-disclosure when interacting under relative anonymity online, they also disclose more to their face-to-face friends and to family members when interacting with these individuals online. In other words, even without the cloak of anonymity, people more readily make intimate disclosures through their Internet interactions than through their face-to-face interactions, even when it comes to their nearest and dearest.

There are a number of unique qualities of the Internet that facilitate self-disclosure and intimacy online. As they have been discussed extensively elsewhere (see Ben Ze’ev, 2004 ; McKenna & Bargh, 2000 ; McKenna & Green, 2002 ; McKenna et al., 2002 ), we list them only briefly. They are: (1) a greater sense of anonymity or non-identifiability that leads to a reduced feeling of vulnerability and risk; (2) the absence of traditional gating features to the establishment of any close relationship—that is, easily discernable features such as physical appearance (beauty is in the eye of the beholder), mannerisms, apparent social stigmas such as stuttering, or visible shyness or anxiety; (3) a greater ease of finding others who share our specialized interests and values—and particularly so when there are a lack of “real world” counterparts (e.g., because of the marginalized or highly specialized nature of the interest, such similar others may not be present in one’s physical community or, if they are, they are not readily identifiable); and (4) more control over one’s side of the interaction and how one presents oneself.

The online environment seems to be particularly suited for “getting more than skin deep.” The results of a laboratory experiment in which undergraduates were randomly assigned (in cross-sex pairs) to meet one another for the first time in an Internet chatroom or to meet face-to-face demonstrates this quite well. McKenna et al. (2002 , Study 3) found that those who met online both liked each other more and felt that they had gotten to know one another better than did those who interacted face-to-face. This effect held when participants met one another twice, once in person and once over the Internet, unaware that it was the same interaction partner in both situations. There was a significant correlation between the degree of liking for the partner and how well the participant felt he or she had gotten to know the other person for those who met over the Internet. However, there was no such correlation in the face-to-face condition. Along similar lines, Walther (1996 , 1997) found that new acquaintances can achieve greater intimacy through online communication than they do in parallel face-to-face interactions.

One of the greatest advantages of the Internet is the ability to tailor and tweak the various requirements to achieve optimal results for specific contact situations. Rather than being limited to a single intensive meeting that takes place over a few hours or days, multiple contacts can be arranged spanning days, weeks, and even months.

The anonymity and identifiability of participants can be manipulated depending on the particular needs of the situation and can be altered over time. The salience of the originating group and that of the “new group” can be heightened or lowered as needed. As research has shown, in some situations it is beneficial for the salience of the outgroup to be quite high and in others such heightened salience is detrimental to a successful contact. The Internet contact has the ability to serve both approaches and to examine which, if either, is more beneficial to the intergroup relationship. Group identity may be emphasized on the Internet by exercising control over the contact environment or, if it is thought appropriate, may be reduced if requested. This may be especially important with groups, which have salient physical characteristics that are impossible to reduce in face-to-face contact.

It is possible for the contact organizer to create an interactive information system on the outgroup that can be accessible to the ingroup participants both before and during the contact. In preparing this, its creators should gather information on the perceptions and stereotypes held by their group of the outgroup. They can then start to tackle the main components of the stereotype. This learning system will accompany participants through the different stages of the contact. The need to learn more about the outgroup may continue to be important for the participants at different stages of the contact process, i.e., before the contact, during the meeting sessions, between meetings, and after the contact program has been completed. Importantly, the ease of receiving needed information about the outgroup on the screen at any given moment can prove a useful aid in the creation of a positive intergroup contact. The ability of the Internet to supply a learning mechanism, where the information is accessible in a wholly interactive form and so can answer the specific requests and concerns of participants, creates a learning environment of a particularly high quality ( Rosenberg, 2002 ).

Finally, when it is deemed necessary for participants to meet face-to-face for a successful contact to occur, it is possible to implement a “gradual model” approach leading up to that face-to-face contact to insure a greater chance of success. Using this model, organizers can make use of an very gradual process to help the individual become comfortable with the contact situation and the other members, and to develop strong bonds with those other members before they ever meet in person. The main steps in this graded contact are as follows:

Communicating by text only: This text-only interaction is the most common form of communication over the Internet. This stage will continue until the participant feels secure in this form of contact and his/her anxiety levels are negligible.

Text + image: Participants will continue to use the text method with which they feel secure, but will simultaneously view a live video image of the person with whom they are interacting. When low-level social anxiety has been established, participants will transfer to the next stage.

Communicating by video + audio: At this stage, people will still interact from their secure environment and still without physical proximity to their conversation partner. However, use of text messages by the subject will be reduced; instead he/she will communicate orally. In addition, a live image of the subject will be transferred to the other participant. Again, when a satisfactory level of comfort has been achieved, participants may progress to the next stage.

Face-to-face interaction: This is the stage of regular face-to-face interaction. It is predicted that this process will successfully bridge the gap between text-only Internet contact and total exposure through a face-to-face encounter, and do so in a way that continually preserves low levels of anxiety among participants. As research by McKenna and colleagues has shown ( Bargh et al., 2002 ; Mckenna et al., 2002 ), when interactions first begin over the Internet and then move to a face-to-face environment, participants not only like one another more than they would were they to have initially begun their interaction in person, but when the face-to-face meeting does take place it serves to heighten already strong feelings of liking and kinship.

The Internet has an enormous potential for providing tools to create effective intergroup contact. Its unique characteristics provide an excellent basis for such a contact, for example, by creating a secure environment, reducing anxiety, cutting geographical distances, significantly lowering costs, and by creating equal status, intimate contact,and cooperation. In addition, it offers the chance to receive approval from the authorities. The Internet is also a major information resource, and its ability to answer questions and provide knowledge in real time makes it a uniquely useful tool for the promotion of intergroup communication. The Internet may be said to provide opportunities for a successful contact that are superior to those provided in a traditional face-to-face meeting.

There are clearly potential obstacles in putting together a contact through cyberspace; and while taking this fully into account, it is our belief that contact schemes over the Internet may prove exceptionally effective tools in the pursuit of improved interpersonal and intergroup relations. One of the mechanisms that may be developed to support such schemes is found in the field of interactive learning systems which interact with each user individually. In addition, future research should reveal more information about different factors which affect such a contact; for example, the impact of personality on the Internet contact ( Amichai-Hamburger, 2002 ).

Despite the questions that remain as yet unanswered, we believe that the advantages of using the Internet for an outgroup contact are exceptionally promising, and we advocate the introduction of the Internet as a vital part of contact between groups.

Abreu , S . ( 2000 ). How to manage telecommuters . Retrieved April 27, 2006, from http://cnnstudentnews.cnn.com/2000/TECH/computing/06/19/telecommuting.idg/

Allport , G. W . ( 1954 ). The nature of prejudice . Cambridge, MA: Addison-Wesley .

Google Scholar

Google Preview

Amichai-Hamburger , Y . ( 2002 ). Internet and personality . Computers in Human Behavior , 18 ( 1 ), 1 – 10 .

Amir , Y . ( 1969 ). Contact hypothesis in ethnic relations . Psychological Bulletin , 71 ( 5 ), 319 – 342 .

Amir , Y . ( 1976 ). The role of intergroup contact in change of prejudice and ethnic relations . In P. A. Katz (Ed.), Towards the elimination of racism (pp. 73 – 123 ). New York: Plenum Press .

Bargh , J. A. , & McKenna , K. Y. A . ( 2004 ). The Internet and social life . Annual Review of Psychology , 55 , 573 – 590 .

Bargh , J. A. , McKenna , K. Y. A. , & Fitzimons , G. M . ( 2002 ). Can you see the real me? Activation and expression of the “true self” on the Internet . Journal of Social Issues , 58 ( 1 ), 33 – 48 .

Ben-Ze’ev , A . ( 2004 ). Love online: Emotions on the Internet . Cambridge, UK: Cambridge University Press .

Blanchard , F. A. , Adelman , L. , & Cook , S. W . ( 1975 ). Effect of group success and failure upon interpersonal attraction in cooperating interracial groups . Journal of Personality and Social Psychology , 31 ( 6 ), 1020 – 1030 .

Bodenhausen , G. V . ( 1990 ). Stereotypes as judgmental heuristics: Evidence of circadian variations in discrimination . Psychological Science , 1 ( 5 ), 319 – 322 .

Bodenhausen , G. V. , & Wyer , R. S . ( 1985 ). Effects of stereotypes on decision making and information-processing strategies . Journal of Personality and Social Psychology , 48 ( 2 ), 267 – 282 .

Brewer , M. B . ( 1979 ). In-group bias in the minimal intergroup situation: A cognitive-motivational analysis . Psychological Bulletin , 86 ( 2 ), 307 – 324 .

Brewer , M. B. , & Miller , N . ( 1984 ). Beyond the contact hypothesis: Theoretical perspectives on desegregation . In M. Hewstone & R. J. Brown (Eds.), Contact and conflict in intergroup encounters (pp. 281 – 302 ). Oxford: Blackwell .

Brown , R. J . ( 2000 ). Group processes: Dynamics within and between groups (2nd ed.). Oxford: Blackwell .

Brown , R. , & Hewstone , M . ( 2005 ). An integrative theory of intergroup contact . In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 37, pp. 255 – 343 ). San Diego, CA: Academic Press .

Brown , R. J. , & Wade , G. S . ( 1987 ). Superordinate goals and intergroup behaviour: The effects of role ambiguity and status on intergroup attitudes and task performance . European Journal of Social Psychology , 17 ( 2 ), 131 – 142 .

Cascio , W. E . ( 2000 ). Managing virtual workplace . The Academy of Management Executive , 14 ( 3 ), 81 – 90 .

Climent , S. , Moré , J. , Oliver , A. , Salvatierra , M. , Taulé , M. , Sanchez , I. , & Vallmanya , L . ( 2003 ). Bilingual newsgroups in Catalonia: A challenge for machine translation . Journal of Computer-Mediated Communication , 9 (1) . Retrieved April 28, 2006, from http://jcmc.indiana.edu/vol9/issue1/climent.html

Cook , S. W . ( 1962 ). The systematic analysis of socially significant events: A strategy for social research . Journal of Social Issues , 18 ( 2 ), 66 – 84 .

Coughlin , D . ( 2001 ). Correlating automated and human assessments of machine translation quality . In Proceedings of MT Summit IX (pp. 63 – 70 ). New Orleans, LA.

Dennis , A. R . ( 1996 ). Information exchange and use in group decision making: You can lead a group to information but you can’t make it think . MIS Quarterly , 20 ( 4 ), 433 – 455 .

Dennis , A. R. , & Kinney , S. T . ( 1998 ) Testing media richness theory in the new media: The effects of cues, feedback, and task equivocality . Information Systems Research , 9 ( 3 ), 256 – 274 .

Galegher , J. , & Kraut , R. E . ( 1994 ). Computer-mediated communication for intellectuall teamwork: An experiment in group writing . Information Systems Research , 5 ( 2 ), 110 – 138 .

Garcia-Marques , L. , & Mackie , D. M . ( 1999 ). The impact of stereotype incongruent information on perceived group variability and stereotype change . Journal of Personality and Social Psychology , 77 ( 5 ), 979 – 990 .

Hamburger , Y . ( 1994 ). The contact hypothesis reconsidered: Effects of the atypical outgroup member on the outgroup stereotype . Basic Applied Social Psychology , 15 ( 3 ), 339 – 358 .

Hewstone , M. , & Brown , R. J . ( 1986 ). Contact is not enough: An intergroup perspective on the Contact Hypothesis . In M. Hewstone & R. J. Brown (Eds.), Contact and conflict in intergroup encounters (pp. 1 – 44 ). Oxford: Blackwell .

Hewstone , M. , & Hamberger , J . ( 2000 ). Perceived variability and stereotype change . Journal of Experimental Social Psychology , 36 ( 2 ), 103 – 124 .

Hogg , M. A . ( 1993 ). Group cohesiveness: A critical review and some new directions . European Review of Social Psychology , 4 , 85 – 111 .

Islam , M. R. , & Hewstone , M . ( 1993 ). Dimensions of contact as predictors of inter-group anxiety, perceived out-group variability, and out-group attitude: An integrative model . Personality and Social Psychology Bulletin , 19 ( 6 ), 700 – 710 .

Lambert , A. J. , Payne , B. K. , Shaffer , L. M. , Jacoby , L. L. , Chasteen , A. , & Khan , S . ( 2003 ). Stereotypes as dominant responses: On the “social facilitation” of prejudice in anticipated public contexts . Journal of Personality and Social Psychology , 84 ( 2 ), 277 – 295 .

Leary , M. R . ( 1983 ). Social anxiousness: The construct and its measurement . Journal of Personality Assessment , 47 ( 1 ), 66 – 75 .

McClendon , M. J . ( 1974 ). Interracial contact and the reduction of prejudice . Sociological Focus , 7 ( 4 ), 47 – 65 .

McKenna , K. Y. A. , & Bargh , J. A . ( 1999 ). Causes and consequences of social interaction on the Internet: A conceptual framework . Media Psychology , 1 , 249 – 269 .

McKenna , K. Y. A. , & Bargh , J. A . ( 2000 ). Plan 9 from cyberspace: The implications of the Internet for personality and social psychology . Personality and Social Psychology Review , 4 ( 1 ), 57 – 75 .

McKenna , K. Y. A. , Buffardi , L. E. & Seidman , G . ( 2005 ). Self presentation to friends and strangers online . In K.-H. Renner , A. Schutz , & F. Machilek (Eds.), Internet and personality (pp. 176 – 189 ). Göttingen, Germany: Hogrefe & Huber Publishers .

McKenna , K. Y. A. , & Green , A. S . ( 2002 ). Virtual group dynamics . Group Dynamics: Theory, Research and Practice , 6 ( 1 ), 116 – 127 .

McKenna , K. Y. A. , Green , A. S. , & Gleason , M . ( 2002 ). Relationship formation on the Internet: What’s the big attraction? Journal of Social Issues , 58 ( 1 ), 9 – 31 .

McKenna , K. Y. A. , & Seidman , G . ( 2005 ). Social identity and the self: Getting connected online . In W. R. Walker & D. Hermann (Eds.), Cognitive technology: Essays on the transformation of thought and society . Jefferson, NC: MacFarland and Company, Inc .

McKenna , K. Y. A. , Seidman , G. , Buffardi , L. E. , & Green , A. S. (in preparation). Social anxiety and self-expression to friends and strangers in Internet and face-to-face interactions .

Miller , N . ( 2002 ). Personalization and the promise of contact theory . Journal of Social Issues , 58 ( 2 ), 387 – 410 .

Miller , N. , & Harrington , H. J . ( 1992 ). Social categorization and intergroup acceptance: Principles for design and development of cooperative learning teams . In R. Hertz-Lazarowitz & N. Miller (Eds.), Interaction in cooperative groups. The theoretical anatomy of group learning (pp. 203 – 227 ). Cambridge, UK: University Press .

Pettigrew , T. F . ( 1971 ). Racially separate or together? New York: McGraw-Hill .

Pettigrew , T. F . ( 1997 ). The affective component of prejudice: Empirical support for the new view . In S. A. Tuch & J. K. Martin (Eds.), Racial attitudes in the 1990s: Continuity and change (pp. 76 – 90 ). Westport, CT: Praeger .

Pettigrew , T. F . ( 1998 ). Intergroup contact theory . Annual Review of Psychology , 49 , 65 – 85 .

Pettigrew , T. F. , & Tropp , L. R . ( 2000 ). Does intergroup contact reduce prejudice? Recent meta-analytic findings . In S. Oskamp (Ed.), Reducing prejudice and discrimination: Social psychological perspectives (pp. 93 – 114 ). Mahwah, NJ: Erlbaum .

Postmes , T. , Spears , R. , Sakhel , K. , & De Groot , D . ( 2001 ). Social influence in computer-mediated communication: The effects of anonymity on group behavior . Personality and Social Psychology Bulletin , 27 ( 10 ), 1243 – 1254 .

Riordan , C. , & Ruggiero , J . ( 1980 ). Producing equal-status interaction: A replication . Social Psychology Quarterly , 43 ( 1 ), 131 – 136 .

Rosenberg , M. J . ( 2002 ). E-learning: Strategies for delivering knowledge in the digital age . New York: McGraw Hill .

Scarberry , N. , Ratcliff , C. , Lord , C. G. , Lanicek , D. , & Desforges , D. M . ( 1997 ). Effects of individuating information on the generalization part of Allport’s contact hypothesis . Personality and Social Psychology Bulletin , 23 ( 12 ), 1291 – 1299 .

Slavin , R. E . ( 1985 ). Cooperative learning: Applying theory in desegregated schools . Journal of Social Issues , 41 ( 3 ), 45 – 62 .

Spears , R. , & Lea , M . ( 1994 ). Panacea or panopticon? The hidden power in computer-mediated communication . Communication Research , 21 ( 4 ), 427 – 459 .

Spears , R. , Lea , M. , & Lee , S . ( 1990 ). De-individuation and group polarisation in computer-mediated communication . British Journal of Social Psychology , 29 ( 2 ), 121 – 134 .

Spears , R. , Postmes , T. , Lea , M. , & Wolbert , A . ( 2002 ). When are net effects gross products? The power of influence and the influence of power in computer-mediated communication . Journal of Social Issues , 58 ( 1 ), 91 – 107 .

Sproull , L. , & Faraj , S . ( 1997 ). Atheism, sex, and databases: The net as a social technology . In S. Kiesler (Ed.), Culture of the Internet (pp. 35 – 51 ). Hillsdale, NJ: Erlbaum .

Sproull , L. , & Kiesler , S . ( 1991 ). Connections: New ways of working in the networked organization . Cambridge, MA: MIT Press .

Stephan , W. G. , & Cookie , W . ( 2001 ). Improving intergroup relations . Newbury, CA: Sage .

Stephan , W. G. , & Stephan , C. W . ( 1985 ). Intergroup anxiety . Journal of Social Issues , 41 ( 3 ), 157 – 175 .

Stephan , W. G. , & Stephan , C. W . ( 1996 ). Cognition and affect in stereotyping: Parallel interactive networks . In D. W. Mackie & D. L. Hamilton (Eds.), Affect cognition and stereotyping: Interactive processes in group perception (pp. 111 – 136 ). Orlando, FL: Academic Press .

Taylor , M. , Dube , L. , & Bellerose , J . ( 1986 ). Intergroup contact in Quebec: Myth or reality? In M. Hewstone & R. Brown (Eds.), Contact and conflict in intergroup encounters (pp. 107 – 118 ). Oxford: Blackwell .

Thompson , L. , & Nadler , J . ( 2002 ). Negotiating via information technology: Theory and application . Journal of Social Issues , 58 ( 1 ), 109 – 124 .

Trew , K . ( 1986 ). Catholic-Protestant contact in Northern Ireland . In M. Hewstone & R. J. Brown (Eds.), Contact and conflict in intergroup encounters (pp. 93 – 106 ). Oxford: Blackwell .

Walther , J. B . ( 1996 ). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction . Communication Research , 23 ( 1 ), 3 – 43 .

Walther , J. B . ( 1997 ). Group and interpersonal effects in international computer-mediated collaboration . Human Communication Research , 23 ( 3 ), 342 – 369 .

Wilder , D. A . ( 1993 ). The role of anxiety in facilitating stereotypic judgments of out-group behavior . In D. M. Mackie & D. L. Hamilton (Eds.) Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 87 – 109 ). San Diego, CA: Academic Press .

Wilder , D. A. , & Shapiro , P . ( 1989 ). Role of competition-induced anxiety in limiting the beneficial impact of positive behavior by an out-group member . Journal of Personality and Social Psychology , 56 ( 1 ), 60 – 69 .

Zajonc , R. B . ( 1965 ). Social facilitation . Science , 149 , 269 – 274 .

Yair Amichai-Hamburger is a social-industrial psychologist, and an assistant professor in the Psychology Department, Bar-Ilan University, Israel. He has written widely on the impact of the Internet on well-being and has recently edited The Social Net: Understanding Human Behavior in Cyberspace . Oxford University Press.

Address: Department of Psychology, Bar-Ilan University, Ramat-Gan 52900, Israel

Katelyn Y. A. McKenna (Yael Kaynan) is a Senior Lecturer at Ben-Gurion University and a lecturer at The Interdisciplinary Center Herzliya in the departments of Communication. Her research interests are in the areas of relationship cognition, the self, and social identity, particularly in terms of their applicability to Internet interactions.

Address: Bldg 72, Room 550, Department of Communication Studies, Ben-Gurion University of the Negev, Beer Sheva, 84105, Israel

Email alerts

Citing articles via.

• Recommend to Your Librarian
• Advertising and Corporate Services

Affiliations

• Online ISSN 1083-6101
• Copyright © 2024 International Communication Association
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Online ordering is currently unavailable due to technical issues. We apologise for any delays responding to customers while we resolve this. For further updates please visit our website: https://www.cambridge.org/news-and-insights/technical-incident

We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings .

Login Alert

• > Journals
• > Behavioural Public Policy
• > Volume 3 Issue 2
• > The contact hypothesis re-evaluated

Article contents

What we have learned since pettigrew and tropp: a 10-year retrospective and update, a timeline of developments in the contact literature, assembling studies for meta-analysis, intergroup contact studies: who, what, where and when, meta-analytic methods, meta-analytic results, the contact hypothesis re-evaluated.

Published online by Cambridge University Press:  10 July 2018

• Supplementary materials

This paper evaluates the state of contact hypothesis research from a policy perspective. Building on Pettigrew and Tropp's (2006) influential meta-analysis, we assemble all intergroup contact studies that feature random assignment and delayed outcome measures, of which there are 27 in total, nearly two-thirds of which were published following the original review. We find the evidence from this updated dataset to be consistent with Pettigrew and Tropp's (2006) conclusion that contact “typically reduces prejudice.” At the same time, our meta-analysis suggests that contact's effects vary, with interventions directed at ethnic or racial prejudice generating substantially weaker effects. Moreover, our inventory of relevant studies reveals important gaps, most notably the absence of studies addressing adults' racial or ethnic prejudices, an important limitation for both theory and policy. We also call attention to the lack of research that systematically investigates the scope conditions suggested by Allport (1954) under which contact is most influential. We conclude that these gaps in contact research must be addressed empirically before this hypothesis can reliably guide policy.

For more than a century, researchers have sought to understand what causes people to harbor and express prejudice against outgroups. Sustained attention to the topic of prejudice reflects the fact that in every era and region, stereotyping, discrimination and xenophobia manifest themselves in ways that contribute to social inequality and sometimes erupt into intergroup violence.

Policy-makers have historically looked to social science for guidance about how to reduce prejudice (Myrdal, Reference Myrdal 1944 ), and social scientists themselves have sought to conduct research on prejudice that could inform programs and policies. In 2006, the American Psychological Association resolved to “call upon psychologists to use findings from relevant psychological research on prejudice, stereotyping and discrimination to inform their research, practice, training and education … [and] to inform anti-prejudice, anti-stereotyping and anti-discrimination positions in public and organizational policy” (APA, 2006 , p. 308).

Among the many prominent theories in this domain (for a review, see Paluck & Green, Reference Paluck and Green 2009 ), the promotion of intergroup contact has arguably become the foremost strategy for reducing prejudice. The palliative effect of intergroup contact is a central theme of Gordon Allport's landmark book, The Nature of Prejudice ( Reference Allport 1954 ), which drew its inspiration from earlier studies suggesting that housing and workplace desegregation in the United States reduced prejudice toward black people (Williams, Reference Williams 1947 ; Mussen, Reference Mussen 1950 ). Although skeptical of the notion that any form of contact diminishes prejudice, Allport conjectured that prejudice

may be reduced by equal status contact between majority and minority groups in the pursuit of common goals. The effect is greatly enhanced if this contact is sanctioned by institutional supports (i.e., by law, custom, or local atmosphere), and provided it is of a sort that leads to the perception of common interests and common humanity between members of the two groups. (p. 281)

The so-called contact hypothesis set in motion decades of research assessing whether and under what conditions intergroup contact diminishes hostility toward outgroups.

A crucial turning point in the prejudice reduction literature came in 2006 with the publication of Pettigrew and Tropp's ( Reference Pettigrew and Tropp 2006 ) influential review of more than 500 studies of the effects of intergroup contact. Their widely cited meta-analysis provided evidence so decisive that the authors concluded “[t]here is little need to demonstrate further contact's general ability to lessen prejudice. Results from the meta-analysis conclusively show that intergroup contact can promote reductions in intergroup prejudice” (p. 751). As Hewstone ( Reference Hewstone 2003 ) put it, thanks to “the Herculean labors of Pettigrew and Tropp,” we may now answer the question of whether contact reduces prejudice “with an emphatic ‘yes’” (p. 352). This conclusion is echoed in an array of social psychology articles and textbooks that describe intergroup contact as a “clearly demonstrated” method for reducing intergroup hostility (Yablon, Reference Yablon 2012 , p. 250).

The Pettigrew and Tropp meta-analysis is also noteworthy for what it did not find. The four special conditions – equal status between the groups in the situation; common goals; intergroup cooperation; and the support of authorities, law or custom (p. 752) – that Allport believed made for propitious intergroup contact received relatively little empirical support. Evidently, “Allport's conditions are not essential for intergroup contact to achieve positive outcomes. In particular … samples with no claim to these key conditions still show significant relationships between contact and prejudice” (p. 766). These effects were said to extend beyond direct exposure, which further magnified the policy implications of intergroup contact:

Indeed, the generalization of contact's effects appears to be far broader than what many past commentators have thought. Not only do attitudes toward the immediate participants usually become more favorable, but so do attitudes toward the entire outgroup, outgroup members in other situations, and even outgroups not involved in the contact. This result enhances the potential of intergroup contact to be a practical, applied means of improving intergroup relations. (p. 766)

A decade has passed since the publication of Pettigrew and Tropp's meta-analysis. Many new studies have been conducted in the meantime, some of them quite elegant and well-powered. How would the meta-analysis look today if the literature were brought up to date and expanded to include pertinent research conducted outside psychology? In this paper, we set ourselves the task of updating the original meta-analysis and broadening its disciplinary scope.

A second and central aim of this paper is to attend specifically to policy-relevant studies that speak to the practical applications of intergroup contact. Scholars who have looked to the prejudice-reduction literature as a guide to public policy have lamented its common design limitations. Non-experimental studies are prone to bias due to well-known threats to validity (Campbell & Stanley, Reference Campbell and Stanley 1963 ), and research has specifically found that the positive correlation between contact and non-prejudiced behavior in observational data can be explained by less-prejudiced individuals seeking contact (Bertrand & Duflo, Reference Bertrand and Duflo 2017 ). Among experimental studies, another limitation is that researchers tend to focus on outcomes that can be measured immediately after an intervention. A policy-maker might reasonably ask whether the effects of contact endure for days, weeks, months or years; yet as Abrams ( Reference Abrams 2010 ) observes in his report to the UK's Equality and Human Rights Commission, “there is a dearth of good-quality longitudinal research on prejudice or prejudice reduction” (p. 68).

When searching for policy-relevant findings among the hundreds of studies that comprise the contact literature to date, we used the following two critiques of the literature as criteria for determining whether a research study was capable of generating actionable research findings: first, did the study assign a contact intervention randomly, allowing for unbiased causal inference about the effects of intergroup contact? Second, were outcomes measured at least one day after the contact intervention began? Footnote 1 Testing whether intervention effects endure beyond the first day of engagement is a minimum policy standard of efficacy. This requirement also reflects the greater stock we put in studies that separate the experimental intervention process from the measurement process. Of the hundreds of studies we reviewed, only 27 experiments track post-intervention outcomes for at least one day. Notably, just 11 of these studies focus on contact across racial or ethnic lines, which has been a concern of courts and policy-makers from the start of intergroup contact research.

We review this select group of studies both qualitatively and quantitatively. After describing the evolution of the contact literature and our procedures for identifying relevant studies, our qualitative review attends to important design nuances, such as details of the contact interventions, the contexts in which they were launched and the outcomes measured. Our quantitative analysis assesses the statistical robustness of the meta-analytic estimates given various coding and estimation choices. (Our database, replication programs and archive of digitized reprints Footnote 2 are publicly available so that readers may retrace our steps.)

The results we obtain are much more equivocal than those presented by Pettigrew and Tropp. Our analysis reveals that effects vary significantly by the type of prejudice addressed. This finding runs counter to one of the main findings from Pettigrew and Tropp's ( Reference Pettigrew and Tropp 2006 ) meta-analysis: namely that racial and ethnic prejudices are affected to approximately the same degree as other prejudices (p. 762). Furthermore, the literature has some important gaps; for example, not one study assesses the effects of interracial contact on people older than 25. Given the narrow scope and mixed findings of the policy-relevant contact literature, we conclude that the jury is still out regarding the contact hypothesis and its efficacy as a policy tool. In particular, we note that the scope conditions suggested by Allport ( Reference Allport 1954 ) under which contact is more likely to be effective have not been systematically investigated.

The birth of the contact hypothesis

The history of the contact hypothesis begins in the early 20th century, when American social scientists initiated empirical studies of the effects of intergroup contact. For example, Williams ( Reference Williams 1934 ) measured the effects of a series of activities involving white and black women in the Young Women's Christian Association aged 14–18, including a field trip and a buffet dinner. Smith ( Reference Smith 1943 ) arranged for a “four-day seminar in Negro Harlem” (p. 26) on black cultural life and accomplishments for white students at Teacher's College, followed by a social tea at which Harlem residents were guests and speakers.

Following the desegregation of the military and other institutions after World War II, the social psychologist Gordon Allport ( Reference Allport 1954 ) distilled ideas about the benefits of intergroup interaction and friendship into a testable proposition. His core idea was that contact reduces prejudice. But, wary that contact could merely affirm existing social group hierarchies, Allport also proposed the set of conditions described above, under which contact should be especially influential. His work grew in prominence with the United States' struggles with desegregation. Many social scientists took up the call to test the contact hypothesis in the service of legal and public policy questions about the effects of school, neighborhood and workplace desegregation (Cook, Reference Cook 1985 ).

Thus, at its inception, the contact hypothesis served both an academic and a policy agenda. In the service of theory, the contact hypothesis characterized prejudice as a product of fear, ignorance, hierarchy or a lack of shared life patterns and goals. In the service of policy, the contact hypothesis has been proposed as a rationale for desegregation policies (Mussen, Reference Mussen 1950 ; Pettigrew, Reference Pettigrew 1979 ), as a guide for designing peacebuilding interventions (Kelman, Reference Kelman 1998 ; Maoz, Reference Maoz 2010 ) and as a theoretical narrative for interpreting the persistence of discrimination and interracial conflict (for an overview of the literature, see Pettigrew, Reference Pettigrew 2016 ). Hundreds of studies followed, gauging the relationship between intergroup prejudice and interaction across racial, ethnic, religious and other group lines.

The canonical meta-analytic result describing the effect of contact on prejudice

When Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ) assembled the contact literature, they counted 515 studies, dating from the 1940s through the year 2000, comprising “slightly more than 250,000 participants from 38 different countries” (Pettigrew et al. , Reference Pettigrew, Tropp, Wagner and Christ 2011 , p. 16). Approximately half of these studies focused on racial or ethnic divisions; the rest investigated prejudice against groups including the mentally and physically handicapped, the elderly, political partisans, and gays and lesbians.

While all studies resulted in some type of empirical estimate of the effect of contact, the studies varied widely in terms of their research designs. Seventy-one percent of the meta-analysis database consisted of observational surveys of broad populations. In one widely cited study, Pettigrew ( Reference Pettigrew 1997 ) surveyed 3806 people in France, Great Britain, The Netherlands and West Germany in 1988. Controlling for seven covariates, Pettigrew found that self-reported contact with members of immigrant outgroups was strongly associated with more positive evaluations of those groups.

Another 24% of the meta-analysis database consisted of observational intervention studies designed to assess outcomes among those experiencing intergroup contact and comparison groups who did not experience contact. For example, Lazar et al. ( Reference Lazar, Gensley and Orpet 1971 ) studied the effects of a four-week curriculum unit developed by one teacher about “creative Americans” for a class of high-IQ children. As part of the curriculum, students interacted with people with physical disabilities and with a special education teacher. The treatment class's scores on an Attitudes Toward Disabled Persons (ATDP) survey were compared to one control classroom whose students were similar in terms of age, IQ and prior attitudes.

Just 5% of the database employed an experimental design. Within that subset, contact interventions, target groups, outcomes and settings vary widely. In a college setting, Pagtolun-an and Clair ( Reference Pagtolun-an and Clair 1986 ) had a gay man answer questions about homosexuality for 90 minutes in a “deviant behavior class” (p. 125) and then post-tested students within the hour. The 35 students who experienced this form of contact with the speaker displayed statistically significant reductions in homophobia vis-a-vis an untreated control group. In one of a handful of experimental studies conducted outside the laboratory or classroom, DiTullio ( Reference DiTullio 1982 ), a special education job coordinator for the school district of Philadelphia, studied the effects of a job-training program that integrated adolescents with intellectual disabilities into custodial positions in Philadelphia elementary schools. Among coworkers and supervisors of the adolescents, this experimentally induced contact induced more positive attitudes toward individuals with intellectual disabilities across a battery of measures.

The next ten years: adding to the meta-analytic database

The value of Pettigrew and Tropp's ( Reference Pettigrew and Tropp 2006 ) monumental collection of studies is beyond dispute. Taken together, the studies provide rich descriptions of contact experiences, develop new approaches to measuring attitudes toward stigmatized versus dominant groups and illustrate the many contexts within which intergroup contact may occur (e.g., within a programmatic intervention setting, an exchange program, a school setting or an incidental encounter in a community). Analyzed as a whole, they provide evidence for an association between contact and reduced prejudice that is robust to substantial variation across time, place and subjects.

However, the value of this collection of studies is less clear in one particular respect: for understanding whether contact causes policy-relevant reductions in prejudice. The vast majority (95%) of studies do not randomly assign contact; of those that do, just eight measure outcomes at least a single day after treatment. Of those eight, three study interracial contact. Thus, evidence for whether contact's effects on racial prejudice persist – the focus of policy and legal work on intergroup contact research and advocacy – is sparse.

In the 10 years since Pettigrew and Tropp's meta-analysis, contact research has entered a more methodologically sophisticated era in which social scientists are paying attention to new and re-emerging issues of research design, analysis and transparency. None of Pettigrew and Tropp's experimental studies, for instance, feature a pre-analysis plan or open-access data. Subsequently, three studies featuring one or both have been published (Broockman & Kalla, Reference Broockman and Kalla 2016 ; Finseraas & Kotsadam, Reference Finseraas and Kotsadam 2017 ; Scacco & Warren, Reference Scacco and Warren 2018 ).

As an example of recent developments, consider two high-quality experiments conducted after Pettigrew and Tropp's ( Reference Pettigrew and Tropp 2006 ) meta-analysis. Scacco and Warren ( Reference Scacco and Warren 2018 ) provided 16 weeks of small computer training classes for low-income Christian and Muslim men in northern Nigeria. Classes were randomly assigned to be religiously homogenous or mixed. The authors found that contact produced “no changes in prejudice,” and that while subjects in heterogeneous classes discriminated less than those in homogeneous classes, this was attributable to “increased discrimination by homogeneous-class subjects” (p. 1) relative to those who had not taken the class. In an American university context, Page-Gould et al. ( Reference Page-Gould, Mendoza-Denton and Tropp 2008 ) brought white and Latinx students into an immersive laboratory friendship-building experience over the course of three consecutive weeks, randomly assigning students to work with a same- or cross-group student partner. In the 10 days following the final session, the authors found statistically insignificant and substantively small effects on participant likelihood of initiating cross-group interaction, although the effects were somewhat larger among participants who scored high on a pre-treatment test of implicit prejudice. Overall, the authors found “benefits of cross-group friendship, particularly among people who are most likely to experience anxiety in intergroup contexts” (p. 1089).

We reassess two core propositions about intergroup contact in light of these and other studies. First, we assess whether contact reduces prejudice. Second, we assess whether Allport's original moderating conditions shape the extent to which contact reduces prejudice.

Following Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ), we “define intergroup contact as actual face-to-face interaction between members of clearly defined groups” (p. 754). To update the universe of relevant studies, we sought all studies that met this definition, randomly assigned contact, had delayed outcome measures and were published (or available as working papers) by July 2016. Next, we summarized the resulting set of studies qualitatively and conducted a meta-analysis using methods similar to those of Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ).

Assembling the collection of studies

First, we identified all studies in Pettigrew and Tropp's database that randomly assigned intergroup contact and measured outcomes more than a day after treatment. To do so, we cross-referenced each study that Pettigrew and Tropp classified as experimental with the bibliography provided in their subsequent book, When Groups Meet: The Dynamics of Intergroup Contact ( Reference Pettigrew, Tropp, Wagner and Christ 2011 ). After removing studies that did not have over-time outcome measures, were mislabeled as randomly assigned, did not feature “actual face-to-face interaction” or did not have a non-contact control group, we were left with eight research reports comprising nine experiments on intergroup contact.

Second, we incorporated studies cited by other recent literature reviews and meta-analyses. For example, Lemmer and Wagner ( Reference Lemmer and Wagner 2015 ) compiled every contact and “imagined contact” study taking place in the field through 2012 that measured outcomes more than one month after treatment; this collection furnished an additional four randomized controlled trials. Footnote 3 A literature review on anti-prejudice interventions (Paluck & Green, Reference Paluck and Green 2009 ) provided three studies comprising four samples, and an unpublished literature review of interracial roommate pairings (Green, Reference Green 2014 ) provided four studies. Lastly, a review of sexual discrimination (Tucker & Potocky-Tripodi, Reference Tucker and Potocky-Tripodi 2006 ) provided one study.

Third, we informally canvassed intergroup contact researchers, discussing our project with, among others, Linda Tropp, Thomas Pettigrew, Kristin Davies and Ryan Enos, as well as attending conferences and reading relevant journals. This furnished four additional studies; those studies' citations led to two more.

Fourth, we searched Google Scholar for all studies that cited Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ) and had the words ‘random’, ‘assign’ and ‘contact’ somewhere in the text, which revealed one further study, bringing our final sample to 27 studies and 31 treatment arms.

Looking within the group of studies that comprise this meta-analysis, we now ask: who are the participants, what were the treatments, where did they take place and when? By answering these questions, we attempt a richer qualitative description of this evidence than the numbers alone can provide.

Who: participants and types of prejudice

First, whose prejudices are being studied? And who are the targets of the prejudice under study? Table 1 summarizes our universe of cases along these two dimensions.

Table 1. Participants and targets of prejudice

Participants

Thirteen of the 27 experiments study college students, and all but one of these experiments took place in the USA. The exception is Burns et al. ( Reference Burns, Corno and La Ferrara 2015 ), who studied students at the University of Cape Town. Scacco and Warren ( Reference Scacco and Warren 2018 ) examined young adults in Nigeria of college age who were not in college.

Of the six studies of adults over 25 years of age, three took place outside of the USA: one in a housing complex in Hyderabad, India (Barnhardt, Reference Barnhardt 2009 ) and two studies with Norwegian military recruits (Finseraas et al. , Reference Finseraas, Johnsen, Kotsadam and Torsvik 2016 ; Finseraas & Kotsadam, Reference Finseraas and Kotsadam 2017 ). In the USA, Dessel ( Reference Dessel 2010 ) studied teachers in an evangelical Christian community; DiTullio ( Reference DiTullio 1982 ) studied custodial teams in Philadelphia; and Broockman and Kalla ( Reference Broockman and Kalla 2016 ) canvassed residents of Miami, Florida.

Elementary and middle-school students participated in studies in the USA (Katz & Zalk, Reference Katz and Zalk 1978 ; Meshel & McGlynn, Reference Meshel and McGlynn 2004 ) and Australia (Clunies-Ross & O'Meara, Reference Clunies-Ross and O'Meara 1989 ), and high-school students participated in Israel (Yablon, Reference Yablon 2012 ), Germany (Krahe & Altwasser, Reference Krahe and Altwasser 2006 ) and the USA (Sheare, Reference Sheare 1974 ; Green & Wong, Reference Green and Wong 2009 ).

Types of prejudice

Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ) point out that the contact hypothesis was

originally developed to address racial and ethnic prejudices, but recent decades have witnessed a massive use of the theory for a range of different target groups. Is this expansion of contact theory justified? And do these nonracial and nonethnic samples yield meta-analytic patterns that are similar to those for racial and ethnic samples? (p. 762)

Their meta-analysis seems to provide a resounding ‘yes’: average effects of contact are strikingly similar across a range of target groups, and confidence intervals always overlap (see their Table 11, p. 764). Like their database, ours features a preponderance of studies focusing on ethnic or racial prejudice: 11 out of 27, or 40%. Camargo et al. ( Reference Camargo, Stinebrickner and Stinebrickner 2010 ), Marmaros and Sacerdote ( Reference Marmaros and Sacerdote 2006 ), Katz and Zalk ( Reference Katz and Zalk 1978 ) and Sayler ( Reference Sayler 1969 ) addressed relations between blacks and whites in the USA, and Burns et al. ( Reference Burns, Corno and La Ferrara 2015 ) addressed relations between whites and blacks in South Africa. Four other studies (Boisjoly et al. , Reference Boisjoly, Duncan, Kremer, Levy and Eccles 2006 ; Green & Wong, Reference Green and Wong 2009 ; Markowicz, Reference Markowicz 2009 ; Sorensen, Reference Sorensen 2010 ) tested relations among blacks, whites and members of other groups, such as Asians, Latinxs and “Native Hawaiian and Other Pacific Islander” (Markowicz, Reference Markowicz 2009 , p. 66). Lastly, Page-Gould et al. ( Reference Page-Gould, Mendoza-Denton and Tropp 2008 ) assessed relations between whites and Latinxs, and Furuto and Furuto ( Reference Furuto and Furuto 1983 ) assessed relations between white and “Polynesian and Oriental” (p. 153) students at Brigham Young University – Hawaii. As Table 1 shows, all of these studies were conducted with populations from elementary school through college.

All other categories of prejudice, discrimination and stigma are addressed by four or fewer studies, and yet contain a great deal of demographic and geographic heterogeneity. One study addressed discrimination against foreign nationals in the USA (Hull, Reference Hull 1972 ), and another against immigrants in Norway (Finseraas & Kotsadam, Reference Finseraas and Kotsadam 2017 ). Three studies tested prejudice, discrimination and stigma against LGBT individuals: either transgender people (Broockman & Kalla, Reference Broockman and Kalla 2016 ) or gays and lesbians (Grutzeck & Gidycz, Reference Grutzeck and Gidycz 1997 ; Dessel, Reference Dessel 2010 ).

The three studies examining contact between religious groups selected Hindus and Muslims in India (Barnhardt, Reference Barnhardt 2009 ), Christians and Muslims in Nigeria (Scacco & Warren, Reference Scacco and Warren 2018 ) and Jews and Arabs in Israel (Yablon, Reference Yablon 2012 ). Footnote 4 Four studies targeted prejudice against the intellectually disabled; three of those took place in the USA (Hall, Reference Hall 1969 ; Sheare, Reference Sheare 1974 ; DiTullio, Reference DiTullio 1982 ) and one in Australia (Clunies-Ross & O'Meara, Reference Clunies-Ross and O'Meara 1989 ). Prejudice against people with physical disabilities was studied in the USA (Evans, Reference Evans 1976 ) and Germany (Krahe & Altwasser, Reference Krahe and Altwasser 2006 ). Finally, we note one study targeting prejudice against the elderly (Meshel & McGlynn, Reference Meshel and McGlynn 2004 ) and one targeting discrimination against women (Finseraas et al. , Reference Finseraas, Johnsen, Kotsadam and Torsvik 2016 ).

What: interventions and measurements

What kind of contact did the study authors randomize across participants? Some contact was crafted by researchers, while other types of contact were more naturalistic; some contact was sustained, while other engagements were very brief. We also describe the variety of outcomes measured following the intervention. Most outcomes were self-reported attitudes and social evaluations, while behavioral outcomes mostly included observed interactions and measures of friendship with members of the other group. Outcome measures also varied in whether they focused on reduced prejudice toward the outgroup involved in the study or on general levels of social tolerance.

What type of contact?

The contact interventions can roughly be characterized as falling along two interrelated dimensions: scriptedness , or the degree to which experimenters control and direct the treatment content and whether they employ confederates as a means to steer the contact experience (e.g., Evans, Reference Evans 1976 ); and duration , which ranges from brief and impersonal exposure to sustained and intimate contact.

In an example of a brief, unscripted encounter, Hall ( Reference Hall 1969 ) examined University of Alabama students who were gathered together with residents of an institution for people with severe intellectual disabilities and encouraged to pair up or assemble groups to sing songs or practice skills like tying shoes. Far more common are brief, scripted interactions. These typically take place in a laboratory or classroom and involve a structured conversation or an activity with a member of a presumed outgroup. Evans ( Reference Evans 1976 ), for instance, randomly assigned college students ( n  = 40) to have one of two types of conversation with a blind woman. In one, they were asked to discuss their hometowns, majors and family; in the other, the woman who was blind explicitly invited questions about blindness. Katz and Zalk ( Reference Katz and Zalk 1978 ) assigned interracial and racially homogenous groups of second and fifth graders to work on puzzles for 15 minutes under observation by their teachers.

Scripted and sustained interventions are typically designed around intergroup dialogue and excursion interventions. Yablon ( Reference Yablon 2012 ) studied the effects of six monthly meetings between Palestinian and Israeli high-school students in which they discussed social issues and concluded with a joint trip to an amusement park. Dessel ( Reference Dessel 2010 ) led discussion groups between straight teachers and LGB volunteers over the course of two months for nine hours in total; Sorensen ( Reference Sorensen 2010 ) and Markowicz ( Reference Markowicz 2009 ) each studied the effects of interracial dialogues held at universities.

Sustained, unscripted intergroup encounters featured extensive contact in a naturalistic environment that researchers cannot directly control or sometimes even monitor. In general, they follow Allport's ( Reference Allport 1954 ) argument that to reduce prejudice, intergroup contact experiences “should occur in ordinary purposeful pursuits, [and] avoid artificiality” (p. 489). Ten studies targeted intergroup living situations, such as interracial college roommates, ranging in duration from a weekend (Hull, Reference Hull 1972 ) to eight weeks (Finseraas et al. , Reference Finseraas, Johnsen, Kotsadam and Torsvik 2016 ; Finseraas & Kotsadam, Reference Finseraas and Kotsadam 2017 ) to a year (Boisjoly et al. , Reference Boisjoly, Duncan, Kremer, Levy and Eccles 2006 ; Camargo et al. , Reference Camargo, Stinebrickner and Stinebrickner 2010 ) or more (Barnhardt, Reference Barnhardt 2009 ).

Another way to describe the type of contact in these interventions is to ask whether they fit the conditions that Allport specified as critical for prejudice reduction. Very few interventions fit all four conditions. Most interventions, because they needed approval to be launched, are characterized by authority approval of the contact (26 out of 27 – all but Broockman & Kalla, Reference Broockman and Kalla 2016 ). Seventeen studies could be characterized as featuring equal status contact, 14 feature cooperation and 12 have a common goal between groups in contact. Several of the interventions, however, are difficult to characterize according to Allport's classification scheme. Given a general lack of detailed description about the interventions, it was particularly difficult to determine whether an intervention involved equal status or a common goal. Roommate studies and more generally naturalistic and sustained interventions are also challenging, given that conditions of cooperation or equal status likely fluctuate over time. Naturalistic studies are also likely to involve some amount of negative contact experiences, like misunderstandings or outright conflict, which could affect outcomes.

What kinds of outcome measures?

Because our collection of studies spans six decades, four continents and a host of different demographic groups, it is not surprising that outcome measures range widely. We group outcome measurements into four broad categories: (1) explicit evaluations of the outgroup; (2) political and cultural attitudes commonly associated with prejudice (e.g., opinions about affirmative action); (3) behavioral measures of actions toward the outgroup, such as white subjects' numbers of black friends or the percentage of all emails that white subjects sent to black peers; and (4) indirect or projective measures of prejudices such as implicit attitude tests or the evaluation of hypothetical vignettes.

Explicit evaluations of the outgroup typically took the form of a series of evaluative questions. Such outcomes were common in studies of prejudice against people with intellectual disabilities; Hall ( Reference Hall 1969 ), for instance, asked participants to rate the “mentally retarded along a clean–dirty axis” (p. 31). Studies of ethnic, racial and religious prejudice also featured explicit outcome measurements in settings where it is (or was) more common to express outright hostility toward an outgroup, such as the USA in the 1960s (Sayler, Reference Sayler 1969 ), contemporary Nigeria (between Christians and Muslims; Scacco & Warren, Reference Scacco and Warren 2018 ) and India (between Hindus and Muslims; Barnhardt, Reference Barnhardt 2009 ).

Experimenters sometimes used more oblique measures to elicit prejudiced attitudes. Some focused on political and cultural attitudes, soliciting opinions about affirmative action (Boisjoly et al. , Reference Boisjoly, Duncan, Kremer, Levy and Eccles 2006 ) or policies discriminating against transgender people (Broockman & Kalla, Reference Broockman and Kalla 2016 ). This category also includes measures of nationalism and world-mindedness (Hull, Reference Hull 1972 ) and general beliefs about the extent of racial privilege in the USA (Markowicz, Reference Markowicz 2009 ).

Other experimenters used behavioral indicators to track interactions with outgroup members. Marmaros and Sacerdote ( Reference Marmaros and Sacerdote 2006 ) unobtrusively tracked how many emails Dartmouth students sent to white and black peers, including and excluding their own roommates. Camargo et al. ( Reference Camargo, Stinebrickner and Stinebrickner 2010 ) asked white students at Berea College with and without black roommates to report how many black friends they have, again with and without their roommates included. Page-Gould et al. ( Reference Page-Gould, Mendoza-Denton and Tropp 2008 ) included a daily diary report following intervention of how likely participants are to initiate a cross-group interaction. A novel behavioral measure of social distance between whites and blacks comes from Katz and Zalk ( Reference Katz and Zalk 1978 ). After 15 minutes of cross-group interaction, children were asked to place a variety of felt objects on a flannel board, with either a black or white examiner standing to one side of the board; the outcome measure was “literally the average distance the subject placed the five forms from the examiner” (p. 451).

Finally, a minority of studies (four) gathered evidence typically tested in social scientific laboratories: behavioral games (Scacco & Warren, Reference Scacco and Warren 2018 ), the implicit attitude test (IAT; Barnhardt, Reference Barnhardt 2009 ; Burns et al., Reference Burns, Corno and La Ferrara 2015 ) and a vignette experiment (Finseraas et al. , Reference Finseraas, Johnsen, Kotsadam and Torsvik 2016 ). Footnote 5 Scacco and Warren ( Reference Scacco and Warren 2018 ) used behavioral games to assess cooperation and trust between Nigerian Christians and Muslims, specifically dictator and destruction games in which individuals allocated real money to their real partners. The vignette experiment in Finseraas et al. ( Reference Finseraas, Johnsen, Kotsadam and Torsvik 2016 ) varied the qualifications of female and male officers in the Norwegian military to measure proclivity to discriminate based on gender.

An additional feature of outcome measurement is whether a dependent variable pertains to a particular outgroup or toward outgroups in general. For example, Green and Wong ( Reference Green and Wong 2009 ) measured tolerance toward a variety of groups, and Markowicz ( Reference Markowicz 2009 ) investigated awareness of racial privilege in the USA. Most studies in our sample focus on how prejudice reduces discrimination toward the group to which the treatment subjects were directly exposed.

Of the 27 studies, 19 (70%) took place in the USA. Of the remaining eight, two were located in Norway and one study was located in each of the following countries: India, South Africa, Australia, Germany, Nigeria and Israel.

Of the seven studies with child or adolescent subjects, five took place in a classroom or in the context of a school-required activity such as a field trip; the two in naturalistic settings took place during after-school activities with the elderly (Meshel & McGlynn, Reference Meshel and McGlynn 2004 ) or during an outdoor hiking expedition (Green & Wong, Reference Green and Wong 2009 ).

Of the 13 studies with college students, six took place in a naturalistic setting like a dormitory, a neighborhood center (Sayler, Reference Sayler 1969 ) or a living facility for the severely intellectually disabled (Hall, Reference Hall 1969 ). Three others were structured, monitored intergroup discussions (Hull, Reference Hull 1972 ; Markowicz, Reference Markowicz 2009 ; Sorensen, Reference Sorensen 2010 ); two were studies that students enrolled as part of a class (Evans, Reference Evans 1976 ; Page-Gould et al. , Reference Page-Gould, Mendoza-Denton and Tropp 2008 ); and one took place in a normal class lecture (Grutzeck & Gidycz, Reference Grutzeck and Gidycz 1997 ). The 13th study took place at Brigham Young University – Hawaii and consisted of 14 weekly “spiritual, cultural and social experiences” in integrated settings, both on and off campus (Furuto & Furuto, Reference Furuto and Furuto 1983 , p. 150).

The breakdown by decade of study is shown in Table 2 . The majority of evidence for our review was generated since 2000.

Table 2. Contact studies by decade. Note that we place Meshel and McGlynn ( Reference Meshel and McGlynn 2004 ) in the 1990s and Dessel ( Reference Dessel 2010 ) in the 2000s, going by the initial publication of the relevant data in the authors' dissertations (Meshel, Reference Meshel 1997 ; Dessel, Reference Dessel 2008 )

This section describes the criteria used to identify the key outcome variable in each study and the procedures used to transform each study's reported results into the inputs for our meta-analysis.

Selecting dependent variables

Some studies in our sample reported a single outcome (Hull, Reference Hull 1972 ), while others reported dozens (Burns et al. , Reference Burns, Corno and La Ferrara 2015 ). Some tested multiple subgroups, such as Broockman and Kalla ( Reference Broockman and Kalla 2016 ), whose pre-analysis plan specified looking for heterogeneous treatment effects by party identification, and Scacco and Warren ( Reference Scacco and Warren 2018 ), who tested for contact effects on both Christians and Muslims. Some experiments involve multiple, conceptually distinct treatment arms, such as Boisjoly et al. ( Reference Boisjoly, Duncan, Kremer, Levy and Eccles 2006 ), who measured separately the effects of having a black roommate and having a non-black minority roommate. Others varied the intensity of one treatment, such as Barnhardt ( Reference Barnhardt 2009 ), who measured the effects of having one, two or three Muslim or Hindu households in one's four-household living unit.

Some outcome measurements are composites of multiple subscales (Sayler, Reference Sayler 1969 ) or multiple items intended to evaluate feelings toward the outgroup (DiTullio, Reference DiTullio 1982 ). One study delineated a ‘main outcome’ (Finseraas & Kotsadam, Reference Finseraas and Kotsadam 2017 ), while others present a collection of response variables with no ranking system. We sought to apply consistent rules for choosing which outcomes to demarcate as representative of a study's overall findings so that we could condense each paper's findings down to a single estimate and accompanying standard error. We decided on the following rules:

• First, we chose estimates evaluating the highest dosage (experimentally varied intensity) of contact whenever possible.

• Second, when estimates are split by dominant versus subordinate or majority versus minority groups, we chose estimates evaluating prejudice reduction among the dominant or majority groups.

• Third, we considered studies to have multiple treatment arms if they met any of the following criteria: (a) featured one ingroup exposed to multiple distinct outgroups; (b) measured the effects of contact on multiple participant groups; or (c) featured one intervention across multiple, distinct settings. Our meta-analysis includes one effect size for each treatment arm.

• Fourth, when studies look at contact between two groups in conflict, in which neither is clearly dominant, we chose effect sizes that measure changes across both populations.

• Fifth, we chose the prejudice outcome on which the author(s) focused primarily.

• Sixth, if there were multiple post-tests, we chose the latest possible post-test.

• Seventh, when faced with a choice among estimators, we chose linear estimators so that we could express treatment effects in terms of standardized units.

• Eighth, when multiple econometric specifications were present, we chose the specification that estimated the treatment effect with the smallest apparent standard error.

After selecting our dependent variables, we next turned to converting them to a common framework.

Creating a common statistical framework

The most common analytic strategy for meta-analyses in the social sciences is to calculate standardized mean difference, commonly referred to as Cohen's d , defined in Cooper et al. ( Reference Cooper, Hedges and Valentine 2009 , p. 226) as

in which μ 1 and μ 2 represent the population averages of the treatment and control groups and σ is the sample standard deviation.

Estimating the numerator of equation (1) is straightforward, whereas there is some debate about how to estimate the denominator. Lemmer and Wagner ( Reference Lemmer and Wagner 2015 ) follow the recommendation of Morris ( Reference Morris 2008 ) to pool standard deviations from the treatment and control groups at pre-treatment as

Using pre-treatment information to estimate population standard deviation has the advantage of not making any additional distributional assumptions about the effects of treatment. However, pre-treatment standard deviations are not available for all of the studies in our sample. To keep comparisons constant across studies, we standardized all changes associated with treatment by the standard deviation of the control group, a statistic commonly called Glass's Δ. Footnote 6 After standardizing effect sizes for each study, we calculated standard errors for each, correcting for bias arising in small studies using Hedge's G correction factor (Cooper et al. , Reference Cooper, Hedges and Valentine 2009 ).

Two studies (Hall, Reference Hall 1969 ; Katz & Zalk, Reference Katz and Zalk 1978 ) and one treatment arm of a study (Sayler, Reference Sayler 1969 ) did not provide enough information about sampling variability to compute standardized effect sizes. We exclude these studies, representing a total of four treatment arms, from our meta-analysis, although they remain relevant for the sign tests that we conduct below. This left us with 25 studies comprising 27 treatment arms.

A graphical overview of results from our 27 comparisons can be found in Figure 1 . This scatterplot depicts the relationship between the effect estimate (vertical axis) and its standard error (horizontal axis) for each of the 27 effect estimates in our analysis. To aid interpretation, we color-coded each study according to whether it focuses on prejudice against ethnic or racial groups, religious groups, immigrants, people with mental or physical disabilities, the elderly, women or LGBT individuals. The participant pool for each study is also indicated according to the polygon used to represent each observation. Study participants who are children are depicted with circles, teenagers with triangles, college students and young adults with squares and adults aged 25 years or older with diamonds.

Figure 1. Standard error and effect size (Glass's Δ ) of each experimental comparison. Colors of points and of labels correspond to target of intervention; shapes of plotted points correspond to population. Fitted line is ordinary least squares and gray bands are 95% confidence intervals. Boisjoly ‘A’ and ‘B’ refer, respectively, to effects associated with having black or ‘other minority roommates’

This graphical overview of the core studies underscores several noteworthy features of the contact literature. First, effect sizes vary considerably by target group, with substantially larger effects observed in studies that target prejudice toward those with disabilities. Second, four out of six studies with adult subjects are clustered on the bottom left of the figure, reflecting both smaller effect sizes and standard errors than the collection of studies on average. Third, vertical positioning of the points indicates that the overwhelming majority of experiments (24 out of 27) report a positive effect of contact. The probability of observing 24 positive estimates out of 27 studies is less than 0.001 under the null hypothesis of no effect. Footnote 7 The distribution of estimated effects seems to offer strong support to the hypothesis that the types of contact facilitated in these studies led to reductions in prejudice.

However, Figure 1 also suggests that caution is warranted when summarizing the results via meta-analysis. The regression line that passes through the points calls attention to the fact that studies with smaller standard errors tend to report weaker effects than studies with larger standard errors. In other words, the larger the study, the smaller the standard error and the smaller the estimated effect. This pattern is symptomatic of a ‘file drawer problem’, in which studies are more likely to be reported when they show significant results (Rosenthal, Reference Rosenthal 1979 ). In light of this pattern, our meta-analysis considers not only the pooled study average effect, but also the study average that would be forecasted as the standard errors tend toward zero.

We begin our quantitative analysis by assessing cross-study heterogeneity using Cochran's Q . The test decisively rejects the null hypothesis of homogeneity of effects across studies ( Q (26) = 173.178, p  < 0.001; I 2  = 0.85). We therefore reject the fixed-effects meta-analysis model in favor of a random-effects meta-analysis model, where the variance of the normal random component is estimated using method of moments. The resulting estimate is 0.39, with a 95% confidence interval ranging from 0.231 to 0.554. This pooled estimate of the effect of contact on prejudice suggests that, on average, the contact induced by these experiments reduced prejudice by more than a third of the standard deviation in the control group.

The pooled average, however, glosses over the heterogeneous effects found in the experimental literature. This heterogeneity is illustrated in Figure 2 , which displays the results of the meta-analysis in the form of a study-by-study forest plot, where the studies have been sorted by their estimated standard errors. If a single causal parameter were at work in these studies, one would expect 95% of the experiments' confidence intervals to overlap with 0.39. In fact, only 20 of the 27 studies produce 95% confidence intervals that overlap with this pooled estimate ( p  < 0.001). The problem of coverage is especially acute among larger studies, which tend to produce relatively precise estimates: of the 10 studies with the smallest standard errors, four have 95% confidence intervals that fall below the overall meta-analytic average effect.

Figure 2. Forest plot of standard errors and effect sizes (ES). Areas of squares correspond to weight given to each study; lines represent 95% confidence intervals (CI). Dotted line is a random effects estimate of average effect ( Δ  = 0.39); solid line is an effect size of 0. Studies are sorted by the inverse of their standard errors

One way to model between-study heterogeneity is to allow effects to vary as a function of their standard errors, on the grounds that publication bias inflates the average effects observed in smaller studies. In this meta-regression model, the slope represents the expected change in effect size when the standard error increases by one unit. The intercept is also of interest, as it represents the expected effect if the standard error were zero (i.e., if the study were of infinite size). The results presented in Table 3 suggest that a one-unit increase in standard error is associated with a 2.09-unit increase in effect size, although the pattern is of borderline significance (two-tailed p -value = 0.049). Notably, the intercept is –0.014 with a 95% confidence interval ranging from –0.46 to 0.43. The implication is that a very large study would be expected to produce a minuscule increase in prejudice. Like the tests for publication bias presented by Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 , p. 758), our results are statistically equivocal. The same may be said for our analysis of p -hacking (Simonsohn et al. , Reference Simonsohn, Nelson and Simmons 2014 ; Head et al. , Reference Head, Holman, Lanfear, Kahn and Jennions 2015 ), which is symptomatic of research discretion that favors significant relationships. These tests may be found in the Online Supplementary Materials.

Table 3. Relationship between standard error and effect size

Another way to model effect heterogeneity is to focus on the targets of prejudice in these studies. Here, we regress effect size on indicator variables for disabilities, gender, LGBT status and age with racial, ethnic, religious and immigrant targets at the base category. In contrast to the rather muted degree of heterogeneity found by Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ), we find some prejudices to be much more responsive than others. An F -test indicates significant heterogeneity in effects across target groups ( p  = 0.01). Especially significant is the contrast between disability and the base category, where p  = 0.001. When ethnic, racial, religious and immigrant studies are considered on their own (i.e., the estimated intercept), the estimated effect is 0.25, with a 95% confidence interval ranging from 0.08 to 0.42. This remains a fairly strong and significant pooled effect, although it is subject to the proviso that four of the five largest studies come in below this average.

In sum, meta-analysis offers qualified support for the contact hypothesis. On the one hand, the overwhelming majority of studies report positive effects, and a random-effects model suggests that the true underlying effect is substantively quite large. On the other hand, the collection of studies has three important limitations. First is the gap in coverage. We know little about the effects of contact on adults over 25 years of age. In particular, the meta-analysis furnishes no evidence about contact's effects on adults' racial or ethnic prejudices, which was the original policy-based motivation for this body of work. Second, the larger experiments tend to produce weaker effects, which suggests that a file drawer problem may be concealing smaller studies with more equivocal findings. Finally, effect sizes vary significantly according to the target of prejudice, suggesting that certain kinds of prejudice are more amenable to contact-based remediation.

Robustness check: including seven borderline studies

Many studies fall just short of the selection criteria we used to identify the most policy-relevant research. Footnote 8 Some studies, for instance, do not use a fully randomized design, but instead capitalize on quasi-experiments. van Laar et al. ( Reference van Laar, Levin, Sinclair and Sidanius 2005 ) studied interracial roommate pairings at University of California, Los Angeles, which we did not include because of uncertainty about randomness of roommate assignment in this particular context. Other studies assign something similar, but not exactly identical, to intergroup contact; Enos ( Reference Enos 2014 ), for instance, randomly assigned physical proximity to outgroup members (Mexican nationals living in the USA) without assigning face-to-face interaction, and Fuegen ( Reference Fuegen 2000 ) varied whether an experimental confederate identifying as a feminist displayed stereotype-confirming or -disconfirming behaviors.

When we augment our meta-analysis using seven studies that fall in this category, our results remain largely unchanged. Random-effects meta-analysis renders an overall estimate of 0.373 (standard error = 0.075), which is very similar to what we obtained above. We continue to find significant heterogeneity in effects across target groups ( p  = 0.0024) and marginally significant evidence that treatment effects diminish as studies' standard errors decrease ( p  = 0.058).

Robustness check: studies with pre-analysis plans

Another way to test for the presence of publication bias is to look separately at studies that meet the very highest standards of experimental quality and research transparency. In our sample, three studies – Broockman and Kalla ( Reference Broockman and Kalla 2016 ), Scacco and Warren ( Reference Scacco and Warren 2018 ) and Finseraas and Kotsadam ( Reference Finseraas and Kotsadam 2017 ) – have pre-analysis plans Footnote 9 . As Olken ( Reference Olken 2015 , p. 69) writes, “[f]or readers, referees, editors, and policy-makers, knowing that analysis was pre-specified offers reassurance that the result is not a choice among many plausible alternatives, which can increase confidence in results.”

The relationship between effect size and study quality has played a central role in the assessment of contact's effects. Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ) contend that “research rigor is routinely associated with larger effect sizes. Put differently, the less rigorous studies sharply reduce the overall relationships observed between contact and prejudice” (p. 759). Revisiting the same theme a decade later, Pettigrew ( Reference Pettigrew 2016 ) writes: “the most rigorous studies tend to provide the largest effects. This phenomenon is repeated in 21st-century research. Recent work is more rigorously executed and yields larger contact effects than earlier work” (p. 14). In their meta-analysis of the effects of contact on sexual prejudice, Smith et al. ( Reference Smith, Axelton and Saucier 2009 ) hypothesized that “those studies with higher methodological quality will have more scientific rigor which will produce results with stronger effects than those studies with lower methodological quality” (p. 181).

In our sample, however, we find that studies conforming to the very highest standards of research quality Footnote 10 show much smaller effects on average than the sample as a whole. Studies with pre-analysis plans have a random effects estimate of 0.016. The studies without pre-analysis plans, by contrast, have a random effects estimate of 0.451. Given broader discussions of replicability in science (Ioannidis, Reference Ioannidis 2005 ; Nosek et al. , Reference Nosek 2015 ), this divergence is of particular concern for policy-makers interested in assessing the reliability of contact as a policy tool.

In reviewing the contact literature, previous authors have lamented a dearth of high-quality designs. Tucker and Potocky-Tripodi ( Reference Tucker and Potocky-Tripodi 2006 ), who reviewed the effects of contact on prejudice against gays and lesbians in the USA, concluded that “[n]o intervention met the criteria of a well established or probably efficacious treatment, as all studies had substantial methodological limitations” (p. 176). Yuker ( Reference Yuker 1994 ) reviewed studies of discrimination against people with disabilities and argued that the “general quality of research … is not very high. Many studies suffer from faults such as inadequate sampling, the lack of adequate control groups, failure to randomly assign subjects to groups, the lack of pretests or retrospective pretests, etc.” (p. 4). Speaking generally, Hopkins et al. ( Reference Hopkins, Reicher and Levine 1997 ) wrote, “the initial hopes of contact theorists have failed to materialize” (p. 306). Footnote 11 Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 , p. 752) explicitly challenged these earlier literature reviews on the grounds that they assembled and analyzed the literature in an unsystematic manner.

To prepare our review, we spent years attempting to gather all of the contact studies that used high-quality research designs. Our assessment of the policy-relevant contact literature falls somewhere between the pessimistic accounts and Pettigrew and Tropp's ( Reference Pettigrew and Tropp 2006 ) declaration that “meta-analytic results provide substantial evidence that intergroup contact can contribute meaningfully to reductions in prejudice across a broad range of groups and contexts” (p. 766).

On the one hand, the vast majority of these experiments do indeed show positive effects of contact. Of the 27 experimental comparisons that seem most policy relevant, 24 reveal positive effects. The average effect across these experiments is substantively large, diminishing measured prejudice by 0.39 standard deviations. Both results are statistically significant at the 0.001 level, and the inclusion of seven quasi-contact experiments studies does not materially affect the size or significance of the meta-analytic estimates.

On the other hand, five features of the contact literature give us pause. First, the set of policy-relevant studies has important gaps. What we know about prejudice reduction comes largely from studies of children or young adults. Few studies address prejudice in adults over 25 years of age. Notably, no studies of ethnic or racial contact include participants over 25 years of age.

Second, the extent to which contact diminishes prejudice seems to vary according to the target of prejudice. Contact seems to work especially well as a strategy for reducing prejudice toward people with mental or physical disabilities. Prejudice toward individuals with disabilities may differ from other types of prejudice due to the distinctive ways in which disabled people are perceived (Fiske, Reference Fiske 2011 ). When studies involving disabilities are excluded, the meta-analytic estimate remains significant but diminishes to 0.20. This finding suggests a rather different theoretical interpretation from the one offered by Pettigrew and Tropp ( Reference Pettigrew and Tropp 2006 ), who found that “[c]omparisons across the racial and ethnic subsets and the nonracial and nonethnic subsets yield virtually identical mean estimates of contact-prejudice effect sizes” (p. 762). It now appears that some types of prejudice may be more malleable than others, or that some combinations of contact and prejudice mesh especially well.

Third, larger studies – those that estimate the effects of contact with greater precision – tend to reveal weaker effects. Weaker effects are also characteristic of studies that adhere to the highest standards of analytic transparency. Time will tell whether these correlations are statistical flukes or a genuine cause for concern.

Fourth, we know little about what happens within the contact interventions we are assessing. The authors of these research reports rarely describe the contact programs in sufficient detail to allow others to recreate the experience with other populations. In particular, few state explicitly whether their contact intervention meets one or more of Allport's four conditions for reducing prejudice. As a result, we learn little about what specific aspects of the contact are reducing participants' prejudice.

Fifth, and relatedly, no randomized study with over-time outcome measurement has systematically varied, as part of its experimental design, Allport's facilitating conditions. Without manipulating the features of group contact or the conditions under which it occurs, one can only speculate about whether divergent results reflect the treatments, subject pools or conditions of contact (such as equal status or a common goal). For example, one recent study found that prejudice increased when non-Hispanics were exposed to, but did not interact with, randomly assigned confederates speaking Spanish at commuter train stations (Enos, Reference Enos 2014 ).

Allport did not believe that ‘mere contact’ would reduce prejudice. Indeed, Allport warned that without moving beyond casual contact into a deeper engagement characterized by the conditions he set forth, “the more contact the more trouble” ( Reference Allport 1954 , p. 263). This prediction stands in direct contrast to Pettigrew and Tropp's conclusion that “Allport's conditions are not essential for intergroup contact to achieve positive outcomes” ( Reference Pettigrew and Tropp 2006 , p. 766). Given the lack of experiments that systematically test the moderating impact of these conditions on prejudice reduction, we conclude that the literature is not in a place where we can adjudicate between these two positions.

Reinvigorating the study of these moderating conditions means rediscovering innovative experimental designs from decades past. An example of how to experimentally manipulate the conditions of contact comes from Cohen and Roper ( Reference Cohen and Roper 1972 ), who attempted to create an experience of equal status contact between white and black male junior high-school students. In the study, groups of four students, some black and some white, played a strategy game involving cooperation and collective decision-making. The investigators varied study participants' perceptions of outgroup status by preparing them differently in terms of skills and behavioral expectations. While neither this study nor a follow-up replication (Riordan & Ruggiero, Reference Riordan and Ruggiero 1980 ) measured prejudice, nor had long-term outcome measures, they make the point that “equal-status contact should not be assumed,” but rather experimentally manipulated and tested (Riordan & Ruggiero, Reference Riordan and Ruggiero 1980 , p. 131).

Discovering whether Allport's conditions are important for prejudice reduction is not just a matter of theoretical importance – it is an urgent policy question. Scholars reviewing the contact literature often express skepticism about the feasibility of orchestrating the kinds of high-quality contact that Allport ( Reference Allport 1954 ) had prescribed. Dixon et al. ( Reference Dixon, Durrheim and Tredoux 2005 ), for example, lament that contact in “rarefied conditions” may not generalize to “everyday life in divided societies” (p. 697). Amir ( Reference Amir 1969 ), meanwhile, writes that

if most studies have appeared to prove that contact between ethnic groups reduces prejudice, it does not necessarily follow that these results are typical of real social situations. Intergroup contact under the circumstances studied is unfortunately quite rare in actual life, and even when it occurs, it produces only casual interactions rather than intimate acquaintances. (p. 337)

If future research concludes that Allport's conditions are in fact necessary, then policy-makers have a challenging but clear recipe for improving intergroup relations. However, if Allport's conditions are not always necessary, this knowledge could contribute to less expensive interventions that are more readily scalable. Thus, we conclude by renewing Pettigrew and Tropp's call for further investigation of the conditions under which contact reduces prejudice. The contact hypothesis has profound policy implications for the potential benefits of bringing groups together in schools, workplaces and housing. The surge in high-quality research outside the lab and outside the USA brings the policy community closer to answers about the long-term effects of intergroup contact, but important gaps must be addressed before this research can reliably guide future policy decisions.

Acknowledgements

For invaluable research contributions, the authors thank Jason Chin and Kulani Dias. For detailed comments on their meta-analyses, we thank Linda Tropp, Thomas Pettigrew, Gunnar Lemmer and Ulrich Wagner. For helpful comments on an early draft, we thank Ethan Busby, Alex Coppock, Ruth Ditlmann, Jamie Druckman, Al Fang, Nour Kteily, Arnfinn H. Midtbøen, Laura Paler, Alexandra Scacco and Jay Van Bavel. For help assembling and evaluating the contact literature, we thank David Broockman, Lucia Corno, Kristin Davies, Rafaela Dancygier, Greg Duncan, Ryan Enos, Joe Evans, Sarah Gaither, Michael Green, Josh Kalla, David Laitin, Winston Lin, Elizabeth Page-Gould, Gautam Rao, Bruce Sacerdote, Anna Schickele, Nicole Shelton, Natalie Shook, James Sidanius, Samuel Sommers, Todd Stinebrickner, Thomas Trail, Colette van Laar, Wolfgang Viechtbauer, Tessa West and Shahar Zaks. All errors are our own. This research was supported by an NSF Grant #1322356 to Elizabeth Levy Paluck.

Supplementary Material

To view supplementary material for this article, please visit https://doi.org/10.1017/bpp.2018.25 .

Replication code for this article has also been published in Code Ocean, a computational reproducibility platform that enables users to run the code, and can be viewed here: https://doi.org/10.24433/CO.f152260c-bebb-4157-a640-44579452b4e4.v4 .

1 Requiring outcomes to be measured at least one day after the intervention began is a more permissive filter than one that asks for outcomes to be measured at least one day after treatment ends . The latter requirement would eliminate ten studies from our sample. Some of those test the effects of a sustained intervention extending over many months and measured outcomes during or on the final day of the intervention. Specific examples are roommate studies where measurements are collected while students live with same- or other-race roommates or sustained diversity trainings for which outcomes are measured on the last day. Another filter we considered was setting a sample size minimum of 30 people per cell, which would have eliminated nine studies from this sample. We decided to retain the most permissive filter aimed at policy relevance.

2 Our manuscript, online appendices and the archive of digitized reprints are available on the Open Science Framework: https://dx.doi.org/10.17605/OSF.IO/TTPVY

3 Additionally, the authors produce a wealth of supplementary information, making their work both transparent and particularly helpful for this project. We depart from their paper, however, in that we look exclusively at randomized controlled trials, perform sub-analyses by target of prejudice, and, in particular, attend to the relationship between effect sizes and precision of estimates.

4 Readers might think of the divide between Jews and Arabs as an ethnic distinction. We classified this as religious. Classifying this study one way or the other does not affect our overall results.

5 We include only a few studies that feature this type of evidence because it is typically collected during or immediately after intervention.

6 In practice, we find that all available methods of standardizing effect sizes produced substantively similar estimates.

7 When we conduct a sign test that includes the four studies for which we could calculate unstandardized but not standardized effect sizes, we find that 26 out of 31 studies show positive effects, the probability of which is lower than 0.001 under the null hypothesis of no effect.

8 For a detailed look at which studies we did not include and why, see Appendix A, ‘An overview of excluded studies’.

9 Broockman and Kalla ( Reference Broockman and Kalla 2016 ) and Scacco and Warren ( Reference Scacco and Warren 2018 ) have publicly accessible code and data as well.

10 While a pre-analysis plan is not a sufficient condition for a high-quality study, it is a strong indicator of a high-quality replication effort; moreover, these three studies are exemplary in other regards, taking place in field settings and featuring some of the smallest standard errors among the studies we analyzed.

11 See also McClendon ( Reference McClendon 1974 ), Riordan ( Reference Riordan 1978 ), Ford ( Reference Ford 1986 ), Stephan ( Reference Stephan 1987 ) and Patchen ( Reference Patchen 1999 ).

Table 2. Contact studies by decade. Note that we place Meshel and McGlynn (2004) in the 1990s and Dessel (2010) in the 2000s, going by the initial publication of the relevant data in the authors' dissertations (Meshel, 1997; Dessel, 2008)

Paluck et al. supplementary material

Paluck et al. supplementary material 1

Paluck et al. supplementary material 2

Paluck et al. supplementary material 3

Paluck et al. supplementary material 4

This article has been cited by the following publications. This list is generated based on data provided by Crossref .

• Google Scholar

View all Google Scholar citations for this article.

Save article to Kindle

To save this article to your Kindle, first ensure [email protected] is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle .

Note you can select to save to either the @free.kindle.com or @kindle.com variations. ‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

• Volume 3, Issue 2
• ELIZABETH LEVY PALUCK (a1) , SETH A. GREEN (a2) and DONALD P. GREEN (a3)
• DOI: https://doi.org/10.1017/bpp.2018.25

Save article to Dropbox

To save this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Dropbox account. Find out more about saving content to Dropbox .

Save article to Google Drive

To save this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you used this feature, you will be asked to authorise Cambridge Core to connect with your Google Drive account. Find out more about saving content to Google Drive .

Reply to: Submit a response

- No HTML tags allowed - Web page URLs will display as text only - Lines and paragraphs break automatically - Attachments, images or tables are not permitted

Your details

Your email address will be used in order to notify you when your comment has been reviewed by the moderator and in case the author(s) of the article or the moderator need to contact you directly.

You have entered the maximum number of contributors

Conflicting interests.

Please list any fees and grants from, employment by, consultancy for, shared ownership in or any close relationship with, at any time over the preceding 36 months, any organisation whose interests may be affected by the publication of the response. Please also list any non-financial associations or interests (personal, professional, political, institutional, religious or other) that a reasonable reader would want to know about in relation to the submitted work. This pertains to all the authors of the piece, their spouses or partners.

• More from M-W
• To save this word, you'll need to log in. Log In

Definition of hypothesis

Did you know.

The Difference Between Hypothesis and Theory

A hypothesis is an assumption, an idea that is proposed for the sake of argument so that it can be tested to see if it might be true.

In the scientific method, the hypothesis is constructed before any applicable research has been done, apart from a basic background review. You ask a question, read up on what has been studied before, and then form a hypothesis.

A hypothesis is usually tentative; it's an assumption or suggestion made strictly for the objective of being tested.

A theory , in contrast, is a principle that has been formed as an attempt to explain things that have already been substantiated by data. It is used in the names of a number of principles accepted in the scientific community, such as the Big Bang Theory . Because of the rigors of experimentation and control, it is understood to be more likely to be true than a hypothesis is.

In non-scientific use, however, hypothesis and theory are often used interchangeably to mean simply an idea, speculation, or hunch, with theory being the more common choice.

Since this casual use does away with the distinctions upheld by the scientific community, hypothesis and theory are prone to being wrongly interpreted even when they are encountered in scientific contexts—or at least, contexts that allude to scientific study without making the critical distinction that scientists employ when weighing hypotheses and theories.

The most common occurrence is when theory is interpreted—and sometimes even gleefully seized upon—to mean something having less truth value than other scientific principles. (The word law applies to principles so firmly established that they are almost never questioned, such as the law of gravity.)

This mistake is one of projection: since we use theory in general to mean something lightly speculated, then it's implied that scientists must be talking about the same level of uncertainty when they use theory to refer to their well-tested and reasoned principles.

The distinction has come to the forefront particularly on occasions when the content of science curricula in schools has been challenged—notably, when a school board in Georgia put stickers on textbooks stating that evolution was "a theory, not a fact, regarding the origin of living things." As Kenneth R. Miller, a cell biologist at Brown University, has said , a theory "doesn’t mean a hunch or a guess. A theory is a system of explanations that ties together a whole bunch of facts. It not only explains those facts, but predicts what you ought to find from other observations and experiments.”

While theories are never completely infallible, they form the basis of scientific reasoning because, as Miller said "to the best of our ability, we’ve tested them, and they’ve held up."

• proposition
• supposition

hypothesis , theory , law mean a formula derived by inference from scientific data that explains a principle operating in nature.

hypothesis implies insufficient evidence to provide more than a tentative explanation.

theory implies a greater range of evidence and greater likelihood of truth.

law implies a statement of order and relation in nature that has been found to be invariable under the same conditions.

Examples of hypothesis in a Sentence

These examples are programmatically compiled from various online sources to illustrate current usage of the word 'hypothesis.' Any opinions expressed in the examples do not represent those of Merriam-Webster or its editors. Send us feedback about these examples.

Word History

Greek, from hypotithenai to put under, suppose, from hypo- + tithenai to put — more at do

1641, in the meaning defined at sense 1a

Phrases Containing hypothesis

• counter - hypothesis
• nebular hypothesis
• null hypothesis
• planetesimal hypothesis
• Whorfian hypothesis

Articles Related to hypothesis

This is the Difference Between a...

This is the Difference Between a Hypothesis and a Theory

In scientific reasoning, they're two completely different things

Dictionary Entries Near hypothesis

hypothermia

hypothesize

Cite this Entry

“Hypothesis.” Merriam-Webster.com Dictionary , Merriam-Webster, https://www.merriam-webster.com/dictionary/hypothesis. Accessed 10 Jun. 2024.

Kids Definition

Kids definition of hypothesis, medical definition, medical definition of hypothesis, more from merriam-webster on hypothesis.

Nglish: Translation of hypothesis for Spanish Speakers

Britannica English: Translation of hypothesis for Arabic Speakers

Britannica.com: Encyclopedia article about hypothesis

Subscribe to America's largest dictionary and get thousands more definitions and advanced search—ad free!

Can you solve 4 words at once?

Word of the day.

See Definitions and Examples »

Get Word of the Day daily email!

Popular in Grammar & Usage

What's the difference between 'fascism' and 'socialism', more commonly misspelled words, commonly misspelled words, how to use em dashes (—), en dashes (–) , and hyphens (-), absent letters that are heard anyway, popular in wordplay, the words of the week - june 7, 8 words for lesser-known musical instruments, 9 superb owl words, 10 words for lesser-known games and sports, your favorite band is in the dictionary, games & quizzes.

Alternatives to the Big Bang Theory (infographic)

Is there another theory for the origin of the universe?

Big Bang Theory assumptions

Big bang alternatives, additional resources, bibliography.

The Big Bang Theory is the primary explanation for how the universe began 13.8 billion years ago. Though it is the leading explanation, some theorists have come up with alternative ideas or extensions to the Big Bang Theory. We explore some of these theories in the detailed infographic below. 

To investigate how the universe came into existence, we first need to ask what exactly is it? The term "observable universe" refers to everything that we can see, according to NASA . 

Due to the connection between distance and speed of light , scientists can peer at a region of space that light 13.8 billion light-years away meaning we can look 13.8 billion light-years in every direction. But it isn't that simple. Due to the expansion of the universe, recent estimations place the diameter of the observable universe sphere at over 90 billion light-years, according to Forbes . 

Related: How big is the universe?  

Scientists make three assumptions about the universe based on theories and observations:

• The laws of physics are universal and don't change with time or location in space.
• The universe is homogenous, or roughly the same in every direction (though not necessarily all of the time).   
• Humans do not observe the universe from a privileged location such as its very center. 

When applied to Einstein 's equations, they indicate that the universe has several properties: 

• The universe is expanding.
• The universe emerged from a hot, dense state at some infinite time in the past. 
• The lightest elements, hydrogen and helium were created in the first moments. 
• A background of microwave radiation fills the entire universe.  

If any of these basic assumptions are wrong, the Big Bang Theory would not explain all the properties of this universe. This leads to some theorists asking "is it possible the Big Bang never happened?"

Related: Was Einstein wrong? The case against space-time theory

One alternative theory is the Steady State universe. An early rival to the Big Bang theory, Steady State posits the continuous creation of matter throughout the universe to explain its apparent expansion, according to NASA Cosmic Times . This type of universe would be infinite, with no beginning or end. However, a mountain of evidence found since the mid-1960s indicates that this theory is not correct.

 – 10 wild theories about the universe

– What happened before the Big Bang?

– The history of the universe: Big Bang to now in 10 easy steps  

Another alternative is the Eternal Inflation theory. After the Big Bang, the universe expanded rapidly during a brief period called inflation. The Eternal Inflation theory posits that inflation never stopped, and has been going on for an infinite length of time. Somewhere, even now, new universes are coming into existence in a vast complex called the multiverse. Those many universes could have different physical laws.

The Oscillating model of the universe involved an endless series of Big Bangs, followed by Big Crunches that restarted the cycle, endlessly. The modern cyclic model involves colliding "branes" (a "membrane" within a higher-dimensional volume called the "bulk"). 

Implications found in quantum gravity and string theory tantalizingly suggest a universe that is, in reality, nothing like how it appears to human observers. It may be a flat hologram projected onto the surface of a sphere, for example. Or it could be a completely digital simulation running on a vast computer.

Listen to ESA"s interview with astrophysicist Professor Joseph Silk on why we may never get to know whether the universe is finite or infinite. Explore the wealth of evidence for the Big Bang with the National Schools' Observatory , Liverpool John Moores University and The University of Western Australia . 

• Nicolai, Hermann. " Complexity and the Big Bang. " Classical and Quantum Gravity 38.18 (2021): 187001. 
• Netchitailo, Vladimir S. " World-Universe Model—Alternative to Big Bang Model. " Journal of High Energy Physics, Gravitation and Cosmology 6.01 (2020): 133. 
• Wallace, David. The emergent multiverse: Quantum theory according to the Everett interpretation. Oxford University Press, 2012. 
• Linde, Andrei. " A brief history of the multiverse. " Reports on Progress in Physics 80.2 (2017): 022001. 
• Bostrom, Nick. " Are we living in a computer simulation? ." The philosophical quarterly 53.211 (2003): 243-255. 

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: [email protected].

Get the Space.com Newsletter

Breaking space news, the latest updates on rocket launches, skywatching events and more!

Karl's association with Space.com goes back to 2000, when he was hired to produce interactive Flash graphics. From 2010 to 2016, Karl worked as an infographics specialist across all editorial properties of Purch (formerly known as TechMediaNetwork).  Before joining Space.com, Karl spent 11 years at the New York headquarters of The Associated Press, creating news graphics for use around the world in newspapers and on the web.  He has a degree in graphic design from Louisiana State University and now works as a freelance graphic designer in New York City.

• Daisy Dobrijevic Reference Editor

China lands Chang'e 6 sample-return probe on far side of the moon, a lunar success (video)

Science and music festival Starmus VII is about to rock Bratislava with a stellar lineup

At long last: Europe's new Ariane 6 rocket set to debut on July 9

Most Popular

• 2 This Week In Space podcast: Episode 114 — Starliners & Starships
• 3 SpaceX wants to build 1 Starship megarocket a day with new Starfactory
• 4 Arrokoth the 'space snowman' probably tastes like sweet soap
• 5 Apollo 8 astronaut Bill Anders, who captured 'Earthrise,' killed in plane crash

What is a scientific hypothesis?

It's the initial building block in the scientific method.

Hypothesis basics

What makes a hypothesis testable.

• Types of hypotheses
• Hypothesis versus theory

Additional resources

Bibliography.

A scientific hypothesis is a tentative, testable explanation for a phenomenon in the natural world. It's the initial building block in the scientific method . Many describe it as an "educated guess" based on prior knowledge and observation. While this is true, a hypothesis is more informed than a guess. While an "educated guess" suggests a random prediction based on a person's expertise, developing a hypothesis requires active observation and background research. 

The basic idea of a hypothesis is that there is no predetermined outcome. For a solution to be termed a scientific hypothesis, it has to be an idea that can be supported or refuted through carefully crafted experimentation or observation. This concept, called falsifiability and testability, was advanced in the mid-20th century by Austrian-British philosopher Karl Popper in his famous book "The Logic of Scientific Discovery" (Routledge, 1959).

A key function of a hypothesis is to derive predictions about the results of future experiments and then perform those experiments to see whether they support the predictions.

A hypothesis is usually written in the form of an if-then statement, which gives a possibility (if) and explains what may happen because of the possibility (then). The statement could also include "may," according to California State University, Bakersfield .

Here are some examples of hypothesis statements:

• If garlic repels fleas, then a dog that is given garlic every day will not get fleas.
• If sugar causes cavities, then people who eat a lot of candy may be more prone to cavities.
• If ultraviolet light can damage the eyes, then maybe this light can cause blindness.

A useful hypothesis should be testable and falsifiable. That means that it should be possible to prove it wrong. A theory that can't be proved wrong is nonscientific, according to Karl Popper's 1963 book " Conjectures and Refutations ."

An example of an untestable statement is, "Dogs are better than cats." That's because the definition of "better" is vague and subjective. However, an untestable statement can be reworded to make it testable. For example, the previous statement could be changed to this: "Owning a dog is associated with higher levels of physical fitness than owning a cat." With this statement, the researcher can take measures of physical fitness from dog and cat owners and compare the two.

Types of scientific hypotheses

In an experiment, researchers generally state their hypotheses in two ways. The null hypothesis predicts that there will be no relationship between the variables tested, or no difference between the experimental groups. The alternative hypothesis predicts the opposite: that there will be a difference between the experimental groups. This is usually the hypothesis scientists are most interested in, according to the University of Miami .

For example, a null hypothesis might state, "There will be no difference in the rate of muscle growth between people who take a protein supplement and people who don't." The alternative hypothesis would state, "There will be a difference in the rate of muscle growth between people who take a protein supplement and people who don't."

If the results of the experiment show a relationship between the variables, then the null hypothesis has been rejected in favor of the alternative hypothesis, according to the book " Research Methods in Psychology " (​​BCcampus, 2015). 

There are other ways to describe an alternative hypothesis. The alternative hypothesis above does not specify a direction of the effect, only that there will be a difference between the two groups. That type of prediction is called a two-tailed hypothesis. If a hypothesis specifies a certain direction — for example, that people who take a protein supplement will gain more muscle than people who don't — it is called a one-tailed hypothesis, according to William M. K. Trochim , a professor of Policy Analysis and Management at Cornell University.

Sometimes, errors take place during an experiment. These errors can happen in one of two ways. A type I error is when the null hypothesis is rejected when it is true. This is also known as a false positive. A type II error occurs when the null hypothesis is not rejected when it is false. This is also known as a false negative, according to the University of California, Berkeley . 

A hypothesis can be rejected or modified, but it can never be proved correct 100% of the time. For example, a scientist can form a hypothesis stating that if a certain type of tomato has a gene for red pigment, that type of tomato will be red. During research, the scientist then finds that each tomato of this type is red. Though the findings confirm the hypothesis, there may be a tomato of that type somewhere in the world that isn't red. Thus, the hypothesis is true, but it may not be true 100% of the time.

Scientific theory vs. scientific hypothesis

The best hypotheses are simple. They deal with a relatively narrow set of phenomena. But theories are broader; they generally combine multiple hypotheses into a general explanation for a wide range of phenomena, according to the University of California, Berkeley . For example, a hypothesis might state, "If animals adapt to suit their environments, then birds that live on islands with lots of seeds to eat will have differently shaped beaks than birds that live on islands with lots of insects to eat." After testing many hypotheses like these, Charles Darwin formulated an overarching theory: the theory of evolution by natural selection.

"Theories are the ways that we make sense of what we observe in the natural world," Tanner said. "Theories are structures of ideas that explain and interpret facts." 

• Read more about writing a hypothesis, from the American Medical Writers Association.
• Find out why a hypothesis isn't always necessary in science, from The American Biology Teacher.
• Learn about null and alternative hypotheses, from Prof. Essa on YouTube .

Encyclopedia Britannica. Scientific Hypothesis. Jan. 13, 2022. https://www.britannica.com/science/scientific-hypothesis

Karl Popper, "The Logic of Scientific Discovery," Routledge, 1959.

California State University, Bakersfield, "Formatting a testable hypothesis." https://www.csub.edu/~ddodenhoff/Bio100/Bio100sp04/formattingahypothesis.htm  

Karl Popper, "Conjectures and Refutations," Routledge, 1963.

Price, P., Jhangiani, R., & Chiang, I., "Research Methods of Psychology — 2nd Canadian Edition," BCcampus, 2015.‌

University of Miami, "The Scientific Method" http://www.bio.miami.edu/dana/161/evolution/161app1_scimethod.pdf  

William M.K. Trochim, "Research Methods Knowledge Base," https://conjointly.com/kb/hypotheses-explained/  

University of California, Berkeley, "Multiple Hypothesis Testing and False Discovery Rate" https://www.stat.berkeley.edu/~hhuang/STAT141/Lecture-FDR.pdf  

University of California, Berkeley, "Science at multiple levels" https://undsci.berkeley.edu/article/0_0_0/howscienceworks_19

Sign up for the Live Science daily newsletter now

Get the world’s most fascinating discoveries delivered straight to your inbox.

'The difference between alarming and catastrophic': Cascadia megafault has 1 especially deadly section, new map reveals

Arctic 'zombie fires' rising from the dead could unleash vicious cycle of warming

Epidurals may lower risk of complications after birth, study hints

Most Popular

• 2 100-foot 'walking tree' in New Zealand looks like an Ent from Lord of the Rings — and is the lone survivor of a lost forest
• 3 Save $400 on Unistellar's new smart binoculars during their early bird Kickstarter
• 4 10 'breathtaking' photos of our galaxy from the 2024 Milky Way Photographer of the Year contest
• 5 James Webb telescope finds carbon at the dawn of the universe, challenging our understanding of when life could have emerged
• 2 Neanderthals and humans interbred 47,000 years ago for nearly 7,000 years, research suggests
• 3 What is the 3-body problem, and is it really unsolvable?
• 4 Razor-thin silk 'dampens noise by 75%' — could be game-changer for sound-proofing homes and offices
• 5 Shigir Idol: World's oldest wood sculpture has mysterious carved faces and once stood 17 feet tall

Cannon-Bard Theory of Emotion: Definition and Examples

Charlotte Nickerson

Research Assistant at Harvard University

Undergraduate at Harvard University

Charlotte Nickerson is a student at Harvard University obsessed with the intersection of mental health, productivity, and design.

Learn about our Editorial Process

Saul Mcleod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul Mcleod, PhD., is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Olivia Guy-Evans, MSc

Associate Editor for Simply Psychology

BSc (Hons) Psychology, MSc Psychology of Education

Olivia Guy-Evans is a writer and associate editor for Simply Psychology. She has previously worked in healthcare and educational sectors.

On This Page:

Key Takeaways

• The primary argument of the Cannon-Bard theory of emotion is that emotions trigger affective “feelings” and physiological responses to stimuli simultaneously in different regions of the brain. This stands in contrast to the James-Lange theory of emotion, which posits that people and animals feel emotions because they consciously process their physiological responses to stimuli.
• The Cannon-Bard theory of emotion differentiates between feelings associated with the sympathetic nervous system (fight or flight responses) and the parasympathetic nervous system (calm responses), and Cannon believed that sympathetic and parasympathetic responses could not happen simultaneously.
• The Cannon-Bard and James Lange theories of emotion have greatly influenced modern research into emotional processing and the brain; however, both theories have garnered great criticism for their overgeneralization of emotion and contradictions between theory and evidence dating as far back as Cannon’s own research.

According to Cannon-Bard theory of emotion, physiological arousal and emotional experience occur simultaneously, yet independently. This theory was proposed in the 1920s and early 1930s by Walter B. Cannon and Philip Bard.

In short, the Cannon-Bard theory of emotion, also known as the Thalamic theory of emotion, states that the lower part of the brain, what neurologists call the thalamus , controls emotional experience.

Meanwhile, the higher part of the brain, the cortex , controls emotional expression. These feelings (through the thalamus) and physical reactions (through the cortex) occur at the same time. For example, seeing a spider may trigger both a physical response (such as jerking back one’s hand) and an affective, emotional one (such as a feeling of fear).

The Cannon-Bard Theory of Emotion represented a shift from the James-Lange Theory of emotion to one which studied central brain mechanisms as the cause of emotions; however, work as far back as the 1860s posited physiological causes for emotion (Dror, 2014).

How Does the Cannon-Bard Theory Work?

The main idea of Cannon’s approach to emotions is that people react to emotional stimuli but that two separate parts of the brain control the conscious feeling of emotion and the body’s physiological response.

Cannon reviewed research on emotions in both animals and people with brain damage, as well as conducted his own experiments. He concluded that the thalamus was the brain region most involved in physiological emotion, while the cortex is responsible for controlling and inhibiting it.

This is supported by modern neuroscience, which believes that the thalamus is responsible for relaying sensory and motor signals to the cerebral cortex, which handles processes related to thought, consciousness, reasoning, and memory.

Stressful Exam

Say that a student is about to take a high-stakes midterm exam that he has not studied for. This stressful stimulus (an upcoming exam) triggers two brain regions separately: the cortex and the thalamus.

The cortex triggers a physiological response to the emotion. For example, the student may be sweating, his heart may race, and his fingers may tremble as he picks up his pencil.

According to the Cannon-Bard theory of emotion, these are physiological responses to stress triggered by the sympathetic nervous system, and the triggering of the sympathetic nervous system means that the parasympathetic nervous system cannot be triggered.

Meanwhile, the stimulus triggers the student’s thalamus separately, leading to the conscious, affective feeling of stress.

The student is not stressed because he is trembling, according to Cannon-Bard, but because his thalamus has been activated by an impending exam.

This emotional reaction would be separate and independent of the physiological arousal, even though they co-occur.

Purring Cat

Taking an example from Bard’s 1936 letter to Cannon, consider a resting cat who is being petted by its owner. As the owner pets the cat, the cat may purr and relax its muscles or slowly fall asleep.

This purring is a physiological response of the parasympathetic nervous system to the stimulus of being a pet, and, according to the Cannon-Bard theory of emotion, exists entirely separately from the cat perceiving a conscious emotion (such as calmness or comfort).

According to the Cannon-Bard theory of emotion, the cat is not calm because it is purring and has relaxed muscles; rather, it is calm because the stimulus of petting has activated its thalamus, allowing the cat to have the conscious, affective emotion of calm.

Comparison to Other Theories

James-Lange theory of emotion came to influence a century of empirical emotional research and rebuttals of the James-Lange theory of emotion, notably Cannon-Bard’s 1927 critique, have spurned long-standing debates in neuroscience and psychology (Lang, 1994).

James-Lange Theory of Emotion

Both the James-Lange and Cannon-Bard Theories of emotion investigated whether emotions originated from a source central or peripheral to the nervous system (Meiselman, 2016).

The James-Lange theory of emotion is ultimately a peripheral approach to understanding emotion. This means that rather than first perceiving some stimulus that could elicit an emotion, experiencing an emotion, and then experiencing some bodily reaction in response, James-Lange assumes that the bodily reactions themselves elicit conscious emotions.

Cannon offers five objections to the James-Lange theory of emotion:

1.) The latency of physiological responses to emotion is too long to account for the immediacy of emotional behavior;

2.) Artificially inducing bodily reactions to emotion does not in itself produce emotions;

3.) The viscera are so-called “insensitive structures;”

4.) Visceral changes are the same between emotions and

5.) Interrupting the feedback that occurs because of visceral bodily emotions does not influence emotional behavior (Fehr and Stern, 1970).

To illustrate the contrast from which these five objections arise, consider somebody who is afraid of dogs. The dog barks within range of the person, and the person is aroused, according to the James-Lange theory of emotion, autonomically.

This person could have a physical response such as trembling. Because of that autonomic arousal (the trembling), the person has a conscious feeling of fear. The person feels afraid because they tremble.

Meanwhile, according to the Cannon-Bard theory of emotion, the dog’s bark triggers two separate regions of the brain: the thalamus and the cortex.

Because the dog triggers the cortex, which controls physical reactions, the person trembles. However, the person feels fear because the bark triggers the thalamus, which controls the conscious feelings of emotion.

In theory, if that person’s thalamus or cortex were dysfunctional, the person may be able to feel fear without trembling or tremble without feeling fear. The dog makes the person tremble and feel afraid, but these are reactions stemming from two entirely different parts of the brain.

Considering his objections, Canon introduced a specificity model of emotions (Dror, 2014). He distinguished different classes of emotions by whether they affected the sympathetic or parasympathetic nervous system.

Emotions that affect the sympathetic nervous system are associated with fight or flight responses , and those that activate the parasympathetic nervous system restore the body to a state of calm (Waxenbaum, 2021). The anatomical, physiological, and metabolic differences between the emotions expressed by the sympathetic and parasympathetic nervous systems create a difference in how people consciously label these emotions (Cannon, 1914).

In contrast to the Cannon-Bard and James-Lange theories of emotion, the so-called “common sense” theory of emotion posits that someone has an emotional response, and this response triggers a physiological reaction.

For example, someone can see a barking dog and consciously feel fear. As a result, they tremble (Meiselman, 2016).

Singer’s Two-Factor Theory of Emotion

In the Schachter-Singer Two Factor Theory of Emotion , one does a conscious cognitive appraisal of their physiological response, labels that response, and feels the emotion that results. Neither physiological nor cognitive arousal in itself is enough to elicit an emotion, according to Schacter-Singer.

The Cannon-Bard theory maintains that emotional experience occurs simultaneous to and independent of physiological arousal.

The Schachter-Singer two-factor theory suggests that physiological arousal receives cognitive labels as a function of the relevant context and that these two factors together result in an emotional experience.

To repeat the example of a barking dog, a person may tremble or feel an increased heart rate. After detecting that they are trembling, the person can then label the physiological response (“I must be scared!”) and only then consciously feel the emotion.

A person, according to the Two Factor Theory of Emotion, must appraise their physiological responses in order to experience emotion (Meiselman, 2016).

Zajonc-LeDoux Theory of Emotion

The Zajonc-LeDoux theory of emotion says that emotional reactions exist separately from cognitive labels on emotional situations.

According to this theory, some emotions that have evolutionarily been necessary (e.g., anger or fear) are activated through quicker pathways than others.

Certain emotions can happen instantly without an active cognitive appraisal. For example, someone can be startled by the bark of a dog before labeling it as a threat (Meiselman, 2016).

Key Research

In 1925, Cannon and Britton introduced a way of studying emotions through a cat whose cortex had been removed from his brain (Cannon and Britton, 1925).

From this cat, Cannon and Britain described a new type of emotion — “sham rage.”

At the time, Cannon noticed that laboratory animals did not consistently develop the desired emotional reactions to stimuli consistently and dependently, so Cannon removed the region of the brain — the cortex — thought to inhibit emotions, allowing Cannon to study a stream of emotions over longer periods of time.

This decorticated cat came to underpin Cannon’s physiological model of emotional experience.

In the mid-1920s, Philip Bard was a doctoral student studying under Walter Bradford, a physiologist at Harvard University, and two years after introducing this method for studying the physiology of emotions, Cannon suggested that Philip Bard section the cat’s brain in order to discover which parts of the brain were responsible for emotional expression (Bard, 1973).

From this, Cannon-Bard discovered that the thalamus generated affective emotions.

Many neurologists have initiated empirical research in order to determine how and in which order emotion affects various regions of the body.

The first of these researchers was Allport, who devised the facial feedback hypothesis (Laird, 1984). In short, the facial feedback hypothesis postulated that one’s facial expressions directly affect one’s emotional experience. For example, someone who tries to smile will feel happier because they have a smiling expression.

To Tomkins (1962-1963), emotions are “sets of muscles and glandular responses located in the face.”

Later, researchers suggested embodied theories of emotion. According to embodied theories of emotion, simply contracting the facial muscles associated with a particular emotion can intensify or elicit the emotion associated with those contractions, even when participants are not contracting those muscles consciously (Niedenthal, 2007; Soussignan, 2002).

Cannon (1927) conducted several animal tests to disprove James-Lange’s peripheralist approach to emotion (Meiselman, 2016). For example, Cannon conducted tests where he artificially induced certain bodily reactions and observed that these in themselves did not elicit emotions (Meiselman, 2016).

Researchers such as Fehr and Stern (1970), among many others, have criticized these tests. One of the predominant criticisms of the Cannon-Bard Theory of Emotion is that the theory assumes that physical reactions do not influence emotions.

For example, Cannon-Bard would assume that someone would not necessarily feel happy if the facial muscles creating a smile were triggered.

However, a large body of research on facial expressions and emotions suggests that this assumption is not true.

Some scholars, such as Dror (2014), have argued that Cannon’s own studies have contradicted his theory of emotion.

For example, one large component of Cannon’s physiological theory of emotion was the belief that either the sympathetic or parasympathetic nervous system could be activated by emotion, but the activation of these systems was mutually exclusive.

Relatively common phenomena — such as the contraction of the bladder and rectum during intense emotional stress — would confuse Cannon and his followers because these represented the simultaneous activation of the sympathetic and parasympathetic nervous systems (Cannon, 1914).

Scholars, and indeed Cannon himself, were aware of contradictions and flaws in his studies of the decorated cat as a model for physiological and emotional experience (Dror, 2014).

For example, as Dror (2014) emphasizes, Cannon and Britton included clear evidence of parasympathetic activation, such as contraction of the rectum and occasionally defecation, in their descriptions of the cat experiencing “sham rage” – a supposedly sympathetic, fight or flight emotion (Cannon and Britton, 1925).

Many other physiologists in the 1930s would confirm this evidence that “shame rage” activated both the sympathetic and parasympathetic nervous systems (Beattie, 1932; Beebe-Center and Stevens, 1938).

The second major flaw of Cannon’s theory, according to Dror’s 2014 review, was that there was a lack of anatomical proof for two major aspects of the Cannon-Bard theory of emotion.

The stipulation of the Cannon-Bard model that the optic thalamus was the region that organized emotional expressions, as well as the belief that the thalamus itself was the source of the affective experience of emotion, lacked anatomical proof (Cannon, 1927).

Thirdly and lastly, scholars have criticized Cannon’s theory for its overextension beyond the evidence provided by the Cannon-Bard experiments.

Although Cannon-Bard studied the decoriated cat’s “sham rage” exclusively, Cannon would go on to extend his model of emotions to joy, grief, and disgust long before Philip Baard himself would send a letter to Cannon observing pleasure in decorticated cats (such as pleasure when being petted) (Dror, 2014).

Bard, P. (1973). The ontogenesis of one physiologist. Annual Review of Physiology, 35 (1), 1-16.

Beattie, J. (1932). Hypothalamic mechanisms. Canadian Medical Association Journal, 26(4), 400.

Beebe-Center, J., & Stevens, S. (1938). The emotional responses: changes of heart-rate in a gun-shy dog. Journal of Experimental Psychology, 23 (3), 239.

Bernard, C. (1866). Leçons sur les propriétés des tissus vivants: Germer Baillière.

Cannon, W. B. (1914). The interrelations of emotions as suggested by recent physiological researches. The American Journal of Psychology, 25 (2), 256-282.

Cannon, W. B. (1927). The James-Lange theory of emotions: A critical examination and an alternative theory. The American Journal of Psychology, 39 (1/4), 106-124.

Cannon, W. B., & Britton, S. W. (1925). Studies on the conditions of activity in endocrine glands: XV. Pseudaffective medulliadrenal secretion. American Journal of Physiology-Legacy Content, 72 (2), 283-294.

Coppin, G., & Sander, D. (2021). Chapter 1 – Theoretical approaches to emotion and its measurement. In H. L. Meiselman (Ed.), Emotion Measurement (Second Edition) (pp. 3-37): Woodhead Publishing.

Dalgleish, T., Dunn, B. D., & Mobbs, D. (2009). Affective neuroscience: Past, present, and future. Emotion Review, 1 (4), 355-368.

Darwin, C. (1872). The expression of the emotions in man and animals by Charles Darwin: John Murray.

de Cyon, E. (1873). Principes d”électrothérapie: Baillière.

Dror, O. E. (2013). The Cannon–Bard Thalamic Theory of Emotions: A Brief Genealogy and Reappraisal. Emotion Review, 6 (1), 13-20. doi:10.1177/1754073913494898

Durant, J. R. (1981). The beast in man: An historical perspective on the biology of human aggression. The biology of aggression, 17-46.

Ellsworth, P. C. (1994). William James and emotion: is a century of fame worth a century of misunderstanding? Psychological Review, 101 (2), 222.

Fehr, F. S., & Stern, J. A. (1970). Peripheral physiological variables and emotion: the James-Lange theory revisited. Psychological Bulletin, 74 (6), 411.

Friedman, B. H. (2010). Feelings and the body: The Jamesian perspective on autonomic specificity of emotion. Biological Psychology, 84 (3), 383-393.

Laird, J. D. (1984). The real role of facial response in the experience of emotion: a reply to Tourangeau and Ellsworth, and others.

Lanska, D. J. (2014). Cannon, Walter Bradford. In M. J. Aminoff & R. B. Daroff (Eds.), Encyclopedia of the Neurological Sciences (Second Edition) (pp. 580-583). Oxford: Academic Press.

Meiselman, H. L. (2016). Emotion measurement: Woodhead publishing.

Niedenthal, P. M. (2007). Embodying emotion. Science, 316 (5827), 1002-1005.

Sherrington, C. S. (1900). Experiments on the value of vascular and visceral factors for the genesis of emotion. Proceedings of the Royal Society of London, 66 (424-433), 390-403.

Soussignan, R. (2002). Duchenne smile, emotional experience, and autonomic reactivity: a test of the facial feedback hypothesis. Emotion, 2 (1), 52.

Tomkins, S. (1963). Affect imagery consciousness: Volume II: The negative affects: Springer publishing company.

Waxenbaum, J. A., Reddy, V., & Varacallo, M. (2019). Anatomy, autonomic nervous system.

Further Information

• Cannon, W. B. (1927). The James-Lange theory of emotions: A critical examination and an alternative theory. The American journal of psychology, 39(1/4), 106-124.
• Cannon, W. B. (1914). The interrelations of emotions as suggested by recent physiological researches. The American Journal of Psychology, 25(2), 256-282.
• Barrett, L. F. (2012). Emotions are real. Emotion, 12(3), 413.
• Ellsworth, P. C. (1994). William James and emotion: is a century of fame worth a century of misunderstanding?. Psychological review, 101(2), 222.

Related Articles

Clinical Psychology

What Makes Work Sustainable For Autistic And ADHD Individuals?

Unprompted Thought and Affective Well-Being: A Systematic Review and Meta-Analysis

Executive Function Deficits Mediate The Relationship Between Employee’s ADHD And Job Burnout

Identification Of Exceptional Skills In Autistic Children

Limiting Social Media Use Decreases Depression, Anxiety, & FOMO In Youth

Social Attention, Physiological Arousal, And Familiarity In Autistic Children