the similarity thesis is based on which similarities between partners

• school Campus Bookshelves
• menu_book Bookshelves
• perm_media Learning Objects
• login Login
• how_to_reg Request Instructor Account
• hub Instructor Commons
• Download Page (PDF)
• Download Full Book (PDF)
• Periodic Table
• Physics Constants
• Scientific Calculator
• Reference & Cite
• Tools expand_more
• Readability

selected template will load here

This action is not available.

8.2: Relationship Formation

• Last updated
• Save as PDF
• Page ID 66589

• Jason S. Wrench, Narissra M. Punyanunt-Carter & Katherine S. Thweatt
• SUNY New Paltz & SUNY Oswego via OpenSUNY

Learning Outcomes

• Understand attraction.
• Ascertain reasons for attraction.
• Realize the different types of attraction.

Have you ever wondered why people pick certain relationships over others? We can’t pick our family members, although I know some people wish they could. We can, however, select who our friends and significant others are in our lives. Throughout our lives, we pick and select people that we build a connection to and have an attraction towards. We tend to avoid certain people who we don’t find attractive.

Understanding Attraction

Researchers have identified three primary types of attraction: physical, social, and task. Physical attraction refers to the degree to which you find another person aesthetically pleasing. What is deemed aesthetically pleasing can alter greatly from one culture to the next. We also know that pop culture can greatly define what is considered to be physically appealing from one era to the next. Think of the curvaceous ideal of Marilyn Monroe and Elizabeth Taylor in the 1950s as compared to the thin Halle Barry or Anne Hathaway. Although discussions of male physical attraction occur less often, they are equally impacted by pop culture. In the 1950s, you had solid men like Robert Mitchum and Marlon Brando as compared to the heavily muscled men of today like Joe Manganiello or Zac Efron.

The second type of attraction is social attraction , or the degree to which an individual sees another person as entertaining, intriguing, and fun to be around. We all have finite sources when it comes to the amount of time we have in a given day. We prefer to socialize with people that we think are fun. These people may entertain us or they may just fascinate us. No matter the reason, we find some people more socially desirable than others. Social attraction can also be a factor of power, for example, in situations where there are kids in the “in-group” and those that are not. In this case, those that are considered popular hold more power and are perceived as being more socially desirable to associate with. This relationship becomes problematic when these individuals decide to use this social desirability as a tool or weapon against others.

The final type of attraction is task attraction , or people we are attracted to because they possess specific knowledge and/or skills that help us accomplish specific goals. The first part of this definition requires that the target of task attraction possess specific knowledge and/or skills. Maybe you have a friend who is good with computers who will always fix your computer when something goes wrong. Maybe you have a friend who is good in math and can tutor you. Of course, the purpose of these relationships is to help you accomplish your own goals. In the first case, you have the goal of not having a broken down computer. In the second case, you have the goal of passing math. This is not to say that an individual may only be viewed as task attractive, but many relationships we form are because of task attraction in our lives.

Reasons for Attraction

Now that we’ve looked at the basics of what attraction is, let’s switch gears and talk about why we are attracted to each other. There are several reasons researchers have found for our attraction to others, including proximity, physicality, perceived gain, similarities and differences, and disclosure.

Physical Proximity

When you ask some people how they met their significant other, you will often hear proximity is a factor in how they met. Perhaps, they were taking the same class or their families went to the same grocery store. These common places create opportunities for others to meet and mingle. We are more likely to talk to people that we see frequently.

Physical Attractiveness

In day-to-day interactions, you are more likely to pay attention to someone you find more attractive than others. Research shows that males place more emphasis on physical attractiveness than females. 5 Appearance is very important at the beginning of the relationship.

Perceived Gain

This type of relationship might appear to be like an economic model and can be explained by exchange theory . 6 In other words, we will form relationships with people who can offer us rewards that outweigh the costs. Rewards are the things we want to acquire. They could be tangible (e.g., food, money, clothes) or intangible (support, admiration, status). Costs are undesirable things that we don’t want to expend a lot of energy to do. For instance, we don’t want to have to constantly nag the other person to call us or spend a lot of time arguing about past items. A good relationship will have fewer costs and more rewards. A bad relationship will have more costs and fewer rewards. Often, when people decide to stay or leave a relationship, they will consider the costs and rewards in the relationship.

Costs and rewards are not the only factors in a relationship. Partners also consider alternatives in the relationship. For instance, Becky and Alan have been together for a few years. Becky adores Alan and wants to marry him, but she feels that there are some problems in the relationship. Alan has a horrible temper; he is pessimistic; and he is critical of her. Becky has gained some weight, and Alan has said some hurtful things to her. Becky knows that every relationship will have issues. She doesn’t know whether to continue this relationship and take it further or if she should end it.

Her first alternative is called the comparison level (CL), which is the minimum standard that she is willing to tolerate. If Becky believes that it is ok for a person to say hurtful things to her or get angry, then Alan is meeting or exceeding her CL. However, if past romantic partners have never said anything hurtful towards her, then she would have a lower CL.

Becky will also consider another alternative, which is the comparison level of alternatives (CL alt ), or the comparison between current relationship rewards and what she might get in another relationship. If she doesn’t want to be single, then she might have a lower CL of alternatives. If she has another potential mate who would probably treat her better, then she would have a higher level of alternatives. We use this calculation all the time in relationships. Often when people are considering the possibility to end a relationship, they will consider all alternatives rather than just focusing on costs and rewards.

Similarities and Differences

It feels comforting when someone who appears to like the same things you like also has other similarities to you. Thus, you don’t have to explain yourself or give reasons for doing things a certain way. People with similar cultural, ethnic, or religious backgrounds are typically drawn to each other for this reason. It is also known as similarity thesis . The similarity thesis basically states that we are attracted to and tend to form relationships with others who are similar to us. 7 There are three reasons why similarity thesis works: validation, predictability, and affiliation. First, it is validating to know that someone likes the same things that we do. It confirms and endorses what we believe. In turn, it increases support and affection. Second, when we are similar to another person, we can make predictions about what they will like and not like. We can make better estimations and expectations about what the person will do and how they will behave. The third reason is due to the fact that we like others that are similar to us and thus they should like us because we are the same. Hence, it creates affiliation or connection with that other person.

However, there are some people who are attracted to someone completely opposite from who they are. This is where differences come into play. Differences can make a relationship stronger, especially when you have a relationship that is complementary . In complementary relationships, each person in the relationship can help satisfy the other person’s needs. For instance, one person likes to talk, and the other person likes to listen. They get along great because they can be comfortable in their communication behaviors and roles. In addition, they don’t have to argue over who will need to talk. Another example might be that one person likes to cook, and the other person likes to eat. This is a great relationship because both people are getting what they like, and it complements each other’s talents. Usually, friction will occur when there are differences of opinion or control issues. For example, if you have someone who loves to spend money and the other person who loves to save money, it might be very hard to decide how to handle financial issues.

Sometimes we form relationships with others after we have disclosed something about ourselves to others. Disclosure increases liking because it creates support and trust between you and this other person. We typically don’t disclose our most intimate thoughts to a stranger. We do this behavior with people we are close to because it creates a bond with the other person.

Disclosure is not the only factor that can lead to forming relationships. Disclosure needs to be appropriate and reciprocal 8 . In other words, if you provide information, it must be mutual. If you reveal too much or too little, it might be regarded as inappropriate and can create tension. Also, if you disclose information too soon or too quickly in the relationship, it can create some negative outcomes.

Key Takeaways

• We can be attracted to another person via various ways. It might be due to physical proximity, physical appearance, perceived gain, similarity/differences, and disclosure.
• The deepening of relationships can occur through disclosure and mutual trust.
• Relationships end through some form of separation or dissolution.
• Take a poll of the couples that you know and how they met. Which category does it fall into? Is there a difference among your couples and how they met?
• What are some ways that you could form a relationship with others? Discuss your findings with the class. How is it different/similar to what we talked about in this chapter?
• Discuss how and why a certain relationship that you know dissolved. What were the reasons or factors that caused the separation?

• Media Center

Similarity Hypothesis

The basic idea, theory, meet practice.

TDL is an applied research consultancy. In our work, we leverage the insights of diverse fields—from psychology and economics to machine learning and behavioral data science—to sculpt targeted solutions to nuanced problems.

Consider the closest friends you meet while backpacking abroad. You likely share many similarities; perhaps a thrill for spontaneity, hobbies, appreciation for culture, music preferences, or food choices. During the trip, you find yourself effortlessly interacting with other backpackers: sharing a relatively-unknown scenic route, a local exhibition to visit, or the best bed and breakfast in town. We often relate and empathize easily with similar individuals – this is a result of the similarity hypothesis.

The  similarity hypothesis  suggests that we tend to be drawn towards those who are similar to ourselves. Similarities can refer to shared attitudes and values, as well as political opinions, cultural background, or even minute details like posture. 1

The experience of interacting with similar individuals jumpstarts cognitive processing, like learning, memory, attention, and reasoning. An aspiring musician might remember all the lyrics to their favorite band’s albums. An employee might pick up skills more quickly when assisted by a mentor they admire or identify with. Even when it comes to making comparisons with others, we tend to look for individuals who share similar attitudes and beliefs because it can be difficult to make accurate comparisons when others are too different from us. 2

Why do we tend to be drawn towards individuals who share similar attitudes and values?

Similarity Hypothesis:  A hypothesis which states that we tend to be attracted towards individuals who share similar important traits, such as attitudes and values.

Cognitive Processing:  A general term to describe any mental function involved in acquiring, storing, interpreting and manipulating information. These functions can be conscious or unconscious, such as attention, memory storage, learning, and reasoning.

Empathy:  Understanding an individual from their point of view and experiencing that individual’s feelings, thoughts and perceptions.

In 1954, Leon Festinger proposed in his social comparison theory: when individuals are uncertain of their abilities and opinions, they tend to make comparisons with other similar individuals to assess the accuracy of their own opinion. Festinger’s influential social comparison theory introduced the similarity hypothesis. Since its introduction in  A Theory of Social Comparison Processes , a large amount of evidence has supported the hypothesis. 3

Festinger’s hypothesis has been used to explain phenomena in a diverse array of fields, from political science to marketing. For instance, in the 1971  The Attraction Paradigm , psychologist Donn Byrne introduced the similarity-attraction theory. Byrne’s theory was based on the similarity hypothesis. He suggested that individuals who share similar “important attitudes” (opinions on family and values) are generally more likely to be attracted to each other, compared to individuals who share similar “less important” attitudes (opinions on a specific type of sink). 4  This holds for friendships as well as romantic partners. Byrne further outlined that individuals associate with those who have similar personality characteristics, such as self-esteem, optimism, and conscientiousness.

According to Byrne, personality similarity has a key role to play in the longevity and happiness of a marriage. 5  Byrne’s similarity-attraction theory stated that individuals are generally romantically attracted to others who share similar physical characteristics and levels of physical attractiveness. Byrne’s work on similarity-attraction was so influential that further research has supported his theory, with individuals’ preference for similarity being demonstrated in various other aspects such as social habits and socioeconomic status. 5

The similarity hypothesis then made its way into the field of economics and decision-making in Amos Tversky’s 1972 book,  Elimination by Aspects: A Theory of Choice . 6  Tversky influenced choice theory in economics by applying the similarity hypothesis to decision-making, changing the way modern economists approached the field. Based on the hypothesis, he suggested that when a new product enters a market, it will take more demand from the share of a similar product than a dissimilar one. This has important implications for brands: when creating a new line of products, they should make it as dissimilar as possible from their current offering to prevent market cannibalization. Tversky’s work influenced marketing managers, who started adopting his use of the similarity hypothesis to help make marketing entry decisions. 7

Leon Festinger

An influential American social psychologist, most renowned for his work on social comparison theory in his 1954 book,  A Theory of Social Comparison Processes . Festinger introduced the similarity hypothesis in this book, which has been followed by an enormous amount of data which has provided evidence to support the hypothesis. Several of Festinger’s theories and research also renounced previously dominant behaviorist views of social psychology.

An American psychologist and  influential contributor of foundational theory in interpersonal attraction. His work on similarity-attraction theory, based on the similarity hypothesis, was groundbreaking for exploring the relationship between similar attitudes and attraction. Byrne was also an early contributor on the psychology of human sexuality. 8

Amos Tversky

One of the founders of behavioral science who helped revolutionize the field of economics and decision-making. Tversky was an influential psychologist who applied the similarity hypothesis to decision-making and choice theory in economics. Along with  Daniel Kahneman , Tversky was also a pioneer in  loss aversion  and  prospect theory .

Consequences

When it comes to attraction, Byrne’s similarity-attraction theory remains relevant today as it provides reassurance that an individual is not alone in their belief. Being attracted to individuals with similar attitudes also enables one to more accurately predict the other’s behaviors in different scenarios, providing an insight into the other’s predilections and “pet peeves” based on similarity. 5

Similarly, when we empathize with a target, such as a novel, our enhanced cognitive processing enables us to facilitate reading comprehension. Our reading accelerates and our memory increases. Likewise, when we fail to empathize with a target, such as a film, we evoke a perception of dissimilarity. This creates the opposite effect, and our cognitive processing is inhibited: we lose focus easily, finding it difficult to recall the plot of the film. 1

Our enhanced cognitive processing is a result of empathy, which arises from our perception of similarity. This affects the way we interact with other individuals, as the perception of similarity can implicitly evoke empathy between two individuals. The perception of similarity is the reason why an employee may be able to learn new techniques more quickly when assisted by a mentor they empathize with.

Understanding the similarity hypothesis can allow us to better design inclusive educational curricula, particularly in scenarios where it is important to understand individuals or experiences which are not necessarily similar to most learners. This can be especially useful in cross-cultural education, history, minority education, and special-needs classes. 1  Applying the similarity hypothesis in these fields of education can help overcome the effort involved in understanding experiences or individuals which are dissimilar.

Controversies

Despite the repeated evidence upholding the similarity hypothesis, one criticism is that individuals frequently seek novelty and difference, with such experiences providing just as much certainty when it comes to self-evaluation. 3

Scholars who disagree with the similarity-attraction theory tend to adopt the complementarity view of attraction. This view states that individuals are more likely to prefer partners who have attributes that are complementary, rather than those who possess replicating attributes. This can be seen when an individual with a certain perceived negative attribute, such as impatience, is more attracted to someone who does not possess that same attribute. The complementarity view of attraction suggests that individuals prefer not to be reminded of their faults by being with someone similar, and therefore they are more attracted to those who will complement and bring out the best in them. 5

Emerging studies are also starting to define more clearly that it is perceived similarity, rather than actual similarity, that influences attraction. A 2012 study by American psychologists at Texas A&M and Northwestern University found that, unlike previous findings, actual similarity did not predict romantic attraction as effectively as previously thought. 9

There are alternative views when addressing how the similarity hypothesis influences opinion comparisons between individuals. Some argue that comparisons with other similar individuals depend on the type of opinion being evaluated. A study in 2000 by Jerry Suls, René Martin, and Ladd Wheeler highlights results which suggest that we prefer comparing with other similar individuals when it comes to the evaluation of preferences. Think about how you are more likely to care about what your best friend thinks of your outfit, compared to the Lyft driver who dropped you off this morning. In contrast, other studies have suggested that we prefer to compare ourselves with dissimilar individuals when it comes to belief assessment, 3  such as evaluating whether a certain statement or proposition is true.

The effects of the similarity hypothesis on memory retrieval.

In 2015, Hidetsugu Komeda conducted a study to observe memory retrieval in typically developing (TD) individuals and individuals with Autism Spectrum Disorder (ASD). The similarity hypothesis predicts that individuals with ASD will be able to easily retrieve other individuals with ASD from their memory. Participants were carefully selected and read 24 stories, before completing a recognition task. The results showed that ASD individuals demonstrated the same level of accuracy as TD individuals, but memory-retrieval patterns between the two groups were different. 1

Individuals with ASD were able to retrieve ASD-consistent stories more easily than ASD-inconsistent stories. TD individuals were also able to retrieve TD-consistent stories more easily than ASD-protagonist stories. These results are consistent with the similarity hypothesis, suggesting that individuals with ASD characteristics are able to help other ASD individuals due to empathy arising from their similarities. 1

Related TDL Content

The Similar-To-Me Effect

Why do we tend to surround ourselves with people similar to ourselves? While it is normal to get along with people who have similar experiences, like your basketball teammate or a fellow college alumnus, favoring people similar to you becomes a problem when it leads to discrimination.

Why do we feel more strongly about one option after a third one is added?

You might never buy the most expensive option, but do you sometimes buy the second-most expensive option? The decoy effect explains why the addition of a third choice can make us spend more money – even if we don’t opt for the new choice.

• Definition of Rapport . (n.d.). Dictionary by Merriam-Webster. Retrieved October 4, 2021, from  https://www.merriam-webster.com/dictionary/rapport
• Rapport Quotes . (n.d.). A-Z Quotes. Retrieved October 4, 2021, from  https://www.azquotes.com/quotes/topics/rapport.html
• Mcleod, S. (2020).  Humanistic approach . Simply Psychology.  https://www.simplypsychology.org/humanistic.html
• Active Listening . (n.d.). Skills You Need. Retrieved October 4, 2021, from  https://www.skillsyouneed.com/ips/active-listening.html
• Vollmer, S. (2010, January 6).  Transference . Psychology Today.  https://www.psychologytoday.com/ca/blog/learning-play/201001/transference
• American Psychological Association. (n.d.).  therapeutic alliance . APA Dictionary of Psychology. Retrieved October 4, 2021, from  https://dictionary.apa.org/therapeutic-alliance
• The Mind Tools Content Team. (2019).  Building Rapport . Mind Tools.  https://www.mindtools.com/pages/article/building-rapport.htm
• What is Rapport? Techniques for Relationship Building . (2018, May 17). Exploring Your Mind.  https://exploringyourmind.com/what-is-rapport-techniques-for-relationship-building/
• Mcleod, S. (2014).  Carl Rogers Theory . Simply Psychology.  https://www.simplypsychology.org/carl-rogers.html
• Coan, G. (1984). Rapport: Definitions and Dimensions.  Advances in Consumer Research ,  11 , 333-336.  https://www.acrwebsite.org/volumes/6269/volumes/v11/NA-11
• Tickle-Degnen, L., & Rosenthal, R. (1990). The nature of rapport and its nonverbal correlates.  Psychological Inquiry ,  1 (4), 285-293.  https://doi.org/10.1207/s15327965pli0104_1
• Buskist, W., & Saville, B. K. (2001). Creating Positive Emotional Contexts for Enhancing Teaching and Learning.  APS Observer , 12-13.  https://www.socialpsychology.org/rapport.htm
• Ardito, R. B., & Rabellino, D. (2011). Therapeutic Alliance and Outcome of Psychotherapy: Historical Excursus, Measurements, and Prospects for Research.  Frontiers in Psychology ,  2 (270).  https://doi.org/10.3389/fpsyg.2011.00270
• Miles, L. K., Nind, L. K., & Macrae, C. N. (2009). The rhythm of rapport: Interpersonal synchrony and social perception.  Journal of Experimental Social Psychology ,  45 (3), 585-589.  https://doi.org/10.1016/j.jesp.2009.02.002
• Drolet, A. L., & Morris, M. W. (2000). Rapport in conflict resolution: Accounting for how face-to-face contact fosters mutual cooperation in mixed-motive conflicts.  Journal of Experimental Social Psychology ,  36 (1), 26-50.  https://doi.org/10.1006/jesp.1999.1395

Mental Accounting

Tree Testing

Service Blueprints

Cognitive Walkthrough

Eager to learn about how behavioral science can help your organization?

Get new behavioral science insights in your inbox every month..

• Relationships

Does Similarity Matter in Relationships?

Do birds of a feather flock together or do opposites attract.

Posted September 25, 2020 | Reviewed by Devon Frye

• Why Relationships Matter
• Find counselling to strengthen relationships

If I asked you to tell me what makes two people compatible, what would your answer be? If you are like the students in my course on the psychology of close relationships, your answer is likely to be “similarity.” People tend to think that having a partner who is similar is important for relationship success. But, of course, some people also argue that “opposites attract.” So which is it?

Imagine taking 200 couples and then mixing up the couples so that you randomly assign partners from different couples to each other. According to research, if you look at the similarity between members of couples and compare it to two people randomly put together, partners in actual relationships tend to be more similar to each other that the randomly paired partners.

But knowing that partners tend to be similar doesn’t necessarily mean that similarity matters. Just because partners tend to be similar does not mean that more similar couples are in more satisfying relationships or that their relationships last longer. Perhaps dissimilar partners are less common but just as satisfied and successful.

What we know about similarity and relationship success is that similarity can matter, but it looks like it's about values and background more than personality . Researchers have shown that similarity in couples tends to be more about shared values and background, such as their social class and religion. And these are the factors that appear to predict relationship success—couples with more similar attitudes, values, and backgrounds tend to experience more lasting satisfaction, companionship, intimacy , and love and are less likely to break up.

Similarity of personality, on the other hand, does not appear to matter as much. While some research found that people report being most attracted to others with similar personalities, similar personalities did not strongly predict relationship outcomes.

How similar partners are might be a function of how they met. One cross-sectional study of 137 married or cohabiting heterosexual couples found that couples who “fell in love at first sight” were less similar than couples who were “friends first,” particularly on levels of extraversion , emotional stability , and autonomy. However, contrary to the researchers’ hypothesis, partners who “fell in love at first sight” did not report lower relationship quality, suggesting that their dissimilar personalities were not necessarily a burden on the relationship.

While similarity in personality might not matter as much as we instinctively think, certain personality traits do seem to matter. In particular, neuroticism (the tendency to experience negative emotions) has been shown to predict lower relationship quality. Though this may not be true across the lifespan: Most of this research is done with younger couples, and in a sample of older, long-term marriages there was no link between neuroticism and relationship satisfaction. Some research has found that being open to experiences, agreeable, and conscientious all bode well for relationship quality, but the findings generally are not as strong as those with neuroticism. When asking people what traits they value most in partners, the answer is loyalty and honesty.

So which factors actually do matter for who we end up with and whether our relationships are successful? Stay tuned for my next post.

Facebook image: WAYHOME studio/Shutterstock

Kenny, D. A., & Acitelli, L. K. (1994). Measuring similarity in couples. Journal of family psychology, 8(4), 417.

Markey, P. M., & Markey, C. N. (2007). Romantic ideals, romantic obtainment, and relationship experiences: The complementarity of interpersonal traits among romantic partners. Journal of social and Personal Relationships, 24(4), 517-533.

O’Rourke, N., Claxton, A., Chou, P. H. B., Smith, J. Z., & Hadjistavropoulos, T. (2011). Personality trait levels within older couples and between-spouse trait differences as predictors of marital satisfaction. Aging & mental health, 15(3), 344-353.

Weidmann, R., Ledermann, T., & Grob, A. (2017). The interdependence of personality and satisfaction in couples. European Psychologist.

Amie M. Gordon, Ph.D., is a social psychologist at the University of Michigan whose research focuses on interpersonal relationships and well-being.

• Find a Therapist
• Find a Treatment Center
• Find a Psychiatrist
• Find a Support Group
• Find Online Therapy
• International
• New Zealand
• South Africa
• Switzerland
• Asperger's
• Bipolar Disorder
• Chronic Pain
• Eating Disorders
• Passive Aggression
• Personality
• Goal Setting
• Positive Psychology
• Stopping Smoking
• Low Sexual Desire
• Child Development
• Therapy Center NEW
• Diagnosis Dictionary
• Types of Therapy

Understanding what emotional intelligence looks like and the steps needed to improve it could light a path to a more emotionally adept world.

• Coronavirus Disease 2019
• Affective Forecasting
• Neuroscience

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

A Little Similarity Goes a Long Way: The Effects of Peripheral but Self-Revealing Similarities on Improving and Sustaining Interracial Relationships

Tessa v. west.

Department of Psychology, New York University

Joe C. Magee

Stern School of Business, New York University

Sarah H. Gordon

School of Public Health, Harvard University

Lindy Gullett

Integrating theory on close relationships and intergroup relations, we construct a manipulation of similarity that we demonstrate can improve interracial interactions across different settings. We find that manipulating perceptions of similarity on self-revealing attributes that are peripheral to the interaction improves interactions in cross-race dyads and racially diverse task groups. In a getting-acquainted context, we demonstrate that the belief that one’s different-race partner is similar to oneself on self-revealing, peripheral attributes leads to less anticipatory anxiety than the belief that one’s partner is similar on peripheral, nonself-revealing attributes. In another dyadic context, we explore the range of benefits that perceptions of peripheral, self-revealing similarity can bring to different-race interaction partners and find (a) less anxiety during interaction, (b) greater interest in sustained contact with one’s partner, and (c) stronger accuracy in perceptions of one’s partners’ relationship intentions. By contrast, participants in same-race interactions were largely unaffected by these manipulations of perceived similarity. Our final experiment shows that among small task groups composed of racially diverse individuals, those whose members perceive peripheral, self-revealing similarity perform superior to those who perceive dissimilarity. Implications for using this approach to improve interracial interactions across different goal-driven contexts are discussed.

Despite the increasing frequency of cross-race contact, for most people, interracial interactions are still experienced more negatively than are intraracial interactions ( Brown & Hewstone, 2005 ; Pettigrew & Tropp, 2008 ; Plant & Butz, 2006 ). For example, for both majority and minority group members, interracial interactions are marked by higher levels of stress and anxiety than intraracial interactions ( Blascovich, Mendes, Hunter, Lickel, & Kowai-Bell, 2001 ; Dovidio, 2001 ; Dovidio, Gaertner, Kawakami, & Hodson, 2002 ; Pearson et al., 2008 ; Trawalter, Richeson, & Shelton, 2009 )—an effect that has remained fairly stable over the past four decades ( Toosi, Babbitt, Ambady, & Sommers, 2012 ). The physiological and psychological discomfort of interracial interaction begins when anticipating the interaction ( Mendoza-Denton, Page-Gould, & Pietrzak, 2006 ), thus reducing the likelihood that people will initiate ( Shelton & Richeson, 2005 ) or maintain interest in prolonged contact ( Pearson et al., 2008 ; Plant, 2004 ; Plant & Butz, 2006 ).

Although researchers have invested significantly in documenting these issues in interracial interactions, comparatively little investment has been made in figuring out how to reduce them, particularly within interpersonal interactions. This is not because the need to address these issues has gone unrecognized. Researchers have made repeated calls for methods that improve the quality of interracial relations across multiple interaction settings (e.g., Paluck & Green, 2009 ; Pettigrew, 1998 ; Pettigrew & Tropp, 2008 ; van Knippenberg & Schippers, 2006; Williams & O’Reilly, 1998 ). Among past approaches that have successfully improved interracial interactions, most were designed for specific interaction contexts (e.g., friendship forming; Mallett & Wilson, 2010 ; Page-Gould, Mendoza-Denton, & Tropp, 2008 ; Pinel & Long, 2012 ), require preexisting common ground between partners (e.g., Nier et al., 2001 ), or necessitate repeated interactions before any improvement occurs (e.g., Page-Gould et al., 2008 ).

We aimed to address these limitations by developing an intervention that promotes positive interpersonal processes during the getting-acquainted stage of cross-race interactions and operates effectively across diverse interracial contexts (e.g., friendship building and task performance). We propose that research on the determinants of close relationship satisfaction can inform how to reduce the unsatisfying experiences common to cross-race interactions, and we adapt theory and methods from studies of friends and other intimate relationships to bring cross-race interaction partners closer. Although cross-race interactions among unfamiliar individuals are qualitatively different from—even seemingly opposite of—repeated interactions between close relationship partners, we argue that one of the core findings in research on the formation and maintenance of close relationships can be effectively integrated with theories of intergroup relations to address problems in interracial interactions. Specifically, we propose that the benefits of perceiving similarity to one’s partner on attributes that are perceived as self-revealing (i.e., attributes that communicate something important about the self) (e.g., Lutz-Zois, Bradley, Mihalik, & Moorman-Eavers, 2006 ) can be extended from close relationships to interracial relationships.

How Can Research on Close Relationships Be Applied to Cross-Race Interactions?

Perceiving similarity with one’s relationship partner facilitates close relationships in a number of ways ( Lemay & Clark, 2008 ; Murray, Holmes, Bellavia, Griffin, & Dolderman, 2002 ). Individuals who believe that they are similar to their romantic partners on self-revealing dimensions, such as values, attitudes, emotional experiences, and personality traits, not only are more committed and satisfied in their relationships than those who perceive less similarity ( Kenny & Acitelli, 2001 ; LaPrelle, Holyle, Insko, & Bernthal, 1990 ) but also have more satisfied and committed partners ( Murray et al., 2002 ). Highlighting that the extent to which the dimensions are self-revealing is critical to the effects of perceived similarity, Lutz-Zois, Bradley, Mihalik, and Moorman-Eavers (2006) found that perceived similarity was positively associated with relationship satisfaction only when the dimensions of similarity were personally valued by the relationship partners.

The benefits of perceived similarity can also be found in less established relationships, even among new acquaintances. For example, Sunnafrank and Ramirez (2004) found that among a sample of newly acquainted undergraduates, perceived similarity in attitudes predicted long-term attraction and frequency of communication between partners 9 weeks later. The authors argued that partners’ estimates of perceived similarity during the initial acquaintance stage predicted their desire to engage in future interaction and the development of a long-term relationship. Similarly, Selfhout, Denissen, Branje, and Meeus (2009) demonstrated that greater perceived similarity in personality profiles at initial stages of interaction predicts stronger relationships over time. These positive relational outcomes appear to have been generated by enhanced communication (i.e., the ease of communication and its frequency) between partners who perceived themselves as more similar.

Perceived similarity is beneficial even when it is not accompanied by actual similarity ( Hoyle, 1993 ; Morry, Kito, Martens, Marchylo, & Stevens, 2005 ; Tidwell, Eastwick, & Finkel, 2012 ). When people believe they have found a partner who is similar in terms of traits, values, and emotional experiences, both partners report higher levels of relationship satisfaction, even if their perceived similarity is somewhat illusory ( Murray et al., 2002 ; see also Selfhout et al., 2009 ; Tidwell et al., 2012 ). These findings converge with a meta-analysis demonstrating that in field studies, the effect of actual similarity on liking is small relative to the effect of perceived similarity (for a review, see Montoya, Horton, & Kirchner, 2008 ).

The tendency to perceive similarity between oneself and one’s close partners is motivated by a desire to feel positively about one’s close relationships in the face of an imperfect reality ( Lemay & Clark, 2008 ; Murray et al., 2002 ). After all, partners find important bases of dissimilarity ( Norton, Frost, & Ariely, 2007 ) and discover patterns of behavior they dislike in each other ( Neff & Karney, 2005 ). These experiences can threaten close relationships, and perceived similarity provides a buffer for relationships by enhancing mutual understanding, communication, and conflict resolution ( Holmes & Rempel, 1989 ; Linden-Andersen, Markiewicz, & Doyle, 2009 ; Murray et al., 2002 ). In close relationships (both established and developing), it makes a great deal of sense to perceive self–other similarity for relationship-protective reasons; however, in getting-acquainted contexts and temporary groups, in which individuals have much less psychological investment, it is less likely that partners and group members will perceive similarity, particularly if they have a visible marker of difference, such as race.

Perceiving Similarity in Cross-Race Versus Same-Race Interactions

Going into an encounter, racial differences serve as a strong basis of assumed dissimilarity ( Byrne & Wong, 1962 ; Frey & Tropp, 2006 ; Robbins & Krueger, 2005 ; Rokeach & Mezei, 1966 ; Stein, 1966 ; Stein, Hardyck, & Smith, 1965 ; Vorauer, Main, & O’Connell, 1998 ). As such, race can be a powerful antecedent to negative experiences in interracial interactions, giving rise to elevated levels of anxiety and uncertainty in anticipation of interaction ( Fiske, Lin, & Neuberg, 1999 ; Gaertner & Dovidio, 1986 ; Mallett, Wilson, & Gilbert, 2008 ; Plant, 2004 ; Stephan & Stephan, 1985 ) and misattributions of partners’ intentions and behavior ( Dovidio, Pearson, Smith-McLallen, & Kawakami, 2005 ; Shelton & Richeson, 2005 ; Trawalter & Richeson, 2006 ; Vorauer, 2006 ; West, Shelton, & Trail, 2009 ). As a result of these negative experiences, people have less interest in sustaining cross-race relationships than same-race ones, even after prolonged contact ( Levin, van Laar, & Sidanius, 2003 ; Swart, Hewstone, Christ, & Voci, 2011 ; West et al., 2009 ).

In contrast, during the development of same-race relationships, a sense of similarity from the onset can provide a buffer against psychological and behavioral processes that could potentially hinder relationship formation. Robbins and Krueger (2005) found that individuals tend to assume that ingroup members share their attitudes and traits more than outgroup members, which might be one reason why they are motivated to affiliate more with ingroup members ( Pettigrew, 1998 ; Robbins & Krueger, 2005 ). This tendency to perceive similarity to ingroup members can perhaps be beneficial when partners are faced with potential disruptions to the relationship, particularly in its early stages. For example, in one study involving interactions over closed-circuit television, same-race partners were actually less anxious when there was a disruptive 1-s delay than when there was no delay, and the delay did not hurt their desire to become friends with their partner ( Pearson et al., 2008 ). Within interracial interactions, however, the delay induced more anxiety for both partners. A study by West et al. (2009) comparing same-race with cross-race roommates parallels these results. Among same-race roommates, anxiety experienced by one roommate positively predicted the other roommate’s interest in living together the following day, suggesting that people are inclined to give the “benefit of the doubt” to their same-race partners by not assuming that their partner’s anxiety is a signal of dislike ( West et al., 2009 ). By contrast, within cross-race roommate relationships, anxiety experienced by one roommate negatively predicted the other roommate’s interest in living together. Presumably, the slack afforded to same-race roommates was not extended to different-race roommates. Furthermore, West, Pearson, and Stern (2014) found that within same-race interactions, anxiety expressed by one partner prompted individuals to engage in compensatory behaviors, such as increasing self-disclosure to help ease the interaction, whereas within cross-race interactions, anxiety expressed by one partner prompted individuals to disengage from the interaction (for a review, see West, 2011 ). Together, these findings demonstrate that cross-race interactions are more fragile than same-race interactions during early stages of relationship formation, and one apparent barrier to more successful cross-race interactions is an assumption of psychological difference between partners ( Mullen, Dovidio, Johnson, & Copper, 1992 ; Rokeach & Mezei, 1966 ).

Theoretical Basis for Our Approach

The negative inferences that plague cross-race interactions might be reduced if individuals perceived greater similarity with their interaction partners. Indeed, the notion that finding common ground with outgroup members can improve intergroup relations has a long history in social psychology. In one classic study, Stein (1966) demonstrated that Whites who were prompted to think about being of the same religion as Blacks reported greater openness to cross-race contact (see also Rokeach & Mezei, 1966 ; Stein, 1966 ; Stein, Hardyck, & Smith, 1965 ).

Contemporary research has continued to emphasize the importance of perceiving similarities as a means of overcoming race-based biases. According to Gaertner and Dovidio’s (2000) common ingroup identity model, making similarities between ingroup and outgroup members salient is a cornerstone of improving intergroup interactions. For example, when members of different groups share a common identity, such as attending the same university, they extend the affective and cognitive benefits of ingroup categorization to members of the outgroup ( West, Pearson, Dovidio, Shelton, & Trail, 2009 ). Though this model is noteworthy for its success at improving intergroup relations (e.g., Houlette et al., 2004 ; Nier et al., 2001 ; Penner et al., 2013 ), it has an important limitation: The groups must share a preexisting, meaningful common identity. In the prior example, individuals must have a strong identification with their university for recategorization to the superordinate identity to occur.

Other researchers have explored using attributes of similarity that do not depend on shared social category membership to enhance intergroup relations. Pinel and Long (2012) called attention to the distinction between two possible sources of interpersonal similarity—stable aspects of the self, such as personal and social identities (as in the common ingroup identity model) versus fleeting, subjective experiences, such as laughing at the same joke—and found that sharing subjective experiences increases intergroup attraction more than sharing common identities. Pinel and Long also found that the sharing of subjective experience only needs to be perceived; if I think we have had a similar reaction to a situation, even if we have not, I will still like you more.

In their study of how to improve cross-race relations, Mallet, Wilson, and Gilbert (2008) also used an approach that deemphasizes social categories. In their studies, making incidental similarities to a different-race partner (e.g., sharing a preference for apples over oranges) salient to participants prior to an initial interaction was sufficient to reduce negative expectancies. Compared with Whites who focused on incidental differences from a cross-race partner, Whites who focused on similarities expected to like their partner more. However, the similarity manipulation had no effect on interpersonal processes during the interracial encounter. Although it is not clear why this was the case in the studies conducted by Mallett and colleagues, their findings are consistent with other research demonstrating that similarity manipulations tend to have large effects on attraction under conditions with no social interaction, but only small effects on attraction when social interaction is required (e.g., Sunnafrank, 1984 , 1986 ; Sunnafrank & Miller, 1981 ; see also Montoya et al., 2008 ).

Combining elements of intergroup relations and close relationships research, our approach to improving interracial interaction is designed to improve both expectations prior to cross-race interactions and processes during those interactions. From research on intergroup relations, we borrow the basic notion that similarity improves cross-race encounters ( Gaertner & Dovidio, 2000 ), but like Mallett et al. (2008) and Pinel and Long (2012) , we emphasize the importance of perceived similarity over actual similarity. Although the notion that perceiving similarity is more important than actual similarity also has roots in the close relationships literature ( Condon & Crano, 1988 ; Hoyle, 1993 ; Montoya et al., 2008 ; Murray, 1999 ; Selfhout et al., 2009 ; Tidwell et al., 2012 ), the key insight we import from research on close relationships to research on intergroup relations is proposing that the dimension of similarity must be perceived as self-revealing for it to improve relational processes ( Lemay & Clark, 2008 ; Murray et al., 2002 ).

We also took into consideration the possibility that within cross-race interactions, individuals may be more sensitive about self-disclosure than within close relationships. Altman and Taylor (1973) found that revealing too much too soon is apt to disrupt new relationships, and this may be particularly true within cross-race relationships, in which individuals tend to spontaneously reveal less about themselves to their partners ( Shelton, Trail, West, & Bergsieker, 2010 ). Thus, if information disclosed during the getting-acquainted stage of an interaction is too personally revealing, individuals might experience an increase rather than a decrease in discomfort and anxiety ( Page-Gould et al., 2008 ), which could contribute further to contact avoidance. This research suggests that attributes of similarity that might bring cross-race relationships closer would only be somewhat revealing, rather than extremely revealing, about the self.

Furthermore, we reasoned that when the attributes on which self–other comparison occurs are peripheral to the interaction task, perceived similarity will improve interpersonal processes without interfering with the main goals of the interaction. As such, we reasoned that a manipulation of perceived similarity would not require participants to actively discuss their attributes of similarity during the encounter. This method of skipping over discussion of the similarities allows for perceived similarity not only to operate above and beyond any actual levels of similarity (and even if partners are actually dissimilar) but also to improve interracial interactions across multiple contexts (e.g., when partners are casually becoming acquainted, or when they are trying to accomplish a task in a limited period of time). Our challenge, then, was to find a source of similarity that was both self-revealing and peripheral to the interaction in order to improve interracial interactions across different contexts.

Basic Prediction and Testing an Underlying Assumption

We predicted that our manipulation of perceived similarity would be particularly beneficial within cross-race interactions because in those interactions, individuals tend to assume significant dissimilarity from their partner. That is, we assumed that there is room to shift cross-race partners’ perceptions toward greater similarity. In contrast, we reasoned that same-race partners would already perceive substantial similarity, putting a ceiling on the extent to which their judgments of similarity could be manipulated.

Although there is evidence that individuals perceive themselves to be more similar to racial ingroup than outgroup members ( Robbins & Krueger, 2005 ), we first sought to directly establish that individuals perceive themselves to be more similar to same-race partners than to different-race partners in the nascent stages of interpersonal relationships. As part of another longitudinal study ( West, Pearson, Dovidio, Shelton, & Trail, 2009 ), we asked 130 randomly assigned first-year roommates at a university in the northeastern United States to report how similar they felt to each other after they had lived together for 2 weeks. Twenty of the 65 roommate pairs were different-race (five Black–White, six Latino– White, nine Asian–White), and 45 were same-race (White–White) pairs. Participants individually completed an online questionnaire during the second week of the fall semester, and embedded within the questionnaire, participants were asked to report the extent to which they agreed with the following statement (from 1 = not at all to 7 = very much ): “My roommate and I seem like similar people.”

We compared the responses of same-race (White–White) with cross-race (White–minority) roommates, and Whites with minorities within cross-race roommate pairs, adjusting for nonindependence of dyad members’ responses ( Kenny, Kashy, & Cook, 2006 ). As we expected, same-race roommates perceived more similarity to each other than did cross-race roommates, t (63) = −3.41, p = .001. Also consistent with our expectations, the similarity mean for same-race roommates was high in an absolute sense ( M = 5.33, SD = 1.25), whereas the mean for cross-race roommates was around the midpoint of the scale ( M = 4.25, SD = 1.64). Within cross-race roommate pairs, Whites and minorities were not significantly different from each other ( p = .23).

Having established that people perceive more similarity to same-race than cross-race partners in the nascent stages of relationship formation and that same-race partners perceived similarity at a high absolute level, we turned to our hypothesis that a manipulation of perceived incidental similarity on a self-relevant attribute would positively influence cross-race interaction partners, but have a negligible effect on same-race interaction partners.

Overview and Predictions

In three experiments, we focus on improving the quality of interpersonal processes within cross-race interactions across different interaction contexts, all of which involve early stage interactions but different interaction tasks. We theorized that perceiving similarity on attributes that are self-relevant but peripheral to the interaction would enhance processes that are particularly important for intergroup relationship development (i.e., reduce anxiety and increase interest in future contact) and also enhance general communication processes between partners (i.e., increase empathic accuracy and improve behavioral coordination during a task that requires communication between partners). We focus on these outcomes in particular because they are widely associated with relationship maintenance and satisfaction in close and casual relationships and have been shown to be improved by perceived similarity within close relationships in particular.

In Experiment 1, we compared the influence of similarity attributes that are relatively high versus low in self-revelation on anxiety prior to delivering a speech to an ostensible cross-race (White–minority) versus same-race (White–White) partner. In Experiment 2, we explored whether making partners aware of actual similarities on self-revealing attributes reduces anxiety and increases interest in sustained contact in cross-race versus same-race getting-acquainted interactions. We also examined empathic accuracy in partners’ perceptions of one another’s interest in the encounter. In Experiment 3, participants worked together in racially diverse groups on a task that requires effective communication in order to perform well. We explored whether or not manipulating perceived similarity on peripheral self-revealing attributes influences behavioral coordination and performance within these groups.

Experiment 1

The goal of Experiment 1 was to directly test the hypothesis that attributes of perceived similarity must be self-revealing to benefit participants in an interracial interaction setting. We compare the extent to which similarity based on self-revealing versus less self-revealing attributes reduces anxiety in anticipation of an ostensible cross-race versus same-race encounter. We focused specifically on anticipatory anxiety given its negative effects on initiating and sustaining intergroup contact ( Islam & Hewstone, 1993 ). Participants learned that they and their partner would each give a speech about why they would make a good friend, prior to interacting. This task has been found to be anxiety-provoking and physiologically stressful ( Kirschbaum, Pirke, & Hellhammer, 1993 ), particularly in the context of cross-race interactions ( Blascovich et al., 2001 ).

We hypothesized that the similarity and self-revealing manipulations would interact to reduce anxiety in anticipation of giving a speech to a cross-race but not a same-race partner. Specifically, we expected that learning that one’s cross-race partner is similar to oneself on peripheral, self-revealing attributes would lead participants to experience less anxiety compared with learning that one’s partner is similar on peripheral but less self-revealing attributes. However, based on our finding that individuals in same-race interactions are apt to assume similarity (see also Robbins & Krueger, 2005 ), we predicted the effects on anxiety would be weaker for individuals anticipating giving a speech to a same-race partner.

Within the cross-race condition, we were also interested in whether similarity on self-revealing attributes would be equally effective at reducing Whites’ and minorities’ anxiety. Although some research has demonstrated that Whites and minorities are equally anxious during cross-race interactions (e.g., Pearson et al., 2008 ; West et al., 2009 ), other studies have revealed that, in cross-race interactions, minorities tend to experience less negative affect than Whites (and comparable levels to minorities in same-race interactions; for a meta-analysis, see Toosi et al., 2012 ). Some minorities might experience less negative affect in cross-race interaction because they are more accustomed than Whites to interracial interactions in their daily lives. Thus, we tested whether minorities and Whites would benefit equally from believing that they were similar on self-revealing attributes, prior to giving a speech on why they would make a good friend.

Participants and design

Participants were 136 Black, White, Latino/a, Asian, and minority multiracial individuals (106 women; M age = 19.88) who were students recruited through the New York University psychology department participant pool ( n = 121) and New York City community members recruited through craigslist.org ( n = 15). The experiment was a 2 Similarity (high vs. low) × 2 Self-Revelation (high vs. low) × 3 Dyad Race Composition (White–White, White–Black, minority–White) between-subjects design. Whites believed they would interact with either a White partner ( n = 40) or a Black partner ( n = 35). All minorities ( n = 61; 13 Black, 14 Latino/a; 30 Asian; four non-White multiracial) believed they would interact with a White partner. Men believed they would interact with a man, and women believed they would interact with a woman. In reality, no interaction actually took place.

Upon arrival, participants were informed that, after completing some initial surveys, they would each give a 1-min videotaped speech on why they would make a good friend to a new acquaintance who ostensibly had already arrived at the session. They were also informed that after watching each other’s speech, they would interact with the new acquaintance. These aspects of the procedure raised the stakes of the interaction for participants, increasing the potential for them to experience anxiety about the task and the interaction. Participants who reported that they did not believe the interaction would take place were excluded from the analyses ( n = 2).

Further, they were told that before the interaction, they would exchange some information about themselves. They completed an information form , which contained the manipulations. The experimenter then provided participants with a completed information form that was ostensibly from their partner (and were told that the partner would see their form). After having time to look at this form, participants completed the dependent measure of anticipatory anxiety .

The information form contained the manipulations of similarity, self-revelation of the similarity, and partner race.

Manipulations

We had two goals in selecting sets of attributes with which to manipulate similarity. First, we took into consideration that learning “too much too soon” about a partner can increase rather decrease anxiety in interracial interactions ( Page-Gould et al., 2008 ), so we selected attributes that individuals felt comfortable disclosing to a new acquaintance. Second, we sought to select attributes for which there is no culturally endorsed or consensual preference. People prefer to interact with others who possess desirable traits, regardless of their similarity on those traits (e.g., regardless of their own level of neuroticism, people prefer to interact with others who are low on neuroticism; Anderson, John, Keltner, & Kring, 2001 ; Klein, Lim, Saltz, & Mayer, 2004 ). Thus, we avoided personality constructs (e.g., Big Five personality traits) and other highly desirable or undesirable attributes. Instead, we selected attributes for which individuals reported their own preferences without any sense of what other people preferred.

Pilot study to select attributes for the similarity and self-revelation manipulations

To select attributes that were relatively high versus low on self-revelation, participants (47 undergraduates from New York University and 82 Amazon Mechanical Turk participants) provided responses to a number of would-you-rather (WYR) questions in which they selected between one of two options. Forty-two potential high-self-revealing items were drawn from Horn (2001) —a book of social games—and 27 potential low-self-revealing items were drawn from Mallett et al. (2008) .

Participants were first prompted with the following:

Imagine that you are about to interact with a new acquaintance. For each of the following Would-You-Rather questions, please think about how you would feel if this person saw your response before you interacted. You would not discuss your answers with this person, they simply would know which of the 2 choices you picked.

For each WYR dilemma, participants selected one of the two options and then responded to two additional questions. To select items that participants felt were high versus low on self-revelation, participants responded to the item: “Does your response convey anything about your personality?” To assess comfort with sharing one’s preferences, participants responded to the item: “How comfortable would you be in knowing that this person [the interaction partner] is going to see your response?” These two questions were assessed on scales that ranged from 1 ( not at all ) to 7 ( very much ).

Participants rated seven of the 27 dilemmas from Mallett et al. (2008) as low on the extent to which they were revealing of their personality (at least 1 point below the midpoint of the scale; range = 2.30–2.97) and that they would be comfortable disclosing them to an interaction partner (range = 6.21–6.38). The average difference in frequency of the two responses was 16.8% (58.4 % vs. 41.6 %); a 0% difference would indicate that both options were chosen with equal frequency. These comprised the low-self-revealing WYR dilemmas (see Appendix A ).

From the 42 dilemmas from Horn (2001) , we selected seven (see Appendix B ) that participants reported as moderately highly self-revealing to their personalities (range = 4.13–5.67). As with the dilemmas low on self-revelation, participants reported that they would feel comfortable disclosing their responses on these seven items to a new acquaintance (range = 5.28–6.24). The average difference in frequency of the two responses was 11% (55.5% vs. 44.5%). These comprised the high-self-revealing WYR dilemmas.

Important for our purposes, the dilemmas were significantly different on the extent to which they are self-revealing and similar on the extent to which participants would be willing to disclose their responses to interaction partners in our experiments. 1

Information form

On the information form, participants completed seven high-self-revealing or seven low-self-revealing WYR dilemmas. In the high-similarity condition, participants saw that they had five of seven answers in common with their partner, and, in the low-similarity condition, they saw that they had two of the seven answers in common. Two versions of the specific items of similarity/dissimilarity were counterbalanced to ensure that similarity on one set of items was not responsible for any effects of the manipulation.

To manipulate partner race , the partner’s information form indicated the partner’s race (White or Black) and name (Caitlin or Bradley in the White partner condition, Shanice or Darnell in the Black partner condition). The form held constant the partner’s gender (same as the participant’s), age (19), and nationality (American), which were included to make the partner’s race appear as part of a standard presentation of demographic information.

Anticipatory anxiety measure

Following the prompt, “Before you give your speech and watch your partner give his/her speech, we would like to get a sense of how you are feeling,” participants responded to the following four items on 7-point scales: anxious, nervous, uncomfortable , and uncertain (adapted from Britt, Boniecki, Vescio, Biernat, & Brown, 1996 , and Stephan & Stephan, 2000 ). These items were averaged to create a composite measure of anticipatory anxiety (α = .81).

We analyzed the data with a general linear model that included terms for similarity, self-revelation, dyad race, and all two-way and three-way interactions. We also included a term for gender. In an initial analysis, we included the interaction between racial composition of the dyad (same- vs. cross-race) and gender, as other researchers have found gender differences in cross-race interactions ( Toosi et al., 2012 ). However, this analysis revealed that men and women did not demonstrate a different pattern of effects in same-race versus cross-race dyads ( p = .15), and this interaction term was not included in the analyses presented below. We did not have any expectations of differences among minorities, and an examination of the means suggested similar effects across races. However, we did not have sufficient power to compare the effect of the manipulation for minorities of different races.

Anticipatory anxiety

The hypothesized Dyad Race × Similarity × Self-Revelation interaction was significant, F (2, 123) = 3.28, p = .041. To decompose the three-way interaction, we conducted pairwise comparisons, which revealed that the Similarity × Self-Revelation interaction was driven by Whites in the cross-race condition. As seen in Figure 1 , top panel, for Whites in the high-self-revealing condition, those who believed they had high similarity with their Black partner felt significantly less anxious than did those who believed they had low similarity, F (1, 123) = 5.62, p = .019. However, for Whites in the low-self-revealing condition who anticipated an interaction with a Black partner, the main effect of similarity was not significant, F (1, 123) = 1.96, p = .164. We also note that for Whites in the high-similarity cross-race condition, those in the high-self-revealing condition were marginally less anxious than those in the low-self-revealing condition ( p = .058), consistent with our hypotheses.

Top panel: Anticipatory anxiety for Whites anticipating an interaction with a Black partner as a function of similarity and self-revelation condition in Experiment 1. Middle panel: Anticipatory anxiety for Whites anticipating an interaction with a White partner as a function of similarity and self-revelation condition in Experiment 1. Bottom panel: Anticipatory anxiety for minorities anticipating an interaction with a White partner as a function of similarity and self-revelation condition in Experiment 1. Error bars represent standard errors of the means.

For White participants in the same-race (White partner) condition and minority participants in the different-race (White partner) condition, there were no differences in anxiety as a function of similarity, self-revelation, or their interaction ( p s > .16) (see Figure, middle and bottom panels). For these participants, anxiety was relatively low across conditions.

Only one other effect was noteworthy, although only marginally significant: Men felt marginally more anxious ( M = 3.61, SD = 1.24) than women ( M = 3.16, SD = 1.33) in anticipation of the speech, F (1, 123) = 3.65, p = .058.

Results of Experiment 1 provide evidence that learning that one’s partner is similar to oneself on self-revealing attributes reduces Whites’ anxiety in anticipation of trying to convince a prospective different-race partner of why he or she would be a good friend. Consistent with our theorizing that individuals within same-race encounters may be less affected by incidental similarity than individuals in cross-race encounters, we found that the similarity and self-revelation manipulations did not exert a significant influence on Whites’ anxiety in the same-race conditions.

Also, in contrast to the effect for Whites in the cross-race condition, we found that minorities were not affected by the similarity and self-revelation manipulations. In fact, minorities in the cross-race condition experienced levels of anxiety similar to Whites’ levels in the same-race condition. In addition to the possibility that minorities are more accustomed to interracial contact than are Whites, it is also possible that the manipulations only exerted their intended effects on Whites because Whites and minorities have different bases of intergroup anxiety and thus different impression management concerns in cross-race interaction. Whites with low levels of prejudice, which probably characterizes the vast majority of Whites in our sample, tend to feel significant anxiety due to a concern about appearing prejudiced to minorities. One way to allay this concern is to try to appear likable when interacting with minority partners ( Bergsieker, Shelton, & Richeson, 2010 ). By contrast, minorities’ anxiety stems more from a concern about being the target of prejudice ( Shelton, Richeson, & Salvatore, 2005 ), and to reduce this concern, minorities might emphasize appearing competent versus likable to gain the respect of their White partner ( Bergsieker, Shelton, & Richeson, 2010 ). We think it is likely that the speech topic—why they would make a good friend—appealed more to Whites’ principal impression management concern of likability than to minorities’ concern about being respected in cross-race interactions. Whites who anticipated that they would be similar on self-revealing attributes to their cross-race partners might have had reduced concerns about appearing prejudiced and thus less anxiety than participants in the other conditions.

In the next experiment, to reduce the possibility that task demands would affect Whites and minorities differently, we used a task involving less explicit emphasis on appearing likable.

Experiment 2

In Experiment 2, we moved beyond the anticipation stage, to a dyadic interaction in which participants have a conversation with a new acquaintance, taking turns asking and answering a set of questions that encourage self-disclosure. This task has two distinct advantages over the task used in Experiment 1. First, it is similarly anxiety provoking for both Whites and minorities in a cross-race setting ( Page-Gould et al., 2008 ). Second, it allows us to explore the effectiveness of perceived similarity during actual encounters. This is important because the psychological experience of actual cross-race interactions diverges from the expectancy stage in some cases ( Mallett et al., 2008 ; cf. Crisp & Turner, 2012 ). A getting-acquainted interaction context also afforded an opportunity to measure partners’ interest in sustained contact, which is a necessary precursor to improving intergroup friendships ( Pettigrew, 1998 ). We measured not only each participant’s interest in sustained contact with their partner but also their empathic accuracy at detecting their partner’s interest in sustained contact.

Empathic Accuracy in Intergroup Interactions

Several recent studies have documented the difficulties that cross-race partners face in reading one another’s thoughts, feelings, and relationship intentions ( Demoulin, 2008 ; Pearson et al., 2008 ; West & Dovidio, 2013 ). Whites and minorities underestimate the extent to which racial outgroup members are interested in forming relationships ( Shelton & Richeson, 2005 ; Vorauer & Sakamoto, 2006 ) and assume that their social overtures communicate more interest to outgroup partners than they actually do ( Vorauer, 2006 ). These inaccurate perceptions can inhibit Whites and minorities from interacting over time ( Shelton & Richeson, 2005 ; West, 2011 ). Thus, improving empathic accuracy, particularly with respect to perceptions of interest in sustaining the relationship, is important for improving race relations in the long term; however, research on improving accuracy in reading out-group partners’ relationship interest is rare.

Why might perceiving similarity on self-revealing attributes improve accuracy within cross-race encounters? In close relationships, empathic accuracy stems in part from a motivation to understand one’s partner’s thoughts and feelings, particularly when individuals feel close to their partners and want to sustain a relationship ( Gagné & Lydon, 2004 ; Luo & Snider, 2009 ; Neff & Karney, 2005 ). We reasoned that by bringing partners closer, our manipulation of similarity on self-revealing attributes would increase individuals’ motivation to understand their cross-race partner’s interest in becoming friends. In trying to understand their partner’s relationship interest, we expected those who believed they were similar to attend more carefully to their partners’ behaviors and thus to exhibit higher levels of empathic accuracy.

In the current experiment, we predicted that in cross-race interactions, greater levels of perceived similarity would be associated not only with less anxiety and greater interest in sustained contact but also with greater empathic accuracy. We also reasoned that within same-race interactions, perceived similarity would have a minimal effect on anxiety, sustained contact, and empathic accuracy, given relatively high levels of perceived similarity that already exist in these encounters.

Having established in Experiment 1 that the attributes of similarity must be self-revealing to reduce anxiety, we used only self-revealing attributes in this experiment. We manipulated perceived similarity between previously unacquainted interaction partners in their responses to self-revealing WYR dilemmas by manipulating whether partners could see each other’s responses prior to interacting. That is, in this experiment, any similarities (and differences) between partners were real, and the manipulation involved whether or not they were made aware of those similarities (and differences). When participants were not shown their partner’s WYR responses, they were in a control condition (with respect to the manipulation of perceived similarity), which allowed us to directly compare dyads whose partners were equivalent in actual level of similarity but differed in whether or not they were aware of that level of similarity. We expected that awareness of similarity would be more important than actual similarity in benefiting cross-race interactions. Following the manipulation, participants completed a “getting-acquainted” exercise and then independently completed a questionnaire containing the dependent measures.

Participants were 100 previously unacquainted Black, White, Latina, and Asian female undergraduate students at New York University who made up 50 dyads (32 White–White, 18 White–minority; eight Black, eight Latino, two Asian). The participants were between 17 and 32 years of age ( M = 19.02). The experiment was a 2 (partner’s self-revealing WYR responses revealed vs. not revealed) × 2 (dyad race: same-race vs. cross-race) between-dyads design with one measured variable that was central to our hypothesis, similarity in WYR responses.

At the beginning of the session, the experimenter ensured that the participants did not know each other and then escorted them to separate rooms where they were informed that they would engage in a 6-min getting-acquainted conversation with their partner. They then completed one of two WYR questionnaires. Each questionnaire included six different WYR dilemmas that had been identified as self-revealing in the selection procedure described in Experiment 1. We used two versions of the WYR questionnaire in this study to ensure that any effects obtained were not dependent on a particular set of dilemmas. 2 Partners always completed the same questionnaire so that similarity could be measured and were given an information sheet on which they reported their age, their student status, and their race.

In the WYR responses revealed condition, participants were provided with their partner’s information sheet and WYR form. These two forms were used in tandem to ensure that participants knew their partner’s race prior to the start of the interaction. In the WYR responses not revealed condition, participants were only provided with their partner’s information sheet. All participants then completed a preinteraction questionnaire assessing their feelings leading up to the interaction.

Prior to the start of the interaction, participants were independently told not to discuss the WYR dilemmas with their partners. Then, they were escorted to a room where they sat facing each other with a visible video camera in front of them. Both individuals were given a list of six questions from Aron, Melinat, Aron, Vallone, and Bator’s (1997) interpersonal closeness procedure designed for initial interactions between new acquaintances and were instructed to take turns asking and answering each question. At the completion of the 6-min interaction, participants were escorted back to their separate rooms where they completed a postinteraction questionnaire. Videotapes of the interactions were then coded to ensure that no one discussed the WYR dilemmas during the interaction.

Actual similarity

We calculated actual similarity by averaging the number of WYR responses that partners had in common. Actual similarity ranged from 1 ( 1 answer out of 6 in common ) to 6 ( all answers in common ). On average, partners responded identically to roughly half of the dilemmas ( M = 3.33, SD = 1.03).

We measured anxiety prior to and after the interaction. Anticipatory anxiety was measured in the same way as in Experiment 1 (α = .85). A similar set of items measured interaction anxiety on the postinteraction questionnaire. Participants reported how they felt “during the interaction” ( uncertain of how to behave, nervous, uncomfortable , and awkward ; α = .84).

Interest in sustained contact

We measured participants’ own feelings of interest in sustained contact and perceptions of their partner’s interest with composites of 7-point scale responses to the following questions: “How much did you (your partner) enjoy the interaction?” “How much would you want to become friends with this person (would this person want to become friends with you)?” “How much did you like this person (did this person like you)?” and “Would you want to have another interaction with this person (would this person want to have another interaction with you)?” (α self = .86 and α partner = .90).

Analysis Strategy

To examine the primary variables of interest, we performed dyadic analyses using the MIXED procedure in SPSS for the analysis of indistinguishable dyadic data, which accounts for non-independence in dyad members’ responses ( Kenny, Kashy, & Cook, 2006 ). We note that this method can yield fractional degrees of freedom (see Kenny et al., 2006 , for an explanation). In all models, whether WYR responses were revealed and similarity in WYR responses (grand mean centered) was treated as between-dyad variables.

As with Experiment 1, we had no expectations about differences between minorities, nor the statistical power to formally test for differences, but an examination of means indicated no differences within cross-race dyads on the outcomes of interest. Thus, we treated all minority racial categories as identical in our analyses (i.e., Black, Asian, and Latina were all treated as minority). To simultaneously examine differences between White/White and White/minority dyads and differences between Whites and minorities within these dyads, the following two contrasts were included in all analyses: between-dyad race, which compares same-race with cross-race dyads, and within-dyad race, which compares Whites with minorities within cross-race dyads. All models included WYR responses revealed condition (hereafter, responses revealed), similarity, between-dyad race, within-dyad race, and interactions between all of these variables. We included a contrast code (1, −1) to indicate which of the two WYR questionnaires the dyad received.

Following the guidelines of Aiken and West (1991) , in Figures 2 and ​ and3, 3 , we plot predicted means for participants that were relatively similar to their partner (i.e., 1 SD above the similarity mean, or 4.63 WYR responses in common) and for those that were relatively dissimilar (i.e., 1 SD below the similarity mean, or 2.03 WYR responses in common).

Top panel: Relationship between similarity and anticipatory anxiety by WYR revealed condition, cross-race dyads (combining Whites and minorities), in Experiment 2. Middle panel: Relationship between similarity and anticipatory anxiety by WYR revealed condition, same-race dyads, in Experiment 2. WYR = would you rather.

Top panel: Relationship between similarity and interest in sustained contact for White participants with minority partners (WM), minority participants with White partners (MW), and White participants with White partners (WW), WYR revealed condition, in Experiment 2. Bottom panel: Relationship between similarity and interest in sustained contact for White participants with minority partners, minority participants with White partners, and White participants with White partners, WYR not revealed condition, in Experiment 2. WYR = would you rather.

The model for anticipatory anxiety revealed a main effect of similarity, t (42) = −2.51, p = .016. Important for our prediction that revealing similarity between partners would benefit cross-race interactions more than same-race interactions, we found a significant three-way Responses Revealed × Similarity × Between-Dyad Race interaction, t (42) = −2.05, p = .046. We did not find a significant Responses Revealed × Similarity × Within-Dyad Race interaction ( p = .311), indicating that the patterns of effects for Whites and minorities within cross-race dyads were not significantly different from each other. Thus, in Figure 2 , top panel, we collapsed the data for Whites and minorities within cross-race dyads and describe the different patterns of data for White/minority and White/White dyads below.

White/minority dyads

As shown in Figure 2 , top panel, there was a significant two-way Responses Revealed × Similarity interaction for participants in cross-race dyads, t (42) = −2.09, p = .043. Consistent with our hypotheses, for participants in cross-race dyads in which WYR responses were revealed, the more similar respondents were to their partners, the less anxiety they felt in anticipation of the interaction, t (42) = −2.49, p = .017. In contrast, when WYR responses were not revealed, there was no effect of similarity on anticipatory anxiety, t (42) = −0.45, p = .65. This finding indicates that when cross-race partners were actually similar in WYR responses, but were not aware that they were similar, there was no effect of similarity on anticipatory anxiety.

White/White dyads

As expected and shown in Figure 2 , bottom panel, the two-way Responses Revealed × Similarity interaction was not significant for participants in White/White dyads, t (43) = .490, p = .627.

Interaction anxiety

For interaction anxiety, we found a main effect of similarity, t (42) = −3.82, p < .01. As with anticipatory anxiety, the three-way Responses Revealed × Similarity × Within-Dyad Race interaction was not significant ( p = .255). Also consistent with anticipatory anxiety, we found a significant three-way Responses Revealed × Similarity × Between-Dyad Race interaction, t (42) = −2.38, p = .022, which we decompose below by separating the effects for White/minority and White/White dyads.

The two-way Responses Revealed × Similarity interaction was significant for participants in White/minority dyads, t (42) = −3.13, p = .003. Consistent with our hypotheses, for participants in cross-race dyads in which the WYR responses were revealed, the more similar participants were to their partners, the less anxiety they felt during the interaction, t (42) = −3.74, p = .001. When WYR responses were not revealed, there was no effect of similarity on interaction anxiety, t (42) = −0.85, p = .40.

Also consistent with the results for anticipatory anxiety, in White/White dyads, the two-way Responses Revealed × Similarity interaction was not significant for interaction anxiety, t (43) = −0.630, p = .532.

The model for interest in sustained contact revealed a main effect of similarity, t (42) = 2.18, p = .035. In contrast to the results for interaction anxiety, the Responses Revealed × Similarity × Between-Dyad Race interaction for interest in sustained contact was not significant, t (42) = 1.56, p = .126; however, the Responses Revealed × Similarity × Within-Dyad Race interaction was significant, t (427) = 2.35, p = .023, indicating that the Responses Revealed × Similarity interaction differed for Whites and minorities within cross-race dyads.

Specifically, in White/minority dyads, the Responses Revealed × Similarity interaction was not significant for minorities, t (83.74) = −0.314, p = .754, but it was significant for Whites, t (83.74) = 2.67, p = .009 (see Figure 3 , top and bottom panels). As seen in Figure 3 , top panel, for Whites in cross-race dyads in which the WYR responses were revealed, the more similar respondents were to their partners in their WYR responses, the more interested they were in sustaining contact, t (83.64) = 2.89, p = .005. When WYR responses were not revealed in cross-race dyads, Whites experienced no effect of similarity on interest in sustained contact, t (84.02) = −0.022, p = .983 (see Figure 3 , bottom panel). This finding is consistent with Experiment 1. 3

Also consistent with Experiment 1 and as seen in Figure 3 , top and bottom panels, the Responses Revealed × Similarity interaction was not significant for Whites in same-race dyads, t (42) = −0.447, p = .66.

Accuracy in reading partners’ interest in sustained contact

To examine whether similarity in WYR responses influenced accuracy in perceptions of partners’ interest in sustained contact, we used West and Kenny’s (2011) truth and bias model. Note that in this model, all participants are perceivers and partners . The model simultaneously estimates two “forces”: the truth force —the effect of the truth variable (i.e., the partner’s interest in contact) on the judgment (i.e., the perceiver’s evaluations of their partner’s interest in contact)—measures accuracy; the bias force —the effect of the bias variable (i.e., the perceiver’s interest in contact) on the judgment—measures assumed similarity (e.g., If I am interested in contact, I assume my partner is also interested in contact). We adjusted for assumed similarity to allow for an examination of direct accuracy—the amount of accuracy left over once we account for accuracy that is achieved indirectly (i.e., through correctly assumed similarity; see West & Kenny, 2011 ). This way of measuring accuracy provides a cleaner test of our hypotheses because it leaves us with a measure of accuracy that is directly attributable to correctly inferring the partner’s interest in sustained contact, above and beyond that which is achieved by relying on one’s own feelings of interest in contact ( Fletcher & Kerr, 2010 ; Overall & Hammond, 2013 ; West & Kenny, 2011 , for a review).

Preliminary analyses revealed no differences in accuracy and assumed similarity between Whites and minorities within cross-race dyads. Thus, we describe the results of a simplified model excluding the within-dyad race contrast. To examine whether accuracy (the truth force) and assumed similarity (the bias force) varied as a function of whether responses revealed condition, similarity, or the between-dyad race variable, we included terms for the interactions between these variables and the partner’s self-ratings of interest in contact and the perceiver’s own interest in contact (for a similar strategy, see Case 4 in West & Kenny, 2011 ).

Assumed similarity was positive and significant, t (9.85) = 84.84, p < .001. Overall, perceivers showed no direct accuracy, t (82.17) = 0.12, p = .91. There was a significant two-way interaction between the truth variable and responses revealed, t (79.42) = 2.54, p = .013, which was qualified by a significant Truth Variable × Responses Revealed × Similarity × Between-Dyad race interaction, t (84.10) = −2.69, p = .009. The pattern of results for this effect revealed that similarity in WYR responses facilitated accuracy in perceptions of interest in contact for cross-race dyads but hindered accuracy in same-race dyads, as we describe in further detail below.

For White/minority dyads, there was a marginally significant Truth Variable × Responses Revealed × Similarity interaction, t (83.11) = 1.98, p = .052. When WYR responses were revealed, the Truth Variable × Similarity interaction was significant and positive, t (84.50) = 2.12, p = .037, indicating that there was a significant positive effect of similarity on accuracy. However, when WYR responses were not revealed, the Truth Variable × Similarity interaction was not significant, t (75.95) = −0.29, p = .773, indicating that there was no effect of similarity on accuracy. Thus, as hypothesized, individuals in cross-race dyads were more accurate in reading their partners’ relationship intentions when their responses were revealed and were similar.

For White/White dyads, the Truth Variable × Responses Revealed × Similarity interaction was significant, t (80.24) = −2.28, p = .025. When WYR responses were not revealed, The Truth Variable × Similarity interaction was not significant, t (77.09) = 0.77, p = .446, indicating that there was no effect of similarity on accuracy. However, when WYR responses were revealed, the Truth Variable × Similarity interaction was significant and negative, t (81.70) = −2.48, p = .016, indicating that similarity was associated with less accurate perceptions of partners’ interest in contact. Surprisingly, learning that one’s same-race partner was similar to oneself decreased accuracy for White perceivers with White partners.

In Experiment 2, we found that when partners within cross-race interactions realized they were similar on self-revealing attributes, they experienced a number of benefits. Whites experienced reduced feelings of anxiety both in anticipation of and during their interactions with minority partners, were more interested in sustaining contact, and perceived their partners’ level of interest in contact more accurately. In summary, Whites in interracial interactions benefited across all measured outcomes from perceiving similarity to their partners.

We reasoned that Whites’ and minorities’ impression management concerns within interracial interactions would be better balanced during the getting-acquainted task than during the friendship speech task of Experiment 1, thereby creating a context in which minorities could also benefit from perceiving similarity. Consistent with this reasoning, we found that minorities who perceived similarity to their White partners experienced less anxiety in anticipation of and during the interaction, and were more accurate in reading their partners’ interest in contact. Contrary to hypotheses, however, minorities’ own interest in contact was no different from the interest in contact reported by minorities in the other conditions. In fact, we were surprised to find that minorities’ interest in contact was relatively high across conditions. We suspect this might be due to the experience that minority group members have with a disproportionate number of interactions occurring across race lines. We discuss this possibility for why the manipulation of perceived similarity had different effects on Whites versus minorities in more detail in the General Discussion.

Whites’ anxiety and interest in contact with other Whites were unaffected by perceiving similarity to their partners. However, we found that Whites who perceived more similarity to their White partners were surprisingly less accurate at interpreting their partners’ interest in sustained contact. Why might this be the case? We have argued throughout that perceiving similarity on self-revealing attributes would motivate perceivers to attend to their partners, which would increase accuracy. However, it may be the case that in same-race interactions, because there is already a strong basis of similarity, partners were less attentive to each other’s behaviors that signal interest in contact, leading to weaker “direct” accuracy (i.e., accuracy adjusting for any accuracy achieved through correctly assumed similarity). Although this explanation might seem counterintuitive, it has support in the close relationships literature. As relationships progress and partners become more familiar with each other, accuracy declines ( Kenny, 1994 ; Kilpatrick, Bissonnette, & Rusbult, 2002 ), presumably because partners are confident that they understand each other, and so they stop attending closely to each other’s behaviors ( Kenny, 1994 ).

We demonstrated in Experiments 1 and 2 that perceived similarity improves outcomes for cross-race partners within friendship-building contexts. In Experiment 3, we examined how our manipulation of perceived similarity influences interpersonal outcomes in another type of context—one that requires individuals to communicate effectively as members of a team working together on a complex task.

Experiment 3

There were several goals for Experiment 3. First, we wanted to demonstrate the broad applicability of our approach. To date, nearly all research on the benefits of perceiving similarity in interracial interactions has focused on affiliative measures, such as liking and the desire for interpersonal contact ( Mallett et al., 2008 ; Pinel & Long, 2012 ). Task groups provide an opportunity to explore performance outcomes and analyze coordination processes that contribute to performance.

Second, we chose to investigate racially diverse task groups, including many groups with more than two members from different minority groups. Despite repeated calls for more carefully crafted methods that could improve outcomes for racially diverse groups ( Paluck & Green, 2009 ; Van Knippenberg & Schippers, 2006; Williams & O’Reilly, 1998 ), there remains a shortage of empirical research on the topic. Racially diverse groups tend to perform worse than racially homogenous groups on many tasks ( Stahl, Maznevski, Voigt, & Jonsen, 2010 ), in part because their members are less effective at communicating and coordinating with each other ( Milliken & Martins, 1996 ). A possible cause of relatively poor communication and coordination in diverse groups is that Whites and minorities tend to infer psychological dissimilarity (e.g., on attitudes and preferences) from demographic dissimilarity (e.g., on race and sexual orientation) ( Chen & Kenrick, 2002 ; Phillips, 2003 ). We reasoned that perceiving similarity on self-relevant attributes might diminish communication and coordination difficulties in diverse groups. After all, perceived similarity on self-relevant attributes motivates individuals to understand one another’s perspectives and enhances communication ( Holmes & Rempel, 1989 ; Linden-Andersen et al., 2009 ; Murray et al., 2002 ). Therefore, we expected that our manipulation of self-revealing similarity to improve coordination and thus performance in diverse task groups.

Third, we wanted to examine the extent to which any effect of the manipulation on task performance was attributable to the perceived contributions of White versus minority group members. At the end of the task, participants made judgments of one another’s task contributions, which allowed us to explore (a) whether the perceived similarity manipulation influenced task performance—a group-level outcome in this context—through perceptions of individual group members’ task contributions and (b) whether perceived similarity equally affected Whites’ and minorities’ perceived contributions to the group’s performance.

In this experiment, groups were randomly assigned to perceived similarity condition such that all members within a group believed either that they were similar or that they were dissimilar in their WYR responses. We hypothesized that groups would perform more efficiently when they were told that their members were similar, and this effect would hold for all groups, given that they were all diverse (i.e., no groups were composed of members who were of one race or ethnicity). We also expected that the effect of group performance would be mediated by members’ task contributions, as rated by fellow group members. Assuming that both Whites and minorities would infer equivalent levels of psychological dissimilarity from demographic dissimilarity, as previous research has demonstrated within task groups ( Chen & Kenrick, 2002 ; Phillips, 2003 ), we reasoned that the similarity manipulation would improve ratings for Whites and minorities equally.

Participants

Participants were 110 (71 female) graduate students enrolled in an introductory management course at New York University (for a different analysis of these data, see West, Heilman, Gullett, Moss-Racusin, & Magee, 2012 ). The mean age was 26.41 years (range = 20–46), and the majority of participants had between 3 and 5 years of work experience. The sample was racially diverse (62 White, 48 racial minorities, among which there were 10 Blacks, 11 Latinos, 11 Asians, 13 multiracials, and three “others”). They completed the experiment in 22 groups of five that we designed to be racially heterogeneous; participants were randomly assigned to groups with respect to all other variables. The mean number of Whites per group was 2.81, and the minorities within each group were always of different races (e.g., no group had two Asians, two Blacks, or two Latinos).

Design and procedure

The study was conducted in four classes with between three and nine groups working simultaneously. All groups within a class were in the same condition: perceived similarity or perceived dissimilarity. Prior to the start of the first class of the semester, participants completed a questionnaire that contained basic demographic questions and six high-self-revealing WYR dilemmas (see Appendix B ). One week later, they were put into groups and were told that their goal was to correctly build a Legoperson puzzle using Lego™ blocks. They were told they would have 30 min to plan as a group, and 30 min to build the puzzle. To facilitate coordination during the building phase, it is beneficial for group members to communicate about integrating the various component parts of the Legoperson during the planning phase ( Heath & Staudenmeyer, 2000 ). Prior to the start of the planning period, groups were told that they were formed on the basis of having similar responses (in the similarity condition) or dissimilar responses to the WYR dilemmas (in the dissimilarity condition). In fact, we did not form groups on the basis of their WYR responses. Group members received no additional information about the WYR questions or about their group members.

The Legoperson exercise required groups to build an exact replica of a figure (the model) resembling a person out of 49 blocks of various sizes. The instructor provided each group with a set of materials, including an instruction sheet and a bag of 49 blocks. On the instruction sheet, students were informed that the exercise was “designed to provide a simulated experience of trying to maximize the efficiency and effectiveness of a work team.”

Groups of students were located at workstations throughout the classroom, and the model was located at the front of the room. Each group was assigned an observer, who monitored and enforced the rules of the exercise, including making sure the group members did not discuss their WYR responses. When groups believed they had completed an exact replica, they brought their Legoperson to the front of the room to be checked against the model by a judge. If the Legoperson was not perfect, the judge told the group that it was incorrect without telling them any more information about the defect(s). Groups continued to bring the model to the judge until it was built correctly. At the end of the study, participants reported on each group member’s task contributions.

After completing measures of task contribution, participants were probed for suspicion. None reported being suspicious that their groups were not actually formed on the basis of similarity.

Performance

We measured performance as the number of times the group’s Legoperson was rejected by the judge before they submitted a perfect replica (i.e., number of attempts minus one). Scores ranged from 0 to 4.

Postinteraction ratings

Group members were randomly assigned to wear a nametag with the letter A, B, C, D, or E and rated each group member on the extent to which they “contributed to,” “were focused on,” and “helped the team complete” the task. These three items formed the composite measure of perceived task contribution (α = .85). All items were measured on a 1 ( strongly disagree ) to 7 ( strongly agree ) scale.

We used different analysis strategies for the judgments of the group as a whole and for judgments of individual team members. For ratings of each group member’s task contribution, which were made at the level of the dyad (e.g., Person A rated Person B, and Person B rated Person A), we examined the main effect of experimental condition and estimated the random effects of perceiver, target, dyad, and group (i.e., a social relations model analysis; see Livi, Kenny, Albright, & Pierro, 2008 , for a full discussion of the strategy using the MIXED procedure). We also adjusted for gender composition of the group (see West et al., 2012 ), which did not have a significant effect on performance.

Consistent with Experiment 2, for ratings of individual group members, we considered actual similarity in WYR responses between the perceiver and target. We also created a group-level index of actual similarity by averaging the similarity scores of all of the dyads in the group. As we expected, actual similarity at the dyadic and group levels was not a significant predictor of any of our dependent measures ( p s ranging from .166 to .797). To simplify our models, we omit actual similarity from all models that we report.

Perceived task contribution

Consistent with our hypothesis, participants in teams in the similarity condition perceived that their team members made greater contributions to the Legoperson task ( M = 6.47, SD = 0.58) than did those in the dissimilarity condition ( M = 6.28, SD = 0.69), t (287.57) = 2.48, p = .018. 4 To test whether minorities and Whites were perceived as contributing equally to the task, we compared whether evaluations of task contributions varied as a function of target race (White vs. minority). We also included effects of perceiver race (i.e., Did Whites and minorities perceive team members’ contributions differently?) and the interaction between perceiver and target race (i.e., Did team members perceive other members of the same race differently than those of a different race?). There were no effects of perceiver’s race, target’s race, or their interactions with similarity condition ( p s > .530). Thus, the manipulation equally affected the extent to which Whites and minorities were perceived as contributing to the task.

Team performance

The number of rejections for Legoperson replicas was normally distributed, so we analyzed these team-level performance data using linear regression (treating group as the unit of analysis). Similar to the results for perceived task contribution, we found a main effect of similarity condition, t (1) = 2.16, p = .043. Groups in the similarity condition had fewer rejections ( M = 1.10, SD = 1.37) than did groups in the dissimilarity condition ( M = 2.33, SD = 1.30; B = 1.23, SE = .57).

Mediation analyses

Recall that we found no difference between Whites and minorities in how much they were perceived to have contributed to the task, demonstrating that one demographic group likely did not drive performance effects. We next examined whether the effect of similarity on performance was mediated by individual team members’ perceived task contributions. To do so, we created a team-level average of ratings of individuals’ task contribution and included it in a model predicting team performance. The effect of perceived task contribution on performance was significant ( B = −.448), t (1) = −2.25, p = .036, and the effect of similarity was no longer significant ( p = .224). The bootstrap confidence intervals did not include zero (CI upper bound = .040; lower bound = .382; significance of indirect effect, p = .022), indicating that perceived task contribution mediated the effect of similarity on team performance. 5 This test of mediation provides evidence that the manipulation improved performance through the contributions of all team members.

In Experiment 3, we found that in a setting involving racially diverse teams, manipulating perceived similarity both improved team members’ perceptions of their teammates’ contribution to a task and improved teams’ performance. These findings move beyond Experiments 1 and 2 by demonstrating that the effects of our manipulation generalize from the subjective experience of dyadic cross-race interactions to objective performance in diverse groups wherein group members needed to communicate effectively with one another to perform well. In the context of these diverse groups, we found that Whites and minorities were perceived as contributing equally to the task, suggesting that the manipulation had similar benefits for Whites and minorities. Against the backdrop of previous research on groups and teams, our results are a rare example of a simple manipulation that can improve outcomes for diverse groups without Whites benefiting more than minorities, or vice versa.

General Discussion

Over the past several decades, there has been considerable interest in improving race relations. Some approaches to this issue have focused on how similarities can be highlighted to overcome barriers to successful cross-race interactions ( Gaertner & Dovidio, 2000 ; Mallett et al., 2008 ; Pinel & Long, 2012 ); however, to our knowledge, none has been shown to improve these interactions on as many different psychological and behavioral outcomes or in as many different interaction settings as the one we have explored here.

In three distinct experimental contexts, we demonstrated that perceptions of similarity on self-revealing attributes that are peripheral to the interaction improve dyadic- and group-level interracial interactions in a number of ways. We demonstrated that our approach can reduce feelings of anxiety with respect to one’s partner (Experiments 1 and 2) and increase interest in sustaining contact with one’s partner (Experiment 2). We also move beyond the outcomes of interest in most prior research concerned with improving interracial interactions, namely, decreasing negative emotion and increasing liking. Specifically, in Experiment 2, perceived similarity improved accuracy at inferring one’s partner’s relationship intentions, and in Experiment 3, perceived similarity benefited groups on an objective measure of task performance. The generalizability of our approach across dyadic and group contexts is important given that there is very little dialogue between those who study interracial dyads and those who study racially diverse groups, despite the fact that interpersonal issues that arise in interracial dyads and diverse groups are similar ( Sommers, Warp, & Mahoney, 2008 ).

We have argued that the attributes of similarity are most effective at overcoming issues in cross-race interactions when they have two basic characteristics. First, the attributes must be perceived as self-revealing . Building on research on close relationships, we proposed that the attributes must communicate something important about the self, and we demonstrated empirically that when they do not, the benefits of similarity are not realized by individuals at the initial stages of relationship formation. We argued that in the initial stages of interracial interactions, perceiving similarity across self-revealing attributes can “set the stage” for a positive encounter by creating a sense of psychological closeness. Indeed, we found that our manipulation facilitated positive encounters within a relatively short time frame, and in Experiments 1 and 2, led to more positive expectancies leading up to encounters.

Second, the basis of similarity should be peripheral to the goals of the interaction; the attributes in and of themselves ought to have no bearing, and ought to be perceived to have no bearing, on success within any given interaction context. Because peripheral attributes are tied neither to individuals’ own goals nor to the goals of the dyad or group, similarity on these attributes can yield benefits across relational contexts. Moreover, because peripheral attributes are unlikely to be revealed through behavior, the perception of similarity on those attributes can easily be manipulated without interference from actual similarity. Had the attributes of similarity been central to participants’ interactions (e.g., personal values in Experiment 2, or conscientiousness in Experiment 3), participants might have been able to detect the level of actual similarity to their partners, which we would have expected to exert a greater influence on our psychological and behavioral outcomes (or would have revealed our manipulation to be disingenuous). These arguments emphasize the utility of using attributes of similarity that are peripheral to the goals of the interaction, but future research could systematically test whether the effectiveness of manipulating similarity hinges on the attributes being peripheral.

The Effects and Functions of Perceived Similarity

In Experiment 2, we found that Whites and minorities were less anxious and more empathically accurate when they perceived similarity to their cross-race partners. It is possible that perceiving similarity improves accuracy for Whites and minorities because it facilitates anxiety reduction. Anxiety during cross-race encounters can inhibit information processing, making it difficult for perceivers to attend to their partners’ behaviors during interactions ( Richeson & Shelton, 2003 ). Moreover, the behaviors that communicate anxiety are interpreted as signs of disinterest ( Dovidio et al., 2002 ), but anxious individuals are not necessarily disinterested ones ( West, 2011 ; West, Dovidio, & Pearson, in press ). These connections between anxiety and perceptions of relationship interest are reflected in a growing body of research suggesting that anxiety can interfere with both the reading of one’s partner’s relationship intentions and the expression of these intentions, particularly within cross-race interactions. For example, West et al. (in press) found that within cross-race roommate relationships, the more Whites and minorities perceived their roommates as anxious, the more they systematically underestimated roommates’ interest in the relationship. Future research could test how perceiving similarity uniquely affects the target’s and perceiver’s anxiety, and the extent to which their levels of anxiety, in turn, influence accuracy.

Our research also suggests that not all participants benefit from perceiving similarity. In same-race interactions, too much similarity may be problematic in the early stages of relationship formation. We found that participants in Experiment 2 were less accurate (in terms of direct accuracy, which adjusts for accuracy achieved via correctly assumed similarity) in reading one another’s relationship intentions when they perceived relatively high levels of similarity in same-race encounters. This finding is consistent with West and Kenny (2011) , who found that among newly acquainted college roommates, as closeness increased, direct accuracy decreased (see also Kenny, 1994 ). Perhaps when participants were of the same race, perceiving additional similarity on self-revealing attributes decreased their motivation to attend carefully to their partner’s behavior for clues about their actual level of interest in the relationship because they instead assumed that they could accurately gauge their partner’s interest, which would indeed result in weaker direct accuracy (vs. total accuracy, which includes accuracy achieved indirectly via correctly assumed similarity). Future research could explore the conditions under which perceived similarity leads to a false sense of understanding one’s partner in newly forming relationships.

In all of our experiments, when partners learned they were similar, they simultaneously learned that their partners were also made aware of their similarity. As such, our manipulation of perceived similarity might have reduced participants’ concerns about what their partner(s) thought of them, which could have paved the way for a more positive encounter. Indeed, these meta-concerns can detrimentally affect cross-race encounters, particularly for Whites (see Vorauer 2006 ), and future research could benefit from directly measuring the effects of perceiving similarity on meta-concerns. It would also be worth testing whether perceived similarity across self-revealing attributes improves relational outcomes when individuals are apprised that they are similar to their partners but are also informed that their partners are not aware of their similarities.

Divergent Effects for Whites and Minorities Within Cross-Race Interactions

In comparison to research focused on Whites’ experiences, surprisingly few studies have explored minorities’ experiences within interracial encounters ( Toosi et al., 2012 ). In our experiments, we directly compared the extent to which perceived similarity benefited Whites and minorities within interracial interactions and found that Whites were more consistent benefactors of perceiving similarity to their partners. Minorities were not adversely affected by perceived similarity. Indeed, on some measures minorities benefited appreciably, but their responses to the manipulation were more uneven. In considering how the pattern of effects for Whites and minorities varied across the three experiments, we think it is important to note that the effects of perceived similarity on Whites compared with minorities were more similar from Experiments 1 to 2 and from Experiments 2 to 3. We discuss two possibilities for the divergent effects between Whites and minorities as well as for the reduction in the extent of the divergence across our experiments.

First, in our interpretation of the results of Experiment 1, we suggested that Whites may be more susceptible to the beneficial influence of perceived similarity in contexts in which liking concerns are most salient. It is possible that concerns with appearing likable decreased and became more balanced with competence concerns from Experiment 1 to Experiment 3. In Experiment 1, participants were trying to present themselves as good friends. By Experiment 3, participants were trying to present themselves as good teammates—likable, but also competent at the task.

In Experiment 1, using a task in which participants faced strong demands to appear likable to an ostensible partner, Whites’ greater emphasis on appearing likable to their different-race partner might explain why perceiving similarity was only effective at reducing anticipatory anxiety for Whites and not for minorities. In Experiment 2, we sought to reduce the salience of likability concerns in the interaction and indeed found that minorities who perceived more similarity experienced less anxiety, much like their White counterparts. In that experiment, there was a lingering difference between minorities and Whites: Minorities who perceived more similarity to their partner did not report greater interest in sustained contact, whereas Whites who perceived more similarity did express more interest in contact. By Experiment 3, these concerns were more balanced, as participants needed to be competent to complete the task, especially because teammates were under time pressure to complete the project. In this study, Whites and minorities appeared to benefit equally from perceiving similarity to their group members in terms of their perceived contributions to the task. Although the salience of liking concerns is one interpretation for why Whites and minorities responded more similarly to the manipulation of perceived similarity in some contexts than in others, future research could systematically vary the extent to which those concerns are salient in combination with manipulating perceived similarity.

Second, it is possible that features of our particular White and minority samples contributed to the patterns of effects for Whites and minorities. It might be the case that minorities had more cross-race close friendships than Whites prior to participating in our studies—a difference that would be particularly important within Experiments 1 and 2 in which friendship formation was the goal. Individuals who have more outgroup friends tend to be more trusting of outgroup members ( Tropp, 2008 ) and perceive stronger overlap between the outgroup and the self ( Page-Gould, Mendoza-Denton, Alegre, & Siy, 2010 ). Thus, a manipulation of perceived similarity in the nascent stages of a cross-race relationship might have less influence on outcomes for individuals who have racially diverse friendship networks. We were able to explore this possibility with the data in Experiments 1 and 2.

For Experiments 1 and 2, we examined prestudy data in which participants reported the extent to which their close friends were from their own versus other racial groups (i.e., the extent of homophily in their close friendship networks) (1 = entirely people of my racial group; 5 = entirely people of another racial group [not my racial group]; for a similar measure, see Page-Gould et al., 2008 ). In both experiments, Whites reported significantly more homophilic friendship networks (Experiment 1 M = 2.45, SD = .69; Experiment 2 M = 2.51, SD = .772) than minorities (Experiment 1 M = 2.74, SD = .95; Experiment 2 M = 2.92, SD = .796; p s < .041). Although we had insufficient statistical power to test whether homophily moderated the effects observed for Whites and minorities in these two studies, we think this preexisting difference is important to note and could have contributed to minorities experiencing consistently low levels of anxiety across conditions in Experiment 1 and relatively high levels of interest in sustained contact in Experiment 2. Future research could test this hypothesis more systematically by examining the moderating role of friendship network homophily on interventions designed to improve interpersonal processes within interracial contact contexts.

Additional Directions for Future Research

In addition to the future research directions already suggested, which could help illuminate the roles that similarity plays in new relationships and help establish the necessary and sufficient conditions to improve cross-race interactions, we see a number of other interesting avenues for researchers to build on the results of our studies.

Our experimental designs focused on interaction processes exclusively during the early moments of relationship development, so it remains to be seen whether the effects of our similarity manipulation are sustained over time. Given that initial interaction processes often shape later intergroup relational processes (e.g., Shook & Fazio, 2008 ; West, Shelton, & Trail, 2009 ), we suspect that improving initial interracial encounters sets the stage for more positive interactions in the long term. Moreover, within close relationships, perceptions of similarity from the onset can help protect the relationship from eventual dissolution if and when important bases of dissimilarity are discovered. It would be interesting to examine whether within cross-race interactions, perceiving similarity on self-revealing attributes from the onset also protects relationships when partners uncover sources of dissimilarity.

We also note that consistent with contemporary research on interracial interactions ( Mallett et al., 2008 ; Toosi et al., 2012 ), we did not find any evidence of overall differences between cross-race and same-race interactions (i.e., there were no main effects of the racial composition of the dyad). Differences emerged only under conditions of manipulated dissimilarity and similarity. It may be the case that in interpersonal contexts designed to be fairly positive experiences (e.g., Mallett et al., 2008 ), differences between same-race and cross-race interactions are quite small and variable enough for a manipulation of perceived similarity to benefit partners. It would be interesting to extend the present research to contexts in which partners are more likely to experience negative affect, such as emotional conflicts.

Our studies used samples drawn from New York City, a multicultural city in which people have more cross-race encounters than people in most other locations. Although many New Yorkers (both Whites and minorities) experience some discomfort in their daily interracial interactions, future research should test whether our findings replicate for populations that have less prior, and potentially less positive, interracial contact, outside of New York City. We would expect that our manipulation of perceived self-revealing similarity would be more effective among individuals with less positive interracial contact experience.

By pairing previously isolated research traditions on intergroup relations and close relationships, we have developed a method that improves cross-race interactions in two different interaction contexts— getting-acquainted dyadic interactions and small task groups. Our findings suggest that perceiving similarity with one’s racial outgroup partners may operate in much the same way as does psychological interdependence in close relationships. This research is an important step toward developing simple interventions that can improve intergroup relations and demonstrates the power of perception in altering processes that often go awry in cross-race interactions. Individuals need not actually be similar to their partners; they just need to believe they are, and this belief can promote anxiety reduction, increased interest in sustained contact, accuracy in interpersonal perception, and behavioral coordination in racially diverse contexts.

Would-You-Rather Items: Low Self-Revealing

Would-you-rather items: high self-revealing.

Note . Items 2, 3, 8, 9, 10, 11, and 12 were used for Experiment 1. For Experiment 2, Items 1–6 were used for Would-You-Rather (WYR) Version 1, and Items 7–12 were used for WYR Version 2. For Experiment 3, Items 3, 4, 8, 9, 10, and 11 were used.

1 Participants rated the high-self-revealing dilemmas as more revealing about their personalities ( M = 4.99, SD = 1.20) than the low-self-revealing dilemmas ( M = 2.65, SD = 1.35), t (137) = −10.02, p < .01. Participants were equally comfortable disclosing their preferences for the low-self-revealing dilemmas ( M = 6.32, SD = 1.62) and to the high-self-revealing dilemmas ( M = 5.88, SD = 1.11), t (137) = 1.66, p = .10.

2 We note that because we had two versions of WYRs, we used five additional questions (i.e., 12 in total; see Appendix B ), which were also pilot-tested prior to Experiment 1. These five items were rated as equally self-revealing by participants as the seven items used in Experiments 1 and 2 ( p = .68). There were no significant differences between the two versions of the WYR dilemmas on how comfortable perceivers felt revealing their answers to their partners ( p = .312), or the extent to which they felt their answers were revealing about their personality ( p = .825). Analyses predicting the dependent measures revealed no significant interactions between versions of the WYR dilemmas and race of the participant or racial composition of the dyad ( p s > .318).

3 For anticipatory anxiety, when Whites and minorities were relatively high in similarity (1 SD above the mean), they were less anxious when WYR responses were revealed than when they were not, although this difference was not significant, t (42) = −1.46, p = .15. In contrast, when they were relatively low in similarity (1 SD below the mean), they were more anxious when responses were revealed than when they were not, t (42) = 2.03, p = .049. For interaction anxiety, when Whites and minorities were relatively high in similarity (1 SD above the mean), they were less anxious when responses were revealed than when they were not, t (42) = −2.29, p = .027. In contrast, when they were relatively low in similarity (1 SD below the mean), they were more anxious when responses were revealed than when they were not, t (42) = 2.88, p = .006. For interest in sustained contact, when Whites and minorities were relatively high in similarity (1 SD above the mean), they were more interested in contact when responses were revealed than when they were not, although not significantly so, t (85.16) = 1.52, p = .13. In contrast, when they were relatively low in similarity (1 SD below the mean), they were less interested in contact when responses were revealed than when they were not, t (84.76) = −3.22, p = .002.

4 We found that there was consensus in team members’ perceptions of each other’s task contributions, as indicated by significant target variance in the social relations model analysis (absolute target variance = .060, SE = .012; p = .001).

5 This test of mediation does not imply causation between experimental condition (the predictor) and perceived task contributions (the mediator) but rather demonstrates that the effect of the manipulation on performance at the group level can be explained in part by participants’ subjective perceptions of one another’s contributions to the task.

Contributor Information

Tessa V. West, Department of Psychology, New York University.

Joe C. Magee, Stern School of Business, New York University.

Sarah H. Gordon, School of Public Health, Harvard University.

Lindy Gullett, Department of Psychology, New York University.

• Aiken L, West S. Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage; 1991. [ Google Scholar ]
• Altman I, Taylor DA. Social penetration: The development of interpersonal relationships. Oxford, England: Holt, Rinehart & Winston; 1973. [ Google Scholar ]
• Anderson C, John OP, Keltner D, Kring AM. Who attains social status? Effects of personality and physical attractiveness in social groups. Journal of Personality and Social Psychology. 2001; 81 :116–132. doi: 10.1037/0022-3514.81.1.116. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Aron A, Melinat E, Aron EN, Vallone R, Bator R. The experimental generation of interpersonal closeness: A procedure and some preliminary findings. Personality and Social Psychology Bulletin. 1997; 23 :363–377. doi: 10.1177/0146167297234003. [ CrossRef ] [ Google Scholar ]
• Bergsieker HB, Shelton JN, Richeson JA. To be liked versus respected: Divergent goals in interracial interactions. Journal of Personality and Social Psychology. 2010; 99 :248–264. doi: 10.1037/a0018474. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Blascovich J, Mendes WB, Hunter SB, Lickel B, Kowai-Bell N. Perceiver threat in social interactions with stigmatized others. Journal of Personality and Social Psychology. 2001; 80 :253–267. doi: 10.1037/0022-3514.80.2.253. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Britt TW, Boniecki KA, Vescio TK, Biernat M, Brown LM. Intergroup anxiety: A person x situation approach. Personality and Social Psychology Bulletin. 1996; 22 :1177–1188. doi: 10.1177/01461672962211008. [ CrossRef ] [ Google Scholar ]
• Brown R, Hewstone M. An integrative theory of intergroup contact. Advances in Experimental Social Psychology. 2005; 37 :255–343. doi: 10.1016/S0065-2601(05)37005-5. [ CrossRef ] [ Google Scholar ]
• Byrne D, Wong TJ. Racial prejudice, interpersonal attraction, and assumed dissimilarity of attitudes. The Journal of Abnormal and Social Psychology. 1962; 65 :246–253. doi: 10.1037/h0047299. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Chen FF, Kenrick DT. Repulsion or attraction? Group membership and assumed attitude similarity. Journal of Personality and Social Psychology. 2002; 83 :111–125. doi: 10.1037/0022-3514.83.1.111. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Condon JW, Crano WD. Inferred evaluation and the relation between attitude similarity and interpersonal attraction. Journal of Personality and Social Psychology. 1988; 54 :789–797. doi: 10.1037/0022-3514.54.5.789. [ CrossRef ] [ Google Scholar ]
• Crisp R, Turner RN. The imagined contact hypothesis. Advances in Experimental Social Psychology. 2012; 46 :125–182. doi: 10.1016/B978-0-12-394281-4.00003-9. [ CrossRef ] [ Google Scholar ]
• Demoulin S. Emotional misunderstandings in intergroup negotiations. In: Demoulin S, Leyens JPh, Dovidio JF, editors. Intergroup misunderstandings: Impact of divergent social realities psychology. New York, NY: Psychology Press; 2008. pp. 331–334. [ Google Scholar ]
• Dovidio JF. On the nature of contemporary prejudice: The third wave. Journal of Social Issues. 2001; 57 :829–849. doi: 10.1111/0022-4537.00244. [ CrossRef ] [ Google Scholar ]
• Dovidio JF, Gaertner SL, Kawakami K, Hodson G. Why can’t we just get along? Interpersonal biases and interracial distrust. Cultural Diversity and Ethnic Minority Psychology. 2002; 8 :88–102. doi: 10.1037/1099-9809.8.2.88. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dovidio JF, Pearson A, Smith-McLallen A, Kawakami K. From the interpersonal to the intergroup: The role of nonverbal behavior in interracial interactions on intergroup relations; Paper presented at the annual meeting of the Society for Experimental Social Psychology; San Diego, CA. 2005. Jul, [ Google Scholar ]
• Fiske ST, Lin M, Neuberg SL. The continuum model: Ten years later. In: Trope Y, Chaiken S, editors. Dual-process theories in social psychology. New York, NY: Guilford Press; 1999. pp. 231–254. [ Google Scholar ]
• Fletcher GJO, Kerr PSG. Through the eyes of love: Reality and illusion in intimate relationships. Psychological Bulletin. 2010; 136 :627–658. doi: 10.1037/a0019792. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Frey FE, Tropp LR. Being seen as individuals versus as group members: Extending research on metaperception to intergroup contexts. Personality and Social Psychology Review. 2006; 10 :265–280. doi: 10.1207/s15327957pspr1003_5. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Gaertner SL, Dovidio JF. The aversive form of racism. In: Dovidio JF, Gaertner SL, editors. Prejudice, discrimination, and racism. Orlando, FL: Academic Press; 1986. pp. 61–89. [ Google Scholar ]
• Gaertner SL, Dovidio JF. Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology Press; 2000. [ CrossRef ] [ Google Scholar ]
• Gagné FM, Lydon JE. Bias and accuracy in close relationships: An integrative review. Personality and Social Psychology Review. 2004; 8 :322–338. doi: 10.1207/s15327957pspr0804_1. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Heath C, Staudenmeyer N. Coordination neglect: How lay theories of organizing complicate coordination in organizations. In: Sutton RI, Straw B, editors. Research in organization behaviour. Greenwich, CT: JAI Press; 2000. pp. 153–191. [ CrossRef ] [ Google Scholar ]
• Holmes JG, Rempel JK. Trust in close relationships. In: Hendrick C, editor. Close relationships. Newbury Park, CA: Sage; 1989. pp. 187–220. [ Google Scholar ]
• Horn R. Zobmondo!! The outrageous book of bizarre choices. New York, NY: Workman; 2001. [ Google Scholar ]
• Houlette MA, Gaertner SL, Johnson KM, Banker BS, Riek BM, Dovidio JF. Developing a more inclusive social identity: An elementary school intervention. Journal of Social Issues. 2004; 60 :35–55. doi: 10.1111/j.0022-4537.2004.00098.x. [ CrossRef ] [ Google Scholar ]
• Hoyle RH. Interpersonal attraction in the absence of explicit attitudinal information. Social Cognition. 1993; 11 :309–320. doi: 10.1521/soco.1993.11.3.309. [ CrossRef ] [ Google Scholar ]
• Islam MR, Hewstone M. Dimensions of contact as predictors of intergroup anxiety, perceived out-group variability, and out-group attitude: An integrative model. Personality and Social Psychology Bulletin. 1993; 19 :700–710. doi: 10.1177/0146167293196005. [ CrossRef ] [ Google Scholar ]
• Kenny DA. Interpersonal perception: A social relations analysis. New York, NY: Guilford Press; 1994. [ PubMed ] [ Google Scholar ]
• Kenny DA, Acitelli LK. Accuracy and bias in the perception of the partner in a close relationship. Journal of Personality and Social Psychology. 2001; 80 :439–448. doi: 10.1037/0022-3514.80.3.439. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kenny DA, Kashy DA, Cook WL. Dyadic data analysis. New York, NY: Guilford Press; 2006. [ Google Scholar ]
• Kilpatrick SD, Bissonnette VL, Rusbult CE. Empathic accuracy and accommodative behavior among newly married couples. Personal Relationships. 2002; 9 :369–393. doi: 10.1111/1475-6811.09402. [ CrossRef ] [ Google Scholar ]
• Kirschbaum C, Pirke KM, Hellhammer DH. The ‘Trier Social Stress Test’—A tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology. 1993; 28 :76–81. doi: 10.1159/000119004. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Klein KJ, Lim B-C, Saltz JL, Mayer DM. How do they get there? An examination of the antecedents of centrality in team networks. Academy of Management Journal. 2004; 47 :952–963. [ Google Scholar ]
• LaPrelle J, Hoyle RH, Insko CA, Bernthal P. Interpersonal attraction and descriptions of the traits of others: Ideal similarity, self similarity, and liking. Journal of Research in Personality. 1990; 24 :216–240. doi: 10.1016/0092-6566(90)90018-2. [ CrossRef ] [ Google Scholar ]
• Lemay EP, Clark MS. How the head liberates the heart: Projection of communal responsiveness guides relationship promotion. Journal of Personality and Social Psychology. 2008; 94 :647–671. doi: 10.1037/0022-3514.94.4.647. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Levin S, Van Laar C, Sidanius J. The effects of ingroup and outgroup friendships on ethnic attitudes in college: A longitudinal study. Group Processes & Intergroup Relations. 2003; 6 :76–92. doi: 10.1177/1368430203006001013. [ CrossRef ] [ Google Scholar ]
• Linden-Andersen S, Markiewicz D, Doyle AB. Perceived similarity among adolescent friends: The role of reciprocity, friendship quality, and gender. Journal of Early Adolescence. 2009; 29 :617–637. doi: 10.1177/0272431608324372. [ CrossRef ] [ Google Scholar ]
• Livi S, Kenny DA, Albright L, Pierro A. A social relations analysis of leadership. The Leadership Quarterly. 2008; 19 :235–248. doi: 10.1016/j.leaqua.2008.01.003. [ CrossRef ] [ Google Scholar ]
• Luo S, Snider AG. Accuracy and bias in newlyweds’ perceptions of each other: Not mutually exclusive but mutually beneficial. Psychological Science. 2009; 20 :1332–1339. doi: 10.1111/j.1467-9280.2009.02449.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Lutz-Zois CJ, Bradley AC, Mihalik JL, Moorman-Eavers ER. Perceived similarity and relationship success among dating couples: An idiographic approach. Journal of Social and Personal Relationships. 2006; 23 :865–880. doi: 10.1177/0265407506068267. [ CrossRef ] [ Google Scholar ]
• Mallett RK, Wilson TD. Increasing positive intergroup contact. Journal of Experimental Social Psychology. 2010; 46 :382–387. doi: 10.1016/j.jesp.2009.11.006. [ CrossRef ] [ Google Scholar ]
• Mallett RK, Wilson TD, Gilbert DT. Expect the unexpected: Failure to anticipate similarities when predicting the quality of an intergroup interaction. Journal of Personality and Social Psychology. 2008; 94 :265–277. doi: 10.1037/0022-3514.94.2.94.2.265. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Mendoza-Denton R, Page-Gould E, Pietrzak J. Mechanisms for coping with status-based rejection expectations. In: Levin S, Van Laar C, editors. Stigma and group inequality: Social psychological perspectives. Mahwah, NJ: Erlbaum; 2006. pp. 151–170. [ Google Scholar ]
• Milliken FJ, Martins LL. Searching for common threads: Understanding the multiple effects of diversity in organizational groups. Academy of Management Review. 1996; 21 :402–433. [ Google Scholar ]
• Montoya RM, Horton RS, Kirchner J. Is actual similarity necessary for attraction? A meta-analysis of actual and perceived similarity. Journal of Social and Personal Relationships. 2008; 25 :889–922. doi: 10.1177/0265407508096700. [ CrossRef ] [ Google Scholar ]
• Morry MM, Kito M, Martens L, Marchylo A, Stevens K. The attraction-similarity model: Friendship and dating studies; Poster presented at the 2005 Society for Personality and Social Psychology Conference; New Orleans, LA. 2005. [ Google Scholar ]
• Mullen B, Dovidio JF, Johnson C, Copper C. In-group-out-group differences in social projection. Journal of Experimental Social Psychology. 1992; 28 :422–440. doi: 10.1016/0022-1031(92)90040-Q. [ CrossRef ] [ Google Scholar ]
• Murray SL. The quest for conviction: Motivated cognition in romantic relationships. Psychological Inquiry. 1999; 10 :23–34. doi: 10.1207/s15327965pli1001_3. [ CrossRef ] [ Google Scholar ]
• Murray SL, Holmes JG, Bellavia G, Griffin DW, Dolderman D. Kindred spirits? The benefits of egocentrism in close relationships. Journal of Personality and Social Psychology. 2002; 82 :563–581. doi: 10.1037/0022-3514.82.4.563. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Neff LA, Karney BR. To know you is to love you: The implications of global adoration and specific accuracy for marital relationships. Journal of Personality and Social Psychology. 2005; 88 :480–497. doi: 10.1037/0022-3514.88.3.480. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Nier JA, Gaertner SL, Dovidio JF, Banker BS, Ward CM, Rust MC. Changing interracial evaluations and behavior: The effects of a common group identity. Group Processes & Intergroup Relations. 2001; 4 :299–316. doi: 10.1177/1368430201004004001. [ CrossRef ] [ Google Scholar ]
• Norton MI, Frost JH, Ariely D. Less is more: The lure of ambiguity, or why familiarity breeds contempt. Journal of Personality and Social Psychology. 2007; 92 :97–105. doi: 10.1037/0022-3514.92.1.97. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Overall NC, Hammond MD. Biased and accurate depressive symptoms and daily perceptions within intimate relationships. Personality and Social Psychology Bulletin. 2013; 39 :636–650. doi: 10.1177/0146167213480188. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Page-Gould E, Mendoza-Denton R, Alegre JM, Siy JO. Understanding the impact of cross-group friendship on interactions with novel outgroup members. Journal of Personality and Social Psychology. 2010; 98 :775–793. doi: 10.1037/a0017880. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Page-Gould E, Mendoza-Denton R, Tropp LR. With a little help from my cross-group friend: Reducing anxiety in intergroup contexts through cross-group friendship. Journal of Personality and Social Psychology. 2008; 95 :1080–1094. doi: 10.1037/0022-3514.95.5.1080. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Paluck EL, Green DP. Prejudice reduction: What works? A review and assessment of research and practice. Annual Review of Psychology. 2009; 60 :339–367. doi: 10.1146/annurev.psych.60.110707.163607. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Pearson AR, West TV, Dovidio JF, Powers S, Buck R, Henning R. The fragility of intergroup relations: Divergent effects of delayed audio-visual feedback in intergroup and intragroup interactions. Psychological Science. 2008; 19 :1272–1279. doi: 10.1111/j.1467-9280.2008.02236.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Penner LA, Gaertner S, Dovidio JF, Hagiwara N, Porcerelli J, Markova T, Albrecht TL. A social psychological approach to improving the outcomes of racially discordant medical interactions. Journal of General Internal Medicine. 2013; 9 :1143–1149. doi: 10.1007/s11606-013-2339-y. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Pettigrew TF. Intergroup contact theory. Annual Review of Psychology. 1998; 49 :65–85. doi: 10.1146/annurev.psych.49.1.65. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Pettigrew TF, Tropp LR. How does intergroup contact reduce prejudice? Meta-analytic tests of three mediators. European Journal of Social Psychology. 2008; 38 :922–934. doi: 10.1002/ejsp.504. [ CrossRef ] [ Google Scholar ]
• Phillips KW. The effects of categorically based expectations on minority influence: The importance of congruence. Personality and Social Psychology Bulletin. 2003; 29 :3–13. doi: 10.1177/0146167202238367. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Pinel EC, Long AE. When I’s meet: Sharing subjective experience with someone from the outgroup. Personality and Social Psychology Bulletin. 2012; 38 :296–307. doi: 10.1177/0146167211433878. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Plant EA. Responses to interracial interactions over time. Personality and Social Psychology Bulletin. 2004; 30 :1458–1471. doi: 10.1177/0146167204264244. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Plant EA, Butz DA. The causes and consequences of an avoidance-focus for interracial interactions. Personality and Social Psychology Bulletin. 2006; 32 :833–846. doi: 10.1177/0146167206287182. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Richeson JA, Shelton JN. When prejudice does not pay effects of interracial contact on executive function. Psychological Science. 2003; 14 :287–290. doi: 10.1111/1467-9280.03437. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Robbins JM, Krueger JI. Social projection to ingroups and outgroups: A review and meta-analysis. Personality and Social Psychology Review. 2005; 9 :32–47. doi: 10.1207/s15327957pspr0901_3. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Rokeach M, Mezei L. Race and shared belief as factors in social choice. Science. 1966; 151 :167–172. doi: 10.1126/science.151.3707.167. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Selfhout M, Denissen J, Branje S, Meeus W. In the eye of the beholder: Perceived, actual, and peer-rated similarity in personality, communication, and friendship intensity during the acquaintanceship process. Journal of Personality and Social Psychology. 2009; 96 :1152–1165. doi: 10.1037/a0014468. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shelton JN, Richeson JA. Intergroup contact and pluralistic ignorance. Journal of Personality and Social Psychology. 2005; 88 :91–107. doi: 10.1037/0022-3514.88.1.91. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shelton JN, Richeson JA, Salvatore J. Expecting to be the target of prejudice: Implications for interethnic interactions. Personality and Social Psychology Bulletin. 2005; 31 :1189–1202. doi: 10.1177/0146167205274894. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shelton JN, Trail TE, West TV, Bergsieker HB. From strangers to friends: The interpersonal process model of intimacy in developing interracial friendships. Journal of Social and Personal Relationships. 2010; 27 :71–90. doi: 10.1177/0265407509346422. [ CrossRef ] [ Google Scholar ]
• Shook NJ, Fazio RH. Interracial roommate relationships: An experimental field test of the contact hypothesis. Psychological Science. 2008; 19 :717–723. doi: 10.1111/j.1467-9280.2008.02147.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Sommers SR, Warp LS, Mahoney CC. Cognitive effects of racial diversity: White individuals’ information processing in heterogeneous groups. Journal of Experimental Social Psychology. 2008; 44 :1129–1136. doi: 10.1016/j.jesp.2008.01.003. [ CrossRef ] [ Google Scholar ]
• Stahl GK, Maznevski ML, Voigt A, Jonsen K. Unraveling the effects of cultural diversity in teams: A meta-analysis of research on multicultural work groups. Journal of International Business Studies. 2010; 41 :690–709. doi: 10.1057/jibs.2009.85. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stein DD. The influence of belief systems on interpersonal preference: A validation study of Rokeach’s theory of prejudice. Psychological Monographs: General and Applied. 1966; 80 :1–29. doi: 10.1037/h0093907. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stein DD, Hardyck JA, Smith MB. Race and belief: An open and shut case. Journal of Personality and Social Psychology. 1965; 1 :281–289. doi: 10.1037/h0021870. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Stephan WG, Stephan CW. Intergroup anxiety. Journal of Social Issues. 1985; 41 :157–175. doi: 10.1111/j.1540-4560.1985.tb01134.x. [ CrossRef ] [ Google Scholar ]
• Stephan WG, Stephan CW. An integrated threat theory of prejudice. In: Oskamp S, editor. Reducing prejudice and discrimination: The Claremont Symposium on Applied Psychology. Mahwah, NJ: Erlbaum; 2000. pp. 23–45. [ Google Scholar ]
• Sunnafrank M. A communication-based perspective on attitude similarity and interpersonal attraction in early acquaintance. Communication Monographs. 1984; 51 :372–380. doi: 10.1080/03637758409390208. [ CrossRef ] [ Google Scholar ]
• Sunnafrank M. Communicative influences on perceived similarity and attraction: An expansion of the interpersonal goals perspective. Western Journal of Communication. 1986; 50 :158–170. doi: 10.1080/10570318609374222. [ CrossRef ] [ Google Scholar ]
• Sunnafrank MJ, Miller GR. The role of initial conversations in determining attraction to similar and dissimilar strangers. Human Communication Research. 1981; 8 :16–25. doi: 10.1111/j.1468-2958.1981.tb00653.x. [ CrossRef ] [ Google Scholar ]
• Sunnafrank M, Ramirez A. At first sight: Persistent relational effects of get-acquainted conversations. Journal of Social and Personal Relationships. 2004; 21 :361–379. doi: 10.1177/0265407504042837. [ CrossRef ] [ Google Scholar ]
• Swart H, Hewstone M, Christ O, Voci A. Affective mediators of intergroup contact: A three-wave longitudinal study in South Africa. Journal of Personality and Social Psychology. 2011; 101 :1221–1238. doi: 10.1037/a0024450. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Tidwell ND, Eastwick PW, Finkel EJ. Perceived, not actual, similarity predicts initial attraction in a live romantic context: Evidence from the speed-dating paradigm. Personal Relationships. 2013; 20 :199–215. doi: 10.1111/j.1475-6811.2012.01405.x. [ CrossRef ] [ Google Scholar ]
• Toosi NR, Babbitt LG, Ambady N, Sommers SR. Dyadic interracial interactions: A meta-analysis. Psychological Bulletin. 2012; 138 :1–27. doi: 10.1037/a0025767. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Trawalter S, Richeson JA. Regulatory focus and executive function after interracial interactions. Journal of Experimental Social Psychology. 2006; 42 :406–412. doi: 10.1016/j.jesp.2005.05.008. [ CrossRef ] [ Google Scholar ]
• Trawalter S, Richeson JA, Shelton JN. Predicting behavior during interracial interactions: A stress and coping approach. Personality and Social Psychology Review. 2009; 13 :243–268. doi: 10.1177/1088868309345850. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Tropp LR. The role of trust in intergroup contact: Its significance and implications for improving relations between groups. In: Wagner U, Tropp LR, Finchilescu G, Tredoux C, editors. Improving intergroup relations: Building on the legacy of Thomas F. Pettigrew. Malden, MA: Blackwell; 2008. pp. 91–106. [ CrossRef ] [ Google Scholar ]
• Van Knippenberg D, Schippers MC. Work group diversity. Annual Review of Psychology. 2007; 58 :515–541. doi: 10.1146/annurev.psych.58.110405.085546. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Vorauer JD. An information search model of evaluative concerns in intergroup interaction. Psychological Review. 2006; 113 :862–886. doi: 10.1037/0033-295X.113.4.862. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Vorauer JD, Main KJ, O’Connell GB. How do individuals expect to be viewed by members of lower status groups? Content and implications of meta-stereotypes. Journal of Personality and Social Psychology. 1998; 75 :917–937. doi: 10.1037/0022-3514.75.4.917. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Vorauer JD, Sakamoto Y. I thought we could be friends, but … Systematic miscommunication and defensive distancing as obstacles to cross-group friendship formation. Psychological Science. 2006; 17 :326–331. doi: 10.1111/j.1467-9280.2006.01706.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West TV. Interpersonal perception in cross-group interactions: Challenges and potential solutions. European Review of Social Psychology. 2011; 22 :364–401. doi: 10.1080/10463283.2011.641328. [ CrossRef ] [ Google Scholar ]
• West TV, Dovidio JF. Intergroup contact across time: Beyond initial contact. In: Hodson G, Hewstone M, editors. Advances in intergroup contact. Malden, MA: Wiley-Blackwell; 2013. pp. 152–175. [ Google Scholar ]
• West TV, Dovidio JF, Pearson AR. Accuracy and bias in perceptions of relationship interest for intergroup and intragroup roommates. Social Psychological and Personality Science in press. [ Google Scholar ]
• West TV, Heilman ME, Gullett L, Moss-Racusin CA, Magee JC. Building blocks of bias: Gender composition predicts male and female group members’ evaluations of each other and the group. Journal of Experimental Social Psychology. 2012; 48 :1209–1212. doi: 10.1016/j.jesp.2012.04.012. [ CrossRef ] [ Google Scholar ]
• West TV, Kenny DA. The truth and bias model of judgment. Psychological Review. 2011; 118 :357–378. doi: 10.1037/a0022936. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• West TV, Pearson AR, Dovidio JF, Shelton JN, Trail TE. Superordinate identity and intergroup roommate friendship development. Journal of Experimental Social Psychology. 2009; 45 :1266–1272. doi: 10.1016/j.jesp.2009.08.002. [ CrossRef ] [ Google Scholar ]
• West TV, Pearson AR, Stern C. Anxiety perseverance in intergroup interaction: When incidental explanations backfire. Manuscript submitted for publication 2014 [ PubMed ] [ Google Scholar ]
• West TV, Shelton JN, Trail TE. Relational anxiety in interracial interactions. Psychological Science. 2009; 20 :289–292. doi: 10.1111/j.1467-9280.2009.02289.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Williams KY, O’Reilly CA. Demography and diversity in organizations: A review of 40 years of research. In: Staw BM, Cummings LL, editors. Research in organizational behavior. Vol. 20. Stamford, CT: JAI; 1998. pp. 77–140. [ Google Scholar ]

• Previous Article
• Next Article

Current Research

Experiment 1 methods, experiment 1 results and discussion, experiment 2, experiment 2 methods, experiment 2 results and discussion, experiment 3, experiment 3 methods, experiment 3 results and discussion, general discussion, conclusions, contributions, acknowledgements, funding information, competing interests, supplemental materials, data accessibility statement, appendix 1: trust items used in experiment 3, the influence of similarity and mimicry on decisions to trust.

• Split-Screen
• Article contents
• Figures & tables
• Supplementary Data
• Peer Review
• Open the PDF for in another window
• Guest Access
• Get Permissions
• Cite Icon Cite
• Search Site

Alexa S. Clerke , Erin A. Heerey; The Influence of Similarity and Mimicry on Decisions to Trust. Collabra: Psychology 4 January 2021; 7 (1): 23441. doi: https://doi.org/10.1525/collabra.23441

Download citation file:

• Ris (Zotero)
• Reference Manager

Research on trust development has generally focused on how similarities between people influence trust allocation. However, similarity in interests and beliefs, which underpins trust development and may be critical to relationship success, is seldom apparent upon initial interaction and thus may not be a primary predictor of initial trust decisions. Here we ask how mimicry, a visible social cue, affects trust decisions alongside similarity. We used a “chat-room” style task to independently manipulate the degree to which participants were similar to a set of avatars and the degree to which those avatars displayed mimicry. We then assessed trust decisions in both financial and social domains. Our results show that together with similarity, mimicry is an important independent predictor of trust decisions. This work has implications for understanding how and when trust is allocated, as well how to facilitate successful interactions.

Decisions regarding whom to trust are critical elements of the social environment. As with decisions in the cognitive domain, people rely on heuristics to make these decisions quickly, often with little information about the true trustworthiness of their social partners (Metzger & Flanagin, 2013) . For example, people may use the degree to which they perceive themselves to be similar to a person when making trust-based decisions because people generally attend to similarities between themselves and others (Wood, 1996) .

Researchers have examined the effects of similarity across a variety of contexts, such as negotiation outcomes (Wilson et al., 2016) and romantic partner selection (Tidwell et al., 2013) . Findings from this work broadly suggest that as similarity between interaction partners increases, so does liking, cooperation, and trust (Fischer, 2009; Lui et al., 2006) , and that high levels of similarity positively influence interpersonal attraction and social perceptions (Bagues & Perez-Villadoniga, 2013; Byrne & Griffitt, 1973; Jamieson et al., 1987; Morry, 2007; Newcomb, 1963) . For example, evidence shows that in an economic trust game, people who interact with a person of another race return less money than people who interact with a person of the same race (Glaeser et al., 2000) . Thus, physical or appearance-related similarity may increase trust behaviours (e.g., DeBruine, 2002, 2005) . Likewise, research indicates that people with similar interests are more likely to trust and cooperate with one another in the personal domain (Ziegler & Golbeck, 2007) , and in the corporate realm, that corporations with similar business models are less likely to employ coercive negotiation strategies (Lui et al., 2006) .

One possible mechanism underlying the similarity-trust relationship is the notion that interactions with similar others feel more fluent and are easier to process (e.g., Gigerenzer & Gaissmaier, 2011; Whittlesea & Leboe, 2000 ). Specifically, the presence of similarity may make it easier to trust someone because similar attitudes and beliefs are more easily accessible. People use their own attitudes and behaviours to interpret those of others (Gordon, 1992) , leading to easier recall of similar others’ attitudes and more accurate predictions of future behaviour (Thornton et al., 2019) . For example, neuroimaging evidence suggests that people use the self as a reference when inferring others’ states and traits and may attribute their own perceptions of another’s trustworthiness to shared similarity (Jenkins et al., 2008) , thereby enhancing perceived trust (Krueger, 1998; Taylor & Brown, 1988) . Thus, the ease or fluency with which people can interpret others’ behaviour may serve as a trustworthiness cue. Indeed, recent research has indicated that increased fluency, in the context of name perception, leads to greater trust in economic games (Zürn & Topolinksi, 2017) .

Similarity is often operationalized and manipulated using group membership (e.g., Chen & Kenrick, 2002; Montoya & Pittinsky, 2011; Vang & Fox, 2013 ). Specifically, people who are similar to the self on some experimentally salient dimension (e.g., race, sex, team assignment) are classified as in-group members and those who are dissimilar on that dimension become the out-group (e.g., Appiah et al., 2013 ). Evidence from this work largely shows that people are much more likely to like, cooperate with, and trust in-group, relative to out- group members, partially because they are more similar with respect to the experimental context (Balliet et al., 2014; Greenwald & Pettigrew, 2014) . However, one important consideration is that there are often more differences within groups than there are between groups, which may mean that this is not the best manipulation of similarity. In addition, more trust exchanges occur between in-group members than between out-group members, making this manipulation difficult to generalize to the context of real-world trust exchanges.

Regardless of group membership, people who share interests, attitudes, and personality features are more likely to like each other than people who do not share these attributes (McPherson et al., 2001; Youyou et al., 2017) . However, similarity need not be objectively present in order to achieve these effects. Rather, the mere perception of similarity is sufficient to convey benefits. For example, one study found that greater levels of perceived similarity upon first meeting a new freshman undergraduate roommate led to more trust over time (Whitmore & Dunsmore, 2014) . In addition, evidence suggests that when people perceive greater levels of similarity in others, they show higher levels of cooperation and trustworthy behaviour (DeBruine, 2002) .

Although similarity is an important driver of trust decisions in the lab, these cause-effect relationships may be more complicated in the real world. For instance, one may actively seek evidence of similarity with a person after experiencing trustworthy, fair, or cooperative behaviour, which could enhance perceptions of similarity that stem from minimizing or forgetting differences. Indeed, in a computerized interaction, researchers have found that participants perceive trustworthy players as showing greater appearance similarity to themselves than untrustworthy players (Farmer et al., 2014) .

Interestingly, people might infer similarity from experienced social behaviour, such as the disclosure of similar information (Sprecher et al., 2013) or nonverbal mimicry (van Baaren et al., 2009) . Indeed, behavioural mimicry, defined as the inadvertent imitation of an interaction partner’s nonverbal behaviour or verbal style, predicts increased liking, cooperation, and trust (Duffy & Chartrand, 2015; Fischer et al., 2013; Lakin & Chartrand, 2003; Seibt et al., 2015) . For example, in a study of interpersonal negotiations, individuals who engaged in mimicry were more likely to achieve successful outcomes than those who did not (Maddux et al., 2008) . Research has also indicated that participants like and trust people and avatars, who mimic more than those who do not mimic (Chartrand & Bargh, 1999; Chartrand & Lakin, 2013; Seibt et al., 2015) . Mimicry may therefore be influential in creating rapport and bolstering interpersonal connections (Seibt et al., 2015) .

The term mimicry has been used by researchers to describe not only direct imitation, but also a broader class of reciprocal social behaviours that include behaviours that are tightly coupled in time (“social synchrony”) and behaviours that might be complementary rather than exact replicas of another’s behaviour, such as head nodding during an interlocutor’s explanation (“social reciprocity”; Hale et al., 2019 ). While there are slight variations in these behaviours, all are fast, sometimes millisecond-level, responses within interactions that likely generate perceptions of social fluency by increasing the degree to which a partner’s actions feel predictable (Delaherche et al., 2012; Wheatley et al., 2012) . Specifically, high levels of mimicry likely signal a form of similarity between interaction partners, which makes it easier to anticipate and access the feelings and future behaviour of an interaction partner (Tamir & Thornton, 2018; Thornton & Tamir, 2019) by allowing one to use their own behavioural style as an archetype.

In unmanipulated face-to-face interactions, instances of direct mimicry are a subset of this broader class of reciprocal social behaviours. For example, evidence shows that people commonly mimic their interaction partners’ smiles, show complementary behaviour and language use patterns in face-to-face encounters (Heerey & Crossley, 2013; Ireland et al., 2010) , and that people report greater liking for others who indicate liking for them (Montoya & Horton, 2012) . This form of mimicry is strongly apparent in face-to-face social interactions (Heerey & Crossley, 2013; Heerey & Kring, 2007) , including those that culminate in trust decisions. Furthermore, it may signal a social partner’s trustworthiness by providing information about the reliability of the social environment (Behrens et al., 2008, 2009; Tamir & Thornton, 2018) and may subsequently support the development of trust and cooperation. For instance, when reciprocal behaviours are tightly coupled in time (“social synchrony”), social interactions result in greater levels of rapport, cooperation, and overall perceptions of conversational “smoothness” (e.g., Hale & Hamilton, 2016b; Kirschner & Tomasello, 2012; Valdesolo & DeSteno, 2011; Wiltermuth & Heath, 2009 ). Furthermore, temporal synchronization may spontaneously emerge when participants are asked to work cooperatively rather than competitively on a task (Bernieri et al., 1994) , suggesting that people may treat the presence of temporal synchrony as a signal of cooperation.

Research investigating mimicry in naturalistic interactions demonstrates that when people deviate from expected frequencies of low-level behavioural reciprocity (e.g., nodding, smiling), interactions feel disfluent, awkward, and uncomfortable (Delaherche et al., 2012) . This leads to poor outcomes, including reduced trust and willingness to cooperate with the social partner (Launay et al., 2013) . These outcomes are considerable and understanding their causal underpinnings is important.

Many experiments have examined the effect of mimicry on trust, but the current mimicry literature suffers from several limitations that make the findings difficult to generalize. Chief among these is whether the observed mimicry is actually genuine mimicry. Specifically, mimicry is the automatic and unintentional imitation of another’s behaviour (Lakin et al., 2008) but it is not uncommon for researchers to code an instance of mimicry 10 or more seconds from the initiating behaviour and when the initiating behaviour is no longer observable (e.g., Stel & Vonk, 2010 ). Because interactions are extremely fast-paced and social cues may be fleeting (Yan et al., 2013) , the contingency between the initiating and response behaviours may be weak, or non-existent by the time 10 seconds have elapsed. For example, in unmanipulated interactions, the likelihood of smile reciprocity reaches asymptotic levels by approximately 4 seconds (Heerey & Crossley, 2013) . Furthermore, in such interactions the time lag of mimicry is approximately 600ms between the leader and follower (Hale et al., 2019) , which supports the notion that mimicry is both fast and reactive and that longer time frames may not be appropriate for investigating mimicry. In addition to the overlong time lapse, researchers also frequently instruct participants to mimic one another or employ confederates to provide a more “standardized” social experience. Since mimicry is automatic, unconscious, and unintentional (Seibt et al., 2015) , this may lead to artificial or contrived interactions, which differ on other characteristics besides the presence or absence of mimicry.

Although findings from the mimicry literature have generally been thought of as robust, recent work has questioned the consistency of these findings (Hale & Hamilton, 2016a, 2016b) . More specifically, because most of the research has focused on the mimicker rather than the reciprocal nature of mimicking within social interaction, the majority of findings within this field may not be generalizable to the process of mimicking as a whole. Furthermore, work that has focused on the reciprocal nature of mimicking and its effects on trust, rapport, and affiliation is fraught with failed replication attempts (Hale & Hamilton, 2016b) . One explanation for these inconsistent findings is that mimicry is often manipulated by asking participants to interact with trained confederates rather than other naïve participants (e.g., Chartrand & Bargh, 1999; Maddux et al., 2008; Stel et al., 2011; Van Swol, 2003 ). While using confederates within the realm of social psychological research is common practice, it is likely to sway participant behaviour on a lab-by-lab basis because the training is inherently biased by the Principal Investigator’s beliefs and hypotheses (Kuhlen & Brennan, 2013) . Results from research that employs confederates is further complicated because it is unclear whether a single confederate is able to act consistently across participants and, more generally, whether many confederates are able to act consistently within one study or a set of studies.

One solution to these problems is to use computerized avatars in lieu of confederates, which promotes more experimental control while maintaining mundane realism (e.g., Garau et al., 2005; Hale & Hamilton, 2016a ). Some of the work employing this methodology suggests that even in computer-mediated communications (e.g., social media platforms) reciprocal verbal behaviour can predict liking and group cohesiveness (Gonzales et al., 2010; Niederhoffer & Pennebaker, 2002) . However, this research has also produced mixed findings regarding whether mimicry leads to increased trust and rapport (e.g., Bailenson & Yee, 2005; Hale & Hamilton, 2016a ). Because mimicry, and more broadly, reciprocal social behaviour, is often thought of as “social glue” (Dijksterhuis, 2005; Hale & Hamilton, 2016a; Lakin et al., 2003) and a cornerstone of relationship development, it is critical to replicate these findings to test and understand the boundary conditions for these effects.

The current research adds needed conceptual replications of these effects to the literature to enhance understanding of how mimicry influences trust in light of the limitations of past research. Here, we test the relationship between similarity, mimicry, and trust in a rigorous experimental paradigm and report three examinations of these effects using computerized avatars in lieu of confederates.

Specifically, we ask whether higher levels of both similarity and mimicry shape interpersonal perceptions and subsequent trust-based decision-making. This is an important question because, while similarity between social partners is a well-established predictor of trust (e.g., DeBruine, 2002; Fischer, 2009; Glaeser et al., 2000; Ziegler & Golbeck, 2007 ), research on mimicry, which may be just as important as similarity in shaping trust decisions, has generated mixed results (Hale & Hamilton, 2016a, 2016b) . Because we treat similarity and mimicry as independent variables, we use a virtual social context that allows us to reliably manipulate both of these variables. Although nonverbal mimicry in virtual settings may be weaker than in the face-to-face domain (Hale & Hamilton, 2016b) , this effect may in part reflect limited opportunities for participants to learn about one another and exchange personal information in online paradigms (see Bailenson & Yee, 2005; Hale & Hamilton, 2016a ). The effects of mimicry may therefore be degraded in the absence of such conversation – i.e., for mimicry to confer genuine feelings of familiarity, some additional evidence of similarity may be necessary. Importantly, our manipulation allows the simultaneous exchange of both personal information, through which participants can gain a sense of similarity, as well as mimicry.

In each of these experiments, participants “interacted” with avatars, whom they believed to be other participants, in the context of a chat-room style “getting- acquainted” task. Importantly, our similarity manipulation evolves in the context of information exchange, as in natural interactions, rather than evoking similarity based on mere group membership. We measured decisions to trust in the context of a “centipede” game (Experiment 1; Rosenthal, 1981 ), an investor-trustee game (Experiment 2a/b; Berg et al., 1995 ), and a social decision task (Experiment 3). Across the three experiments, we hypothesized that both similarity and mimicry would be significant predictors of trust. Further, in Experiment 3, we hypothesized that mimicry might drive trust decisions to a greater degree than similarity because it appears to be highly apparent to participants.

Participants . Sixty-nine participants completed this study in exchange for partial course credit and a small monetary bonus, based on their performance in the trust game. Of these 69 participants, we discarded 13 from the analysis due to deception failure (i.e., they did not believe they had played real participants). The final sample therefore included 56 undergraduates (14 male, Mage=18.45, SD=.83). All participants gave informed consent and the University’s Ethics Committee approved all study procedures (likewise for Experiments 2 & 3).

The sample size was determined a priori using power analyses based on previously published effects (e.g., Balliet et al., 2014; Vicaria & Dickens, 2016 ). A G*Power (v3.1) analysis revealed that, assuming a small to moderate effect size for the main effects of similarity and mimicry of η p 2 =.04, 55 participants would achieve 95% power (at α=.05). We therefore over-recruited assuming that approximately 5 to 15 participants’ data would be removed due to deception failure (based on similar work conducted in the lab). Sample sizes for Experiments 2 & 3 were determined in similar fashion using the effect sizes obtained in the present experiment.

Procedures . Participants arrived at the lab in groups of five for a study “about how people get to know one another in an online environment.” In reality, participants interacted with computerized avatars, which allowed precise manipulation of both similarity and mimicry and ensured that the manipulation was identical across participants (Heerey, 2015; Schilbach et al., 2006) .

To begin the task, participants selected one of 16 possible avatar images (8 female and 8 male) to represent them for the duration of the experiment. To help maintain the cover story, the computer told all participants that another player had already selected that avatar (we adopted this procedure based on pilot testing that suggested it enhanced believability of the cover story). They then received the second avatar they selected. Moreover, the participant’s first choice of avatar always appeared in the experiment as a highly similar avatar. The computer selected the other three avatars with which participants interacted based on pre-rated appearance similarity. One of these avatars was always similar in appearance to the participant’s own avatar. The remaining two avatars were lower in appearance similarity to the participant’s own avatar. This is based on the idea that participants choose avatars that look similar to themselves (Schultze, 2014) and likely assume that others will do the same. Importantly, we were not inherently interested in appearance similarity, however pre-testing suggested that this selection algorithm increased the believability of the manipulation.

To manipulate objective similarity, we first asked participants to respond to 20 “getting-acquainted” style multiple-choice questions. Four of these items were designated as “attitude” questions, based on an independent participant sample’s ratings (i.e., “Would you consider yourself a feminist?”, “What would you most like to be someday?”, “Would you tend to see yourself as more liberal or more conservative?”, and “Do you have a religious affiliation?”). The remaining items were related to interests/hobbies/preferences (e.g., “What’s your favourite cuisine?”). Participants saw and responded to the questions in random order.

After participants responded to these questions, they “exchanged” answers with each of the other avatars ( Figure 1a ). They viewed responses of all 20 questions for each avatar individually, in fully randomized order (80 trials total). The computer manipulated similarity based on the participant’s responses to the 20 questions. Two avatars were “high” in similarity. These avatars always matched participants’ responses on the four attitude questions, along with a random set of eight other items. The two avatars that were low in similarity matched on fewer of the participant’s own responses to the 20 questions. These avatars did not match on any of the attitude questions but did match on a random set of four of the interest items.

The mimicry manipulation occurred conjointly with the similarity manipulation. Participants gave like/dislike feedback after viewing each avatar’s response to the similarity items using an emoji-style rating scale ( Figure 1b ) similar to those in some social media applications. After participants indicated their feedback response, they saw a screen that simultaneously displayed both their own feedback and that of the avatar ( Figure 1c ). The avatar’s responses could either mimic the participant’s response (matching) or not mimic (non-matching) the participants’ response. Two avatars (one high in similarity and one low in similarity, randomly assigned) were low-mimicry, and provided matching emojis on only 20% of trials. The remaining two avatars were high-mimicry avatars and provided matching feedback on 80% of trials. Importantly, the computer selected non-mimicked feedback that was either 1-level more positive or 1-level more negative on the emoji scale than participants’ own responses, such that the average discrepancy in the feedback positivity between each avatar and a participant was zero. To enhance the believability of the cover story that participants were interacting with real people, the computer inserted brief, random-length (range: 0ms to 1200ms) delays (e.g., “waiting for data…”) in between the response collection ( Figure 1b ) and result display ( Figure 1c ).

After this interaction phase of the game, participants played a “centipede” economic game (Rosenthal, 1981) to measure trust. In a traditional centipede game, two players take turns passing pools of money, until one of them chooses to defect or until some number of exchanges has occurred. On any given turn, the active player receives two pools of points, one large and one small. The player then chooses to either take the larger of the two pools (giving the smaller to the other player) or to pass both pools to the other player. If a player chooses to pass the pools, then both pools double in size ( Figure 1d ). We selected this game because, unlike the prisoner’s dilemma and similar games (e.g., Kanazawa & Fontaine, 2013; Sparks et al., 2016 ), it is designed as an iterated game wherein the incentive to defect increases with each pass.

The dominant strategy in this game is for the first player to defect on round one (McKelvey & Palfrey, 1992) . However, people rarely adhere to the dominant strategy in simple economic games (Mailath, 1998) . Indeed, if one trusts one’s partner not to defect, one’s payout is likely to be significantly larger with a cooperative strategy. Therefore, the number of rounds participants choose to pass the pools indicates trust behaviour. In our game, the task ended when either the participant defected or when the game had reached ten rounds with a given avatar. Participants received their game earnings as bonus money at the end of the experimental session. Participants played each avatar in random order with pool values organized in ascending order for each avatar. The dependent variable in this experiment was the number of rounds participants passed to each avatar. The data for this experiment may be accessed at: https://osf.io/ehpuf/ (likewise for Experiments 2a/b and 3 below; Heerey & Clerke, 2021 ).

To examine perceptions of avatar similarity, participants rated each avatar after the getting-acquainted task on the following item: (“I see [avatar] as similar to me”). Participants rated each item on a 7-point Likert scale (1 = Disagree strongly; 7 = Agree strongly) and rated the avatars in random order. To mask the fact that we were interested in similarity ratings, we embedded this rating in the Ten-Item Personality Inventory (TIPI; Gosling et al., 2003 ), which participants completed for each avatar. Question order was randomized within avatar. The additional TIPI items were not analyzed. 1

To ensure that our similarity manipulation appropriately enhanced perceptions of similarity, we conducted a 2 x 2 repeated measures ANOVA with similarity (low or high) and mimicry (low or high) as within-subjects factors and the degree to which participants rated avatars as “similar to me” as the dependent variable. Results revealed that participants rated high similarity avatars as more similar to themselves than low similarity avatars, F(1,55)=34.03, p<.001, η 2 =.382 ( Figure 2a ). Participants had a tendency to rate high mimicry avatars as more similar to themselves, although this effect was not statistically significant, F(1,55)=3.89, p=.054, η2=.066. There was no interaction between similarity and mimicry, F(1,55)=.054, p=.818, η 2 =.001. These results suggest that our similarity manipulation achieved its desired effect.

Blue fill represents low-similarity avatars and grey fill depicts high-similarity avatars. Hatched plots show high mimicry avatars and un-hatched plots denote low-mimicry avatars. Within each violin, white dots represent the median and the white notches the 95%CI of the median, the horizontal lines show the means, the dark gray bars represent the interquartile range (IQR), and the light gray lines represent 1.5 times the IQR. The shape of the violin shows the probability density function of the data distribution. Individual data points are represented by coloured dots.

To examine trust behaviour, we conducted an ANOVA with similarity (low or high) and mimicry (low or high) as within-subject factors and the total rounds played (see Table 1 ) with each avatar as the dependent variable. Results revealed main effects of similarity, F(1, 55)=7.42, p=.009, η 2 =.119, and of mimicry, F(1,55)=11.75, p=.001, η 2 =.176 ( Figure 2b ). In both cases, consistent with hypotheses, higher levels of similarity and mimicry led to greater trust. The similarity by mimicry interaction was non-significant, F(1,55)=1.59, p=.213, η 2 =.028.

These results suggest that both similarity and mimicry shape decisions to trust. Specifically, people are more trusting of those whom they perceive to be high in similarity as well as those whose behaviour demonstrates greater mimicry. These results suggest that manipulated attitude similarity enhances trust, as in previous research, and that mimicry independently contributes to trust decisions.

The aim of Experiment 2 was to replicate Experiment 1 using a different economic game. Here, we chose an “investor-trustee” game (Berg et al., 1995) . In the typical version of an investor-trustee game (see Berg et al., 1995 ), the “investor” receives an endowment and must then choose how much of the endowment to invest with a “trustee.” If money is invested, the trustee receives a “matured” investment (typically triple the initial investment; see King-Casas et al., 2005; Kosfeld et al., 2005; Shore & Heerey, 2013 ). The trustee then chooses how much of the investment to return to the investor. The amount invested indicates the degree to which the investor trusts the trustee. The investor- trustee game is frequently used to operationalize trust in experimental contexts (e.g., Bailey et al., 2015; Kosfeld et al., 2005; Shore & Heerey, 2013 ).

As above, we predicted significant main effects of both similarity and mimicry on trust behaviour, such that greater levels of each would enhance trust, as measured by investment amounts (Experiment 2a). We also conducted a direct replication of the results using the exact same methods with an independent sample (Experiment 2b). The purpose of this replication was to increase our confidence in our results while engaging in open science practice. As such, Experiment 2b was pre-registered on the Open Science Framework (OSF; osf.io/dv7np).

Participants . Seventy-four participants completed Experiment 2a in exchange for partial course credit and a small monetary bonus, which was based on their performance in the investor- trustee game. Of these 74, nine were discarded from the analysis due to deception failure. The final sample therefore included 65 undergraduates (19 male, Mage=18.63, SD=.86). Our sample size for Experiment 2b was 72 participants (19 male, Mage=18.70, SD=1.14), after excluding 13 individuals for deception failure.

Procedure . Here, we used the same protocol as above to manipulate similarity and mimicry. After the manipulation, participants played a 10-round version of the investor-trustee game in which each participant played 5 rounds as the investor and 5 rounds as the trustee with each avatar. Asking participants to play both game roles allowed us to maintain the deception that participants played real partners, however we did not analyze trustee behaviour. Participants played avatars individually and all game rounds occurred in random order without feedback, to ensure that the presence of investment feedback did not influence results.

In the investor role, participants received a 10-point endowment and chose what proportion to invest with their partner on that round using a key press (allowing the full endowment range; 0-10 points). Participants played 40 trials of the game in random order, with 10-rounds being played with each avatar. To ensure that participants’ behaviour when they played the trustee role was not affected by differential investment amounts across the avatars (i.e., one avatar being more “trusting” than another), each avatar invested 3, 4, 5, 6 and 7 points over the course of the five trials in which participants served as trustees. Trial order was fully randomized, and debriefing confirmed that no participant guessed this manipulation. Finally, participants knew that they would receive their game earnings as a monetary bonus at the end of the experimental session.

To test whether our manipulation effectively altered perceptions of similarity we examined their ratings of the avatars on two items, how “similar to me” and “in sync with me” they thought each avatar was. To examine the reliability of this 2-item scale, we used the Spearman-Brown coefficient (see Eisinga et al., 2013 ). Across Experiments 2a and 2b, this correlation averaged .775, suggesting that these items likely measured a general “similarity” construct. We therefore averaged the two items within these datasets (and in Experiment 3 below). To examine these ratings, we conducted a pair of 2 x 2 repeated measures ANOVAs with similarity (low or high) and mimicry (low or high) as within-subjects factors and avatar ratings as the dependent variables for Experiment 2a and 2b (Figure 3a and 3b ). Results revealed that participants rated high similarity avatars as more similar to themselves than low similarity avatars (Experiment 2a: F(1, 64)=15.19, p<.001, η 2 =.192; Experiment 2b: F(1, 71)=48.97, p<.001, η 2 =.408). As in Experiment 1, they also rated high mimicry avatars as more similar to themselves than low mimicry avatars, though this result did not quite reach threshold for statistical significance in Experiment 2b (Experiment 2a: F(1, 64)=28.725, p<.001, η 2 =.310; Experiment 2b: F(1, 71)=3.91, p=.052, η 2 =.052). There was no interaction between these variables (Experiment 2a: F(1, 64)=2.09, p=.153, η 2 =.032; Experiment 2b: F(1, 71)=.51, p=.478, η 2 =.007). These results suggest that our manipulation of similarity generally achieved its predicted effect.

To examine participants’ trust decisions, we calculated the average number of points that participants invested with each avatar (see Table 2 ). These data served as the dependent variable in our analyses. As above, repeated-measures ANOVAs with similarity (low or high) and mimicry (low or high) as within-subjects factors and the average points invested with each avatar as the dependent variables, revealed both main effects of similarity (Figure 3c and 3d ; Experiment 2a: F(1, 64)=6.59, p=.013, η 2 =.093; Experiment 2b: F(1,71)=13.46, p<.001, η 2 =.159) and mimicry (Experiment 2a: F(1, 64)=6.28, p=.015, η 2 =.089; Experiment 2b: F(1, 71)=8.10, p=.006, η 2 =.102). In both cases, consistent with our hypotheses, higher levels of similarity and mimicry led to greater trust. As above, the similarity x mimicry interactions were not significant (Experiment 2a: F(1, 64)=.33, p=.570, η 2 =.005; Experiment 2b: F(1, 71)=.10, p=.749, η 2 =.001).

As in Experiment 1, we found that people trust those who are more similar and who mimic more frequently than people who are lower in similarity and mimicry. These results establish mimicry as an important predictor of decisions to trust, alongside similarity, at least in the context of economic games.

The purpose of this experiment was to test the robustness of similarity and mimicry as predictors of trust behaviour in decision-making, as it relates to social scenarios. In addition, this experiment sought to disentangle the contributions of similarity and mimicry to trust decisions. Because each avatar has both a similarity value (low or high) and a mimicry value (low or high) in a fully crossed design it is difficult to examine the degree to which these variables might be related. We therefore apply an idea from utility theory (von Neumann & Morganstern, 1947) to test this question. Specifically, we ask how participants apportion their choices across pairs of avatars, depending on how those avatars differ in similarity and mimicry. Participants make repeated choices with each possible avatar pairing. Based on how much they prefer high similarity or high mimicry avatars relative to low, we can estimate the degree to which similarity and mimicry independently and jointly determine participants’ avatar preferences across a set of social trust decisions. This type of design is a common method for examining the degree to which particular variables contribute to preferences within a decision space (e.g., Glimcher & Rustichini, 2004; Kandasamy et al., 2014 ). As above, we predicted that greater levels of both similarity and mimicry would enhance trust and that they would do so independently.

We further hypothesized that mimicry might be a more important predictor of choices than similarity because in real-world interactions, partner mimicry may be apparent before information about that partner’s attitudes and beliefs is known. Thus, after appearance similarity, mimicry may the first interpersonal cue that people send. These hypotheses were pre-registered on the OSF (osf.io/dv7np) prior to data-collection.

Participants . Ninety-seven participants completed this study in exchange for partial course credit. Of these 97, we discarded a total of 12 participants: seven due to deception failure, three due to inattentive task performance, and two for both deception failure and inattentive task performance. Inattentive performance on the decision task was defined as having made a choice in less than 350ms on 25% or more of the test trials. This data exclusion decision was based on research suggesting that 350-400ms is necessary for people to read and understand short phrases, like those in our decision task (Kutas & Federmeier, 2011; Kutas & Hillyard, 1980) . Thus, responses shorter than 350ms in this task are likely to be anticipatory responding that is not representative of deliberate decision-making. The final sample therefore included 85 undergraduates (23 male, Mage=18.68, SD=1.29).

Procedures . As in Experiments 1 and 2, the task began with our manipulation of similarity and mimicry. However, here we measured trust using a social decision task in which participants identified a preference for one of two avatars for a hypothetical trust scenario (e.g., “who would you ask to be your designated driver”). The trust items were selected using pre- ratings from an independent set of participants based on 1) the extent to which each scenario affects university undergraduates and 2) the extent to which the item is a good indicator of whether one trusts someone (see Appendix 1.).

Participants viewed each of the six possible avatar pairings (e.g., avatar 1 & avatar 2, 1 & 3, 1 & 4, 2 & 3, etc.) for each of 14 trust items (84 test trials). The placement of the avatars within each choice pair was counterbalanced so that each avatar appeared on the left and on the right side of a pairing with equal frequency. How participants apportion their choices in this task, depending on the relative differences between the avatars within a pairing, demonstrates the degree to which similarity and mimicry guide their choices. For example, when selecting between avatars with equal similarity values, do participants show a preference for the avatar that displays mimicry to a greater degree? If a participant consistently prefers avatars that demonstrate greater mimicry, this indicates that mimicry is an important driver of trust decisions to that participant.

Data Analysis . To examine the degree to which avatar similarity and mimicry shaped choice behaviour, we individually modeled each participant’s choices using a logistic model. The model estimated the likelihood that a participant would select the avatar on the left, given relative differences in avatar characteristics. The model included terms for avatar similarity and mimicry, coded as the difference between the left and right avatar for each variable. We used a standard logistic model to fit the choice data.

The parameter θ in the logistic equation was estimated as:

In this equation, the β s are the estimated regression weights for each model term. β 0 refers to the intercept; β 1 is the degree to which similarity influenced choice behaviour; β 2 is the estimated regression weight for mimicry; and β 3 is the similarity x mimicry interaction. The X s in the equation represent the difference between the avatar on the left and the avatar on the right. X 1 coded similarity. For example, if the left avatar in a pairing was high in similarity and the avatar on the right was low, we coded this as 1. If these values were reversed, this was coded as -1. If the similarity values were identical, the trial received a code of 0. We used a similar coding system for mimicry ( ⁠ X 2 ⁠ ). The similarity x mimicry interaction ( ⁠ X 3 ⁠ ) was the product of the coded differences. The model used an iteratively re-weighted least squares algorithm to obtain the maximum likelihood estimate for each of the terms (O’Leary, 1990) . The more the estimated regression weights differ from zero, the greater the influence of each term on a participant’s decisions. Importantly, we determined the model coefficients on a subject-by-subject basis because that allowed us the opportunity to estimate the distribution of these preferences across individual participants, rather than generating a single coefficient for the entire sample as a whole. The model coefficients were subsequently used as dependent variables in additional analysis.

As in Experiments 1 and 2, we conducted a repeated measures ANOVA with similarity (low or high) and mimicry (low or high) as within-subjects factors and participants’ ratings of the avatars as the dependent variable. Results ( Figure 4a ) revealed that participants rated highly similar avatars as more similar to themselves, F(1, 84)=30.22, p<.001, η 2 =.265, and also rated high mimicry avatars as more similar to themselves, F(1, 84)=25.65, p<.001, η 2 =.234 ( Figure 4a ). The similarity x mimicry interaction was not significant, F(1, 84)=1.32, p =.254, η 2 =.015.

Figure 4b shows the proportion of choices participants allocated to a particular avatar, given relative differences in avatar characteristics within a given pair. For example, when choosing between avatars that were identical in similarity (e.g., both low or both high) but different in mimicry, results show that participants selected the high versus the low mimicry avatar about 63% of the time ( Figure 4b , top bar). Likewise, when choosing between avatars with identical mimicry rates participants preferred the high versus the low similarity avatar significantly more often than chance. When they chose between the avatar that was low on both similarity and mimicry and the avatar that was high on both variables, they selected the high similarity/mimicry avatar about 77% of the time. Interestingly, participants demonstrate indifference when choosing between the low similarity/high mimicry avatar, and the high similarity/low mimicry avatar, meaning that they do not have a significant preference for one avatar over the other in this pairing (bottom bar). This suggests that participants weigh similarity and mimicry approximately equally when making trust decisions. To test whether similarity and mimicry influenced trust decisions, we conducted a set of one-sample t-tests (against a test value of 0) with the individually estimated unstandardized regression weights as the test variables. Results revealed that both similarity, t(84)=6.10, p<.001, and mimicry, t(84)=4.46, p<.001, were significantly greater than zero indicating that both influenced trust decisions such that avatars higher in both traits were preferred across the sample ( Figure 4c ). The interaction between the two variables was not significant, t(84)=-1.72, p=.090, suggesting that these effects may be additive, rather than interactive. A paired-samples t-test with the regression weights of similarity and mimicry as the test variables revealed that, contrary to our expectation the two variables were not significantly different from one another, t(84)=.72, p=.47, meaning that their effects on trust decisions are likely to be similarly important. These results support two of our three hypotheses. As in Experiments 1 and 2, people trust those who are more similar and demonstrate greater mimicry than people who are lower in similarity and mimicry. However, contrary to prediction, both similarity and mimicry made equal contributions to participants’ trust decisions. Thus, mimicry did not contribute more significantly to trust decisions than similarity, even though it may be apparent in real-world decisions before interaction partners have enough information to judge attitude similarity.

Results from Experiments 1 and 2 demonstrate that the presence of mimicry made an important contribution to both financial and social trust decisions, over and above the contribution of similarity. Interestingly, the magnitude of the mimicry effect was quite comparable to that of similarity but did not interact with it. These results suggest that mimicry explains important and unique variance in social outcomes and suggest it as a central element in trust decisions. This is significant given that previous work has conceptualized similarity as one of the largest contributors to trust decisions (e.g., DeBruine, 2002; Williams, 2001; Ziegler & Golbeck, 2007 ).

We have argued that mimicry is an equally important contributor to social decisions because it is evident before people gain knowledge of another’s attitudes and beliefs. As people become acquainted, they exchange not only personal information but also contingent social cues (Cialdini & Goldstein, 2004; Heerey & Crossley, 2013) . As relationships deepen, so does disclosure depth (Collins & Miller, 1994) , suggesting that only after a suitable acquaintanceship might someone disclose highly personal views and attitudes. Thus, after visual cues, such as whether someone looks trustworthy, knowledge of mimicry has temporal precedence over knowledge of attitude similarity in the real-world and may help to explain its contribution to decisions.

Both interpersonal similarity and mimicry have been linked to the perception of interaction smoothness, which appears to be necessary for successful interactions (Byrne, 1971; Delaherche et al., 2012) . Because mimicry is so frequent (Oullier et al., 2008) , people may come to expect its presence (Heerey, 2015; Heerey & Crossley, 2013) . Without expected levels of mimicry and other contingent social behaviours (e.g., Hale et al., 2019) , people may perceive greater disfluency and awkwardness during interactions, leading to perceptions of dissimilarity. Indeed, interaction disfluency is associated with negative social judgments and subsequent dislike (Heerey & Kring, 2007) . Thus, mimicry may indeed be a subtle but important aspect of social interaction.

Mimicry may also change the way people actively acquire social information. If mimicry is high and initial interactions feel coordinated people may experience this as positive or rewarding (Delaherche et al., 2012) , thereby enhancing the likelihood that they actively seek out points of similarity between themselves and their interaction partners. This may mean that people perceive interaction partners who engage in more mimicry as more similar to themselves, even when objective similarity levels are low. This is consistent with the notion that similarity need not be objective to enhance trust (Sanders et al., 2015; Whitmore & Dunsmore, 2014) . Thus, mimicry may underpin similarity perception and consequently shape trust decisions. Our research speaks to the need to include mimicry as an important predictor of trust behaviour.

Despite these interesting findings, this work has several limitations. One obvious limitation of this set of experiments is that it describes “interactions” that were completely computer controlled. Although this is the only way to perfectly control and manipulate social cues and information (Böckler et al., 2014; Schilbach et al., 2006) , such interactions lack elements of ecological validity. For example, this design made participants’ experiences of mimicry less automatic than they might be in face-to-face interactions. Specifically, participants deliberately chose emoji feedback, reducing the spontaneity of this process relative to real-time interactions. Even though evidence from studies of naturalistic interactions suggests that mimicry may be a conscious process at least some of the time (e.g., Hale et al., 2019; Heerey & Crossley, 2013 ), mimicry in face-to-face interactions is generally considered to be automatic (Chartrand & Lakin, 2013) . The present procedures brought this type of mimicry into the more deliberate realm. In addition, participants had no control over the types of information they exchanged, as they would in any natural interaction. Nonetheless, even though these avatar interactions were not as realistic as true interactions, all the participants we included in our analyses genuinely believed they had interacted with other people (for a comparison of those included in the analysis and those excluded for deception failure on age, gender, and personality traits see Supplementary Materials). The fact that we observed such consistent results even in the minimal social context of the present research, suggests that these interactions are an excellent proxy for real face-to-face behaviour.

A second limitation of this design is that participants liked avatar responses that were more “similar” to their own more than they liked dissimilar responses (see Figure S2 in Supplementary Materials). This meant that the highly similar avatars received and returned more positive feedback than those low in similarity. Although we note that over 75% of the feedback across all three experiments was positive, the low-similarity avatars returned sad emojis somewhat more frequently. However, this avatar behaviour was a consequence of our methodology that we anticipated. We opted to allow this aspect of the experimental design to occur naturally because pretesting suggested that avatars that were uniformly positive were simply not believable because they did not respond in the same manner as real humans. Interestingly, evidence suggests that even sad emojis may be viewed positively (Kralj Novak et al., 2015) . We therefore opted to avoid specific instructions to participants indicating that they should interpret the emojis in any particular way (e.g., mood indicators) and allowed the avatars to provide more human-like feedback. Despite this procedure, the low-similarity/high-mimicry avatars were preferred at a similar rate to the high-similarity/low-mimicry avatars, suggesting that this design decision was unlikely to have substantially affected results.

Finally, although the manipulation we used allows us to conclude that similarity and mimicry are important and independent predictors of trust decisions, it does not allow us to make strong conclusions about the relationship between them. We have argued that mimicry may precede perceptions of similarity and trust and that it may act to enhance interaction “fluency,” such that people perceive similarity even when it is low in an objective sense. This idea suggests an interesting future research direction. As Hale & Hamilton (2016b) note, the effects of mimicry are reduced under the tight control of the virtual environment, perhaps because these settings limit objective evidence of similarity (see Bailenson & Yee, 2005; Hale & Hamilton, 2016a ). We saw rather stronger effects when examining mimicry in the context of the exchange of personal information, suggesting that the effects of mimicry may be attenuated in the absence of disclosure. Evidence shows that disclosure (and the reciprocity of disclosure) is a natural part of social behaviour, even online (Barak & Gluck-Ofri, 2007; Collins & Miller, 1994) . Thus, some element of personal disclosure may be important in catalyzing feelings of similarity, even when mimicry is present.

Trust is a crucial aspect of interpersonal relationships as it underpins successful relationship develop. Therefore, understanding how trust develops and which factors predict it is important. Similarity and mimicry have been frequently investigated and found to be predictors of trust; however, the relationship between mimicry and trust may not be as robust as was once thought (Hale & Hamilton, 2016b) . Our work tested the reliability of these effects in three tightly controlled experimental designs in an effort to add some clarity to the existing literature. The data from these experiments demonstrate that people trust others more when they are highly objectively similar and when they engage in high levels of mimicry, meaning that both variables are likely to be important precursors to the feelings of trust that underpin relationship development.

Contributed to conception and design: ASC, EAH

Contributed to acquisition of data: ASC

Contributed to analysis and interpretation of data: ASC, EAH

Drafted and/or revised the article: ASC, EAH

Approved the submitted version for publication: ASC, EAH

We wish to thank Taylor Tracey for her assistance with data collection.

No external funding contributed to this work.

The authors have no competing interests to report.

Figure S1. Demographic differences between participants who were included in the analysis and excluded for deception failure on a) age, b) gender, c) personality characteristics. Error bars show the 95% CI.

Figure S2. Average emoji feedback chosen by participants for each avatar a) when responses to the similarity manipulation matched and b) when responses to the similarity manipulation did not match.

The full datasets used in this paper are openly available on the Open Science Framework (OSF; osf.io/ehpuf/)

Lend your car keys to

Believe lied to you about something

Ask for a character reference for a job from

Give your computer password to

Do a group project with

Believe intentionally gave you bad advice for an assignment

Let watch your pet while you are away

Give a spare house key to

Choose for a housemate

Lend $20 to

Let hand in an assignment on your behalf

Get class selection advice from

Ask to be your designated driver

Ask to take notes for you if you cannot make it to class.

Additional personality measures were collected in these paradigms as part of a Master’s thesis but are not analyzed for the purposes of this project.

Supplementary data

Recipient(s) will receive an email with a link to 'The Influence of Similarity and Mimicry on Decisions to Trust' and will not need an account to access the content.

Subject: The Influence of Similarity and Mimicry on Decisions to Trust

(Optional message may have a maximum of 1000 characters.)

Citing articles via

Email alerts, affiliations.

• Recent Content
• Special Collections
• All Content
• Submission Guidelines
• Publication Fees
• Journal Policies
• Editorial Team
• Online ISSN 2474-7394
• Copyright © 2024

Stay Informed

Disciplines.

• Ancient World
• Anthropology
• Communication
• Criminology & Criminal Justice
• Film & Media Studies
• Food & Wine
• Browse All Disciplines
• Browse All Courses
• Book Authors
• Booksellers
• Instructions
• Journal Authors
• Journal Editors
• Media & Journalists
• Planned Giving

About UC Press

• Press Releases
• Seasonal Catalog
• Acquisitions Editors
• Customer Service
• Exam/Desk Requests
• Media Inquiries
• Print-Disability
• Rights & Permissions
• UC Press Foundation
• © Copyright 2024 by the Regents of the University of California. All rights reserved. Privacy policy    Accessibility

This Feature Is Available To Subscribers Only

Sign In or Create an Account

The role of similarity and complementarity in the selection of potential partners for open innovation projects in family firms

• Open access
• Published: 11 August 2022
• Volume 60 , pages 1347–1367, ( 2023 )

Cite this article

You have full access to this open access article

• Julia K. de Groote   ORCID: orcid.org/0000-0002-2457-3562 1 ,
• Sabrina Schell   ORCID: orcid.org/0000-0002-4694-7713 2 , 3 ,
• Nadine Kammerlander   ORCID: orcid.org/0000-0002-7838-8792 1 &
• Andreas Hack   ORCID: orcid.org/0000-0003-1849-9420 2  

3598 Accesses

4 Citations

3 Altmetric

Explore all metrics

A Correction to this article was published on 08 September 2022

This article has been updated

Despite the increasing importance of open innovation endeavors, the process by which firms select partners for open innovation is not well understood. Even less is known about how family firms, which are characterized by their resource scarcity and desire for control, handle these processes. We aim to address this gap in the literature by investigating this selection process using a qualitative approach. Our findings are based on data gathered in 53 interviews from ten case studies and expert interviews, as well as secondary data. We find that, in order to engage in collaboration, family firms must manage their perceptions of the similarities and complementarities between themselves and their potential partner and integrate these into an accepted level of anticipated fit. During the selection phase, the elements of fit are weighed in light of the openness of the given firm and preferred levels and mechanisms of control, which are influenced by the family in the family firm. If the fit is deemed sufficient to enter into a partnership, the partnership is then advanced to the collaboration phase, where anticipated fit is translated into experienced fit, and aspects of similarity and complementarity are reassessed; this may potentially end existing partnerships, feeding back to future evaluations of fit and accordingly influencing future partnerships.

Plain English Summary

Family firms have unique ways to balance the openness, which is required to enter open innovation collaborations, and their preferences for control. They evaluate and reevaluate the fit with the potential innovation partner at several points in time. This might lead to the end of the relationship if the potential collaboration partner turns out to show less fit than expected. In the early phase of the process, firms tend to focus on complementary, e.g., in terms of bringing in skills that the family firm has not. While in the later phase of partner selection similarity, e.g., in terms of values and working style becomes more important. Being aware of the mechanisms that are at play might help firms develop more successful open innovation partnerships, as they can deliberately take measures to address the lack of perceived similarity by engaging in trust-building activities.

Similar content being viewed by others

Innovation as a mixed gamble in family firms: the moderating effect of inter-organizational cooperation

David and goliath: causes and effects of coopetition between start-ups and corporates.

An Open Innovation Project Typology of Exploration and Exploitation: Managerial Implications and Empirical Applications

Avoid common mistakes on your manuscript.

1 Introduction 

Selecting the right partner is an important success factor for interfirm collaborative endeavors (Geringer, 1991 ; Hitt et al., 2000 ). This includes open innovation, in which a firm works with customers, suppliers, universities, or even competitors to make innovation strategies more efficient and effective (Chesbrough, 2003 ). It is hence surprising that research on the selection process for open innovation partners remains sparse (De Groote and Backmann, 2020 ).

Initial insights into the open innovation processes of family firms indicate that peculiarities induced by family influence, such as resource scarcity and nonfinancial goals, influence the attitudes and behavior of family firms regarding open innovation (Brinkerink et al., 2017 ; Casprini et al., 2017 ; Lambrechts et al., 2017 ). The extant literature suggests that family firms tend to be rather reluctant to relax their firm boundaries and tend to have fewer open innovation partners than nonfamily firms (Classen et al., 2012 ). The tension between the family’s ability and willingness to enter open innovation relationships affects their behavior during the selection process (De Massis et al., 2014 ; Kotlar et al., 2020 ). Furthermore, because of the known heterogeneity among family firms (Chua et al., 2012 ), selection processes regarding open innovation partners might also differ substantially among these firms.

We know little about how family firms tackle the issue of the unwillingness to lose control when collaborating with external parties (Feranita et al., 2017 ). Furthermore, scant research specifically examines the relative role of various fit characteristics in the selection process (Shah and Swaminathan, 2008 ), and its temporal importance during the open innovation process. The importance of each selection criteria might vary depending on the type of project (Hitt et al., 2000 ) and the peculiarities of the family firm (Brinkerink et al., 2017 ). Because open innovation in general (Faems et al., 2005 ) and selecting the right partner for open innovation, in specific, are important success factors (Geringer, 1991 ; Hitt et al., 2000 ), it is important to better understand the partner selection process in family firms. Therefore, we pose the following research question: How do family firms select open innovation partners?

We aim to answer this research question by inductively and deductively analyzing ten case studies, using data based on 53 interviews with family firm members and open innovation experts. The present study contributes to the literature in several ways. First, we contribute to the literature on open innovation in family firms by developing a process model of the selection process in open innovation partnerships of family firms. We shed light on how perceptions of fit between partners evolve throughout the process of screening, selecting, and collaborating with external partners. We show that firms weigh facets of similarity and complementarity continuously throughout the process and that this weighing is closely linked to the difference between anticipated fit and experienced fit. While prior research has highlighted the importance of fit perceptions and the related concepts of complementarity and similarity (Bierly III and Gallagher, 2007 ; Emden et al., 2006 ), the mechanisms underlying the evaluation process of these criteria, and the interplay with firm characteristics, have thus far remained uninvestigated. Second, herein, we contribute to the understanding of how family firms are able to handle their unwillingness to lose control when entering open innovation partnerships (Feranita et al., 2017 ). Third, we further elaborate on the idiosyncratic preferences and drivers in family firms that influence this weighing process, for example, the family CEO him- or herself, thereby addressing calls to shed further light on the heterogeneity of family firms (Neubaum et al., 2019 ; Rau et al., 2019 ) and contribute to a better understanding of the role of the owner family in open innovation processes.

2 Theoretical background

2.1 open innovation.

Our definition of “ open innovation ” is based on the original definition provided by Chesbrough and Crowther as “the use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively. [This paradigm] assumes that firms can and should use external ideas as well as internal ideas and internal and external paths to market, as they look to advance their technology” (2006: 1). In family firms, cooperation with network partners compensates for the common scarcity of resources (De Massis et al., 2018 ; Lee et al., 1999 ) by enabling the firm to develop new technologies despite those limited resources (Parida et al., 2012 ; Werner et al., 2018 ). For example, companies purchase technological applications that have already proven useful to innovate their own products (Chesbrough and Crowther, 2006 ) or aim to acquire relevant knowledge (Casprini et al., 2017 ).

Although studies on family firms and innovation have recently proliferated (Broekaert et al., 2016 ; Calabrò et al., 2019 ; Cassia et al., 2012 ; Decker & Günther, 2017 ), very few studies on family firms and open innovation have been conducted (Brinkerink, 2018 ; Classen et al., 2012 ; Kotlar et al., 2013 ). The question addressed by most of these articles is whether family firms are more receptive toward open innovation than nonfamily firms, or vice versa (Gjergji et al., 2019 ). These articles suggest that family social capital and the desire to connect with stakeholders might enhance the ability and willingness to engage in open innovation partnerships (Miller et al., 2008 ; Sirmon and Hitt, 2003 ). Barriers to open innovation in family firms mentioned in the literature, which point to a more closed innovation focus, are limited diversity in management, the protection of control, and other socioemotional wealth components (Chrisman et al., 2016 ; Classen et al., 2012 ; Nieto et al., 2015 ). Most quantitative studies evince that family firms acquire external technological resources to a lesser extent than firms without significant family involvement (for a review, see Gjergji et al., 2019 , and for individual studies, see, for example, Classen et al., 2012 ; Kotlar et al., 2013 ; and Nieto et al., 2015 ), indicating that ability barriers and a lower proclivity toward open innovation are pronounced in family firms.

The existing qualitative studies focus on the comparison between family firms and nonfamily firms (Cassia et al., 2012 ), and they suggest that family firms are less willing to collaborate than nonfamily firms and that this has a negative effect on new product development. In addition, the role of trust (Hatak and Hyslop, 2015 ), the flow of knowledge (Casprini et al., 2017 ), and engagement in open innovation partnerships or barriers have been considered (Lambrechts et al., 2017 ).

2.2 Partner selection

The selection of open innovation partners can be broken down into three phases: screening potential partners (screening phase), actually making a selection decision (selection phase), and entering a collaboration (collaboration phase), which often includes the decision concerning whether to continue the collaboration or not (Emden et al., 2006 ). Partner selection is not only of importance in the context of open innovation but has been investigated in related contexts, such as the selection of strategic alliances (De Groote et al., 2021 ; Shah and Swaminathan, 2008 ). As the existing knowledge on open innovation partner selection in family firms is limited, we draw additionally on findings from this related stream of literature.

Choosing the right partner for cooperation means finding desirable matches between the resources, goals, and strategies of those partners (Das and Teng, 2003 ). From the outset, and with consideration given to the overall lifecycle of strategic alliances, firms have to look for a certain degree of fit between partners. Fundamentally, most criteria that result in a fit between partners of strategic alliances can be allocated to two main clusters, which we label as “complementarity” and “similarity.” Beyond these two clusters, the concept of “fit” is often used; in some studies, this term is used to refer to either complementarity or similarity, or a combination thereof. Footnote 1

Partner complementarity is typically defined as the extent to which a partner contributes resources and capabilities that the other partner lacks to the partnership (Dyer and Singh, 1998 ; Manotungvorapun and Gerdsri, 2016 ). Companies are more likely to enter into a partnership if the external partner has complementary resources, which the company can use in addition to its own resources (Chung et al., 2000 ). Such resources include, for example, specific market or technology knowledge (Emden et al., 2006 ). Partner similarity is typically investigated regarding cultural and organizational characteristics (Russo and Cesarani, 2017 ; Swoboda et al., 2011 ) and, in particular, the content dimensions of values, norms, and mindsets (Yoon and Song, 2014 ).

The understanding of “fit” does contain a positive connotation. It refers to aspects related to the similarities between firms but is also related to aspects beyond that. It can also encompass aspects of different resource bases that may be leveraged (i.e., complementary resources), and in some contexts, it is regarded as an input into (and in others, as an outcome of) relationships. To be able to potentially discover new content dimensions of characteristics, which are of relevance to our research context, and to understand patterns of how these dimensions interact, we conceptually differentiate among “similarity,” “complementarity,” and “fit.” We define the “similarity” of the partners as those shared characteristics that explicitly neither include a positive (or negative) connotation nor imply positive outcomes. We understand “complementarity” as characteristics of the firms that are dissimilar in their nature or value. We define “fit” broadly as the perception of how well firms mesh together.

2.3 Partner selection for open innovation in family firms

Several studies have investigated whether family firms differ from nonfamily firms in the depth and breadth of their search for external partners (Basco and Calabrò, 2016 ; Classen et al., 2012 ; Lazzarotti and Pellegrini, 2015 ). Such studies, however, have not focused on the partner search process in general but rather on the motivation for entering into a collaborative external partnership with a particular partner. To minimize the risk associated with uncertain innovation activities, family firms are assumed to work primarily with local partners, such as local suppliers or customers (Basco and Calabrò, 2016 ), since collaboration with competitors is less attractive for family firms, because of their inherent fear of loss of control (Chrisman et al., 2015 ). With regard to this topic, and the importance of complementarity and similarity considerations, two empirical studies on open innovation in family firms offer some initial insights and pointers for future research. While it was not at the core of their study’s intended goal, Casprini et al. ( 2017 ) found evidence concerning partner selection in family firms. When elucidating the distinctive barriers to knowledge transfer, the problems of finding a partner with the right fit regarding similarity (e.g., same language, same family values) were noted. One important aspect mentioned by Lambrechts et al. ( 2017 ) in terms of partner selection is that family firms aim to keep control and influence over the family firm, even in open innovation partnerships. This objective contrasts significantly with engagement in open innovation activities, which typically involve open innovation partners surrendering part of this control. Interestingly, family firms might stay in control during the collaboration phase by taking up the role of orchestrator in the collaboration phase. However, loss of control might also play an important role in the partner selection process, for example, by influencing trading up between complementarity and similarity perceptions.

As our research addresses questions of “how,” we chose an explorative qualitative approach (Yin, 2014 ). Our overall approach is based on recent state-of-the-art recommendations for qualitative research projects in the field of family business (Kammerlander and De Massis, 2020 ; Leppäaho et al., 2016 ; Reay and Zhang, 2014 ). We followed the case study approach by Eisenhardt ( 1989 ) and collected data based on semistructured interviews as well as secondary data (e.g., business reports, company websites). The collected data were subsequently inductively and deductively analyzed (Langley and Abdallah, 2011 ) using the MaxQda software package. To ensure transparency and replicability, we followed recent guidelines by Aguinis and Solarino ( 2019 ). Table A1 in the online Appendix shows a summary of how we addressed the different criteria of transparency and replicability.

3.1 Data collection

We conducted a total of 53 interviews; 40 of these were conducted with 38 representatives of 10 Swiss family firms. No consensus has been reached so far in the scientific community on how a “family firm” is to be defined (Diaz-Moriana et al., 2019 ; Hernández-Linares et al., 2018 ). For the current study, we have chosen a definition that encompasses the aspects of management and ownership but also the intergenerational element (Howorth et al., 2010 ). Based on Chua et al. ( 1999 ), we define a firm as a “family firm” when it meets the following criteria: at least 50% of the business is owned by a family, at least one family member is part of the top management team, and succession is planned/is in progress/has already occurred. To be part of our sample, family firms had to fulfill all of these criteria.

We collected our data in Switzerland. Previous research has demonstrated that national institutional environments influence entrepreneurial activity and innovation (Spencer et al., 2005 ) and the strategic choices of firms (Dacin et al., 2007 ). Specifically, Vasudeva et al. ( 2013 ) empirically showed that strategic partner selection is contingent upon corporatist institutional structures, which reflect differences in underlying cooperative norms, such as the importance of a partner’s social versus technological values.

Since our focus lies on family firms and their peculiarities in the selection process, we needed to find an institutional context that was not dominated by a pure market logic (which might place too much attention on complementarity issues) but, equally, was not dominated by a pure family logic (which might place similarity considerations at the fore) (see Cardinal et al., 2017 ). In this regard, moderately regulated regions, such as Switzerland, should provide an appropriate context for finding a balanced view between the family and the business for selection logics and mechanisms of how selection takes place (see the CASE project, Culturally sensitive Assessment Systems and Education, Gupta and Levenburg, 2010 ).

We started data collection with an in-depth explorative case study (C1) comprising 14 interviews. Over the course of conducting the interviews, we developed a trust-based relationship with the firm, specifically with the family CEO. Preliminary results were discussed with this key informant. Afterward, we followed a theoretical replication logic (Yin, 2014 ). Accordingly, we chose family firms from different industries, of different sizes and with different levels of ownership and management participation of the owner family (Morse et al., 2002 ). Using this approach, we were able to identify similarities and “contrasting results but for anticipatable reasons” (Yin, 2014 : 57). To identify further cases, the authors contacted firms that were known to them due to the firm’s innovation activities. Additionally, based on internet research, a research assistant compiled a list of family firms that presented themselves as being innovative. One data source for this compilation was a list of the winners of Swiss innovation awards. To identify relevant cases and gain additional insights and outside perspectives, we contacted Swiss (open) innovation experts and conducted interviews with them. The experts were then asked to refer further experts and relevant family firms. During data collection, we reached a saturation point, where new interviews tended to add little new information (O’Reilly and Parker, 2013 ; Wray et al., 2007 ).

We conducted semistructured interviews (Horton et al., 2004 ; Rabionet, 2011 ), and an outline of the questionnaire is shown in the online Appendix . For the expert interviews (Bogner et al., 2009 ), the focus was more on their overall insights into open innovation partnerships of family firms and less on the idiosyncratic processes in one company. All interviews and, whenever possible, phone calls and discussions of the preliminary research findings were recorded and transcribed. During interviews and meetings, notes were taken and discussed afterward among the research team. Respondents were generally part of the owner family and the top management team. In addition to the 40 family firm-specific interviews, we conducted 13 interviews with 15 Swiss open innovation experts. All interviews were recorded and transcribed verbatim immediately afterward, resulting in 1304 pages of transcript (348 pages for the expert interviews; 956 for family firm informants).

3.2 Data analysis

To structure and reduce the mass of data and information, we needed a data structure in the analysis process, for which we applied the Gioia method (Gioia et al., 2013 ) to support our positivist case study approach (Eisenhardt, 1989 ; Gehman et al., 2018 ; Leppäaho et al., 2016 ). We initially extracted first-order concepts from the interviews. In a subsequent step, we interpreted these concepts in the study context and aggregated them to second-order concepts. This aggregation was an iterative process and was undertaken by all authors moving back and forth between the data and the literature (Eisenhardt, 1989 ). In the final step, we subsumed the second-order concepts into three overarching themes. We began the data analysis by individually analyzing the cases. We then moved on to the cross-case analysis. To identify common patterns and themes across cases (Eisenhardt and Graebner, 2007 ), we iteratively analyzed the data while considering the extant literature (Eisenhardt, 1989 ), and we steadily refined emerging themes and patterns by revisiting the single cases. A major part of our effort was dedicated to ensuring the reliability of our analysis (Golafshani, 2003 ). We triangulated our data whenever possible by comparing the responses of the interviewees with the information we gained from secondary data, as well as with information provided by the external experts and, in some cases, innovation cooperation partners of the firms (Flick, 2004 ). Any differences of opinions among the authors about the interpretation of the data were discussed until a consensus was reached (Onwuegbuzie and Leech, 2007 ). Based on these discussions and feedback from interviewees and external experts, we refined our interpretation. We condensed our findings into a visual process model, which is presented in the findings section. To externally validate our model and the mechanisms of initiating and sustaining open innovation partnerships in family firms, we presented initial versions of the model to study participants, as well as experts external to the present research (e.g., other researchers, members of family firm unions), and we carried out discussions with them (Tables 1 and 2 ).

3.3 Data structure

An overview of the data structure is presented in Fig.  1 . Fit emerged from our data as the main driver of selection and, later, of relationship building between open innovation partners. We found that aspects of complementarity and similarity represented the subdimensions of fit. The interview partners stated that they look for partners who can offer knowledge, skills, and/or resources that they lack, i.e., complementarity . While we find that firms search for complementarity, they simultaneously look for similarity between themselves and their partners. This similarity is not restricted to factors such as the values and cultures of the firms, and it also refers to criteria such as firm size or geographical location (in our case Switzerland and the specific canton (state) in Switzerland). Moreover, the firms in our sample differ in their general standpoint on open innovation, which we label as openness in the following. In the data analysis, two main drivers of this openness emerged: the family-internal and family - external influences . The family-internal influences are driven by the openness of the owner family, the owner family’s decision maker (e.g., risk propensity), and its lifecycle. Older generations still formally or informally involved in the firm may act as obstacles to the process of opening firm boundaries, especially in multigenerational firms. The openness of the family firms within the study was also determined by family-external influences . We found one important driver to be the industry in which the firm operates. In some of the industries that were represented in the study, firms would simply be unable to operate without taking advantage of open innovation activities. For example, firms in the medical engineering industry are obliged to rely on fundamental research conducted by universities (e.g., C1). Industry characteristics might, however, also limit open innovation activities. Moreover, we found that the specific geographical context of Switzerland plays an important role. For example, we observed that some of the firms were influenced in their decisions by the “Swissness” of potential partners, with this term denoting firms that are perceived as being particularly Swiss.

Data structure

All family firms in our sample demonstrated that they wished to remain in control. The way firms sustain this preferred level and mechanisms of control , despite opening their firm boundaries to outsiders, is driven by two main factors: trust and contracts . Trust is mentioned in most cases as an important driver of relationship building in the open innovation process. This category encompasses, for example, positive experiences in the past, the goodwill of the partner in the network, and the type of partner, which is often based on perceived trustworthiness. Contracts are used to realize control, e.g., to ensure intellectual property rights, or to ensure that information is protected. Table A1 in the online Appendix provides an overview of sample quotes for all categories of the coding structure.

4 Weighing complementarity and similarity in the process of partner selection for open innovation in family firms

We integrated our findings into a process model, which is displayed in Fig.  2 . In what follows, we derive the model from our data.

A process model for weighing complementarity and similarity in partner selection for open innovation

4.1 Screening phase

The screening phase is the first phase of the selection process. When defining partner search criteria, most firms focus on complementary resources, technology, and knowledge because a firm that already has all the knowledge and resources it requires would be unlikely to perceive a need for collaboration. This phenomenon is illustrated by the following quote:

“Complementarity is certainly always important, in the sense that they need to bring in more or different knowledge. Most importantly, they need to have deeper knowledge in a specific field, and we have to need this knowledge.” (C2-I3)

However, beyond looking for complementary knowledge, resources, and technology, firms also prefer companies that have attributes similar to theirs for collaboration:

“In our case, this also means that we prefer companies that have a similar structure. For example, we have 500 employees; we are a family-run business. Now, if I take a big corporation that has 30,000 employees and an extremely complex corporate structure and we are there, even if I take a few million, we are small fry. That never really works [laughs]; they won’t align themselves with us either.” (C5-I3)

Thus, companies look for partners that are complementary and similar to them, e.g., in terms of company size, values, being owner-managed, or speaking a common language. The practical problem is that firms that are clearly complementary tend to be less similar. For example, while a complementarity in knowledge (e.g., being a software provider) is not by definition linked to dissimilarity in firm culture (e.g., status-based mindset), in practice, it is much more likely that dissimilarities in one dimension (e.g., being technology driven) are accompanied by dissimilarity in another dimension (e.g., company values). This discrepancy can be illustrated by a recurring example that we encountered during our research. Some of the family firms in the sample intended to work with start-ups in search of complementary resources. However, as early as during the screening (and selection) phase, the family firms became aware of dissimilarities, e.g., pace of decision-making or working methods, which then ended the potential collaboration before it began (e.g., C1, C2, C5, and E11). It is often impossible to find a partner that fulfills both criteria simultaneously, and hence, a perfect fit is unlikely to be found.

Companies deal with the difficulty of finding a perfect fit both in terms of similarity and complementarity by integrating perceptions of complementarity and similarity into an expectation of how well the firms will work together in the future. We label this expectation as “anticipated fit.” To reach an accepted level of anticipated fit, firms have to reach minimum levels of complementarity as well as similarity. Firms further assume that low levels of one (complementarity/similarity) may be compensated by high levels of the other (similarity/complementarity). Reaching an accepted level of anticipated fit is the basis for further evaluation of a potential partner.

“We want a sense of whether the team is reliable, whether you can work with them, whether you have the same philosophy, whether you somehow sense they’re giving you the razzle-dazzle. We don’t want salespeople on the other side; we want technology experts. These are the most important factors.” (C1-I1-T2)

We use the term “anticipated fit” because, at this point in time, firms have no definite inside knowledge regarding the different dimensions of complementarity and similarity, which constitute fit. Hence, as soon as the firms have developed a perception of the fit between the partners that they expect in the future relationship, they enter the next phase of the process, which we designate the “ selection phase .”

Proposition 1:

To enter an open innovation partnership, family firms weigh their perceptions of anticipated similarity and complementarity to reach an accepted level of anticipated fit during the screening phase, which represents the precondition to enter the selection phase.

4.2 Selection phase

We refer to the second phase of the process as the “ selection phase .” In this phase, family firms start to closely evaluate the different criteria, for which they previously screened their partners. We identify two main drivers (preferred level and mechanisms of control; openness) that influence whether the anticipated fit is translated into final partner selection or not.

Across cases, we found that all firms were anxious not to lose too much control in the partnership and, thus, tended to reach for an accepted level of perceived control. However, family firms differ in the preferred level of control and in the preferred mechanisms to secure control. For example, C5, a medium-sized family firm in the second generation, with family members in management and on the supervisory board, has a high need for control and generally prefers formal mechanisms (i.e., contractual agreements) to guarantee its influence in the partnership. Emphasis is placed on the importance of independence for the firm (as a family firm and as a cooperation partner). C5 knows and appreciates the concept of open innovation, but it has a very strong need for formal control, which impedes the realization of open innovation projects. This phenomenon is reflected, for example, in the observation that the supervisory board has secured numerous veto rights, even against its internal innovation team:

“We give the (Scrum) teams much freedom. However, we still have a right of veto. That means proposals from below, so to speak, and approval by the Board of Directors.” (C5-I1)

C6, for example, has a preference for parity and, thus, greater willingness to relinquish some control over the process, which they see as key to a functioning collaboration relationship; however, some firms prefer a clear asymmetry in power (e.g., C1). Case 2 has defined a product area where they are actively opening to outside partners (new products with fruits). For this product area, they show low levels of need for control and put much trust in their external partners. A similar approach, demonstrating trust and low needs for control for defined product areas, is demonstrated in case 10.

Typically, family firms ensure this asymmetry by relying on contracts or choosing partnerships in which they are neither obliged to share too much knowledge from within their company nor likely to become dependent in any way on the collaboration partner. However, how exactly the individual preference is displayed in a specific partnership also depends on the given situation. For example, in cases where there is a lack of similarity-based fit, family firms might use the mechanism of increasing their perceived control by regulating their partnerships with contracts as a way to bridge this gap. Trust and contracts complement each other as measures of increasing perceived control. This phenomenon is shown by firms that have established long-term relationships with open innovation partners.

“But besides the knowledge and the know-how and so on, the interpersonal level must be right. That is perhaps the basic thing, but on top of it there’s then also the knowledge, and then it will work out. Those are actually the two most important things. These two elements and trust.” (E1-I1)

These findings lead to our second proposition:

Proposition 2:

Family firms differ both in their level of need for control and in the preferred mechanisms to reach control in open innovation partnerships.

Apart from their preferred level and mechanisms of control, family firms also differ in their openness to engage with a broad set of external actors in their innovation activities. This level of openness arises either from a family-internal propensity to be open or from family-external pressures to do so, or a combination of both.

For instance, C8, a small family firm in the 7th generation and operating in the field of clothing production, has a high degree of openness, as displayed in their numerous innovation partnerships with customers and companies from noncompetitive industries. The following quotes from the owner-manager suggest that this openness is mainly driven by a family-internal proclivity to openness:

“Yes, for me, it is very clear that family firms that are managed by a family figurehead, or at least strongly influenced by a family owner are more willing to accept changes, to do something new in order to become stronger, than purely management-driven companies.” (C8-I1) “From my point of view, that’s partly connected with culture. You also have to look at it from a psychological perspective, a little bit of ego, of arrogance. Things like: ‘Yes, we’ve [the owner family] been doing this for 150 years, it’s good, we’re on top of it’.” (C8-I1)

Other family firms exhibit a similar openness. For example, C1 uses means such as tech-scouting to integrate ideas from the outside and, consequently, establishes a bridge to other firms. In contrast to C8, though, this firm conducts open innovation projects not because they are internally convinced of the general benefits of openness but because they feel they are forced to do so by external factors such as industry dynamics. These features are especially pronounced in the medical engineering industry in which C1 is competing. Hence, C1 opens up its firm boundaries with the awareness that openness is an inevitable prerequisite to remaining innovative.

Other family firms, such as C2, are also strongly influenced by family-external factors:

“When it comes to the core area, what we call our secrecy area, where it concerns questions of production, of ingredients, the procurement of innovations is done in-house, by a relatively small circle of people who deal with these topics that have to be kept secret. They have to get the ideas from somewhere, of course. (C2-I6) I: Ok. And is there or was there a cooperation with an external person who somehow touched this core area. C2-I6: Yes, yes, there is, yes. At some point we reach our limits and we have to.”

The perceived external pressure goes beyond the influence of the owner family and positively influences the openness for innovation of the firm, which leads us to the following proposition:

Proposition 3:

Family firms differ in their openness, which is determined by family-internal as well as family-external factors.

Once an accepted level of anticipated fit is reached, family firms engage in a process to come to a final selection decision. Within the selection phase, the firms’ preferences for control and openness play an important role and interact with one another in a complex weighing process.

“But it always has to be weighed up, how open I am to the outside world and how much is important now or how important it is to protect my know-how at this point. Through patents, etc. Yes. And if they do, once that is protected, the know-how, then I can become a bit more open.” (C5-I2)

As outlined above, family firms are heterogeneous in their preferred levels and mechanisms of control, as well as their openness. Based on these individual configurations, they develop their unique approach toward the final decision concerning whether a potential partner is further taken into consideration for collaboration or not. The following examples illustrate the diverse forms that might be assumed by the weighing-up process.

C3 is a small family firm in the second generation that is active in the real estate sector. Its openness toward engaging with external actors in the innovation process is mainly driven by family-external factors. The decision-makers of the firm would prefer to work completely independently, but simultaneously, they feel that the fast pace of today’s world obliges them to open up more. Making things even more complicated, they generally show a high need for control. To dissolve these conflicting needs, C3 decided to cooperate only with external actors without any direct competition to its own business model in their own region and who are members of a trade union. This approach has been followed for many decades.

Interestingly, their specific cooperation approach has had a minimizing effect on their preferred level of control because collaborations with external actors not deemed to be competitors are seen as less threatening both to their existing business model and to their competitive position in the industry. Consequently, C3 relies upon less formal control mechanisms; that is, they trust that their external partners adhere to the obligatory ethical code of the trade union.

A quite different weighing-up process and resulting approach could be identified for C6. C6 is a medium-sized family firm in its third generation that is run by a family CEO. The company is a manufacturer of customized special machines. The company is therefore often the innovation partner of other companies. The firm also feels strong external pressure to collaborate with external partners. Simultaneously, the family effect in fostering openness is very pronounced. Employees refer frequently to the owner family when talking about innovation activities of the firm. While being perceived as entrepreneurial, their need for control is also very pronounced, especially because of the CEO’s personality. C6 has chosen an interesting solution to being open to external partners while simultaneously meeting their strong need for control:

“There was a top engineer who worked in a big corporation as a technology leader. And he went into business for himself. Actually, at first, the idea was that he and his team would found a subsidiary in Italy and he would be hired by us and would do all these developments for us. But I didn’t actually want that, because I said that the risk was far too big for us and that he should also take entrepreneurial risk. We said, ‘you have to set up the company yourself, you have to hire the people you need […] (C6-I1)

A possible collaboration with a much bigger firm was perceived as more threatening (due to a perceived power imbalance) in comparison to working with a start-up. Asking the potential collaboration partner to set up his own business shifted this perceived power imbalance toward the external partner. This shift resulted in a less pronounced need for control for C6 and, as a consequence, a less pronounced need to formally control the external partner. To “compensate” the external partner, C6 builds on reciprocal trust as an informal control mechanism. They offered exclusiveness and guaranteed the partner’s revenue stream.

“[…] we guarantee you [the engineer], that for one to two years, we will fill your pipeline so that if you work exclusively with us, you will have enough work coming your way to employ these people’. That’s what we actually guaranteed.” (C6-I1)

Both examples highlight the complex interplay of several influencing factors. The family firms in our sample find different ways to weigh preferred levels and mechanism of control with their levels of openness. Thus, we posit the following proposition:

Proposition 4:

Given an accepted level of anticipated fit, the weighing-up of partner characteristics and situational characteristics against preferred levels and mechanisms of control and openness influences if and how partnerships are taken forward to the collaboration phase .

4.3 Collaboration phase

The considerations based on anticipated fit, perceived control, and openness define the starting point of the collaboration phase . After the decision to collaborate has been made, the fit between the firms is experienced. The anticipated fit between them is transformed into an experienced fit. Some family firms emphasize that it takes time to determine whether their partners truly fit. Above all, it is important to develop a common understanding of the project, as the following quotation illustrates.

“And what I actually believe is that we really want to collaborate with partners. A partner only becomes a partner when we work together on several projects and on several topics. You know each other, you’re faster, you get to know the language. Now we have two or three companies with which we collaborate… when a topic comes up, we ask them first. We know each other, it’s fair, the costs are fair, and we know what we can expect from each other.” (C6-I1)

The way similarity and complementarity perceptions are weighed as the basis of fit changes over the course of the selection and collaboration process. When initiating collaborations (in the screening phase), family firms search for new input, for example, in the form of technologies. As outlined above, these complementarities typically come with differences in, for instance, company size or age. For example, the partners tend to speak a different language, have different processes, and operate at different paces. While differences in language and culture do not, as a rule, harm the perceived value of the knowledge or technology to be acquired (which is the driver of the selection in the early phase), they might substantially harm the partnership in the long run when they become obvious in the collaboration.

“It’s a bit like the foundation of the project. What would I give more weight to? If I have confidence in someone and he can’t live up to it, that doesn’t help me either. So, I think it’s important that the technical requirements have to be met; otherwise, we won’t even get into conversation with them. And then the soft factors come to the fore over time I think.” (E11-I1)

This quote illustrates that, over the course of the selection and collaboration process, the characteristics related to similarity tend to become more important than at the beginning, where the focus often lies on acquiring complementary knowledge, resources, and technology. There are frequently discrepancies between previous perceptions of fit and the experience of fit. First, it becomes apparent during cooperation whether complementarity and similarity have been correctly assessed.

“If someone tells me: ‘Yes, I can do that’, and afterwards can’t do it, then the topic is also finished [...]” (C6-I3)

Second, it is also noticeable when partners change, and the fit is no longer sufficient for the relationship to continue.

“I have my requirements. We also had an external partner who changed in the course of the collaboration. He just grew to be too big for his boots, and I told him it doesn’t work like that, and then we just said goodbye.” (C7-I2).

One of the main reasons for ending a potential partnership was negative experiences in the past, where the family firms had learned that their values, culture, and language were not compatible with, or sufficiently similar to, those of the partner with whom they intended to collaborate (C1, E1, C2, and C6).

“There have been cases where we have cancelled a project because it did not fit into our strategy. That happens when a customer says to us: ‘Can you develop that for us?’ Yes, that sounds extremely exciting, but we start to get bogged down, we are too small. Then we say: ‘No, we won’t do that,’ because it just doesn’t fit into our strategic orientation. […] Plus of course ethical reasons and all sorts of such things, that’s logical, that’s clear, we would not do it then either. Money rules the world, but not us.” (C6-I4)

This leads to our fifth proposition:

Proposition 5:

If the experienced fit deviates substantially from the previously anticipated fit and is too low, the partnership is terminated.

In most of these cases, when a partnership was ended due to a gap between anticipated and experienced fit, this insight not only ended this specific partnership but also had consequences for future collaborations. The experience gained also allowed firms to define more concrete ideas and expectations for future partnerships.

For example, in C6, the importance of not only the common language “German” but also the subcategory “Swiss German” was emphasized. This was associated with the finding that regional proximity and the opportunity to collaborate at short notice can create proximity and, above all, pay off in cooperation. Prior experience has shown that cost advantages can be exploited through cooperation, e.g., with companies from southern Germany or Italy, but in the firm’s view, the common language (Swiss German) and regional proximity outweigh these advantages.

“So, you have to be, I’m not saying rigid... but you must have a clear idea about what is important to you in a partnership.” (C2–I1)

The way partnerships evolve, especially in terms of experienced fit, feeds back to how firms evaluate complementarity and similarity in future projects, how open they are to future collaborations, and what their accepted level of perceived control is.

“The implementers, i.e. those who make a business out of it, work differently than the innovators or those who find out where and how it actually works, i.e. the applied researchers. And when they talk to each other, they don’t understand each other, because the same words have different meanings. They think differently and they perceive the world differently. And because of this, there is an inherent risk in every case, of misunderstanding. That means you have a… you can say it like this, it’s like they speak a different language. They speak the same language, so they all speak, I don’t know, Swiss German, but it’s a different way of thinking, so to speak. And that needs a translation, like when you have a change of language. And you have to plan for that. You forget it, because you haven’t discovered it yet.” (E2-I1)

Firm characteristics that are easily perceived in the early phases and are not judged as problematic (e.g., a very small firm size) can become more salient and negatively weighted based on negative experiences (e.g., a collaboration that went wrong with a small firm), which feeds back to how family firms will evaluate potential partners in the future. This applies to unacquainted partners as well as partners the family firm has worked with before. This phenomenon leads to our final proposition:

Proposition 6:

The way firms experience fit based on different facets of similarity and complementarity in the collaboration phase influences how they weigh complementarity and similarity in assessing anticipated fit in future potential collaborations.

5 Discussion

Open innovation is becoming increasingly important for firms of all sizes and types; however, research on this topic has been slow to progress in the context of family firms (Brinkerink et al., 2017 ; Feranita et al., 2017 ). Our findings highlight that family firms can develop successful strategies for open innovation that minimize the perceived loss of control but simultaneously enable effective innovation collaboration (Feranita et al., 2017 ).

5.1 Theoretical implications

This study makes at least three contributions to extant research. First, it sheds light on the partner selection process for open innovation projects. Our findings highlight the complex interplay between complementarity and similarity. While “fit” is a commonly used concept in extant literature (Bierly III & Gallagher, 2007 ), researchers typically are ambiguous regarding the meaning of “fit.” Our findings highlight that perceptions of fit are the most important criterion for partner selection in open innovation processes of family firms and that these perceptions of fit are driven by similarity and complementarity. While complementarity is commonly perceived as the main driver of building open innovation partnerships (Bierly III and Gallagher, 2007 ; Chung et al., 2000 ; Emden et al., 2006 ; Manotungvorapun & Gerdsri, 2016 ), the interplay of complementarity perceptions and similarity perceptions has been neglected to date. While research has highlighted the importance of both (De Groote et al., 2021 ), a deeper understanding of the interplay of both in a concrete decision situation has been missing thus far. Furthermore, our process perspective implies that the gestalt of the “ideal open innovation partner” differs across stages, a “one-for-all” solution typically does not exist, and firms hence need to accept trade-offs regarding their partners throughout the process. For example, firms were willing to accept lower levels of complementarity in favor of higher similarity, even if this reduced the potential open innovation outcomes.

Second, while complementarity and similarity as partner selection criteria have been discussed in the broader alliance literature (Chung et al., 2000 ), the importance of each selection criterion has thus far remained unclear and might depend on the type of alliance project (Hitt et al., 2000 ). Furthermore, only a little research has examined the relative role of various fit characteristics in the selection process (Shah & Swaminathan, 2008 ) and how their importance changes during the open innovation process. Hence, our study extends prior research by studying the specific context of (Swiss) family firms and analyzing the relative importance of complementarity and similarity along the collaboration process. Counter to what one might expect from a perspective residing in a financially driven logic, family firms do not take advantage of cost benefits and prefer working with local (as opposed to international) partners whom they already know from their network or whom they could visit on site and, thus, control if necessary. While this is in line with prior findings highlighting that family firms might forgo financial benefits to preserve SEW (Feranita et al., 2017 ) and due to their risk aversion (Chrisman et al., 2015 ), the extent to which this kind of behavior was observed in our sample was still surprising, as even partners from different yet comparatively similar and close countries (i.e., Italy and Germany) where in some cases not considered for collaboration despite considerable financial incentives.

Third, our study also highlights the heterogeneity of family firms’ innovation behavior. Especially in the domain of innovation research, insights into family firms are often restrained to comparisons of family vs. nonfamily firms (Classen et al., 2012 ). Therefore, scholars have recently called for more research that seeks to understand family firm heterogeneity (Chua et al., 2012 ). Our data reveal that family influence can impact open innovation in two ways. First, firm characteristics, such as willingness to engage in open innovation, are typically closely tied to the values of the involved family members. This focus is especially salient during the selection phase, when the influence of the specific characteristics of the family firm decision-makers as well as the general openness of the owner family comes into play. Second, in line with the existing literature (Feranita et al., 2017 ), family firms do indeed strive to retain control, even when entering open innovation partnerships. However, adding a new nuance to extant research, we identified substantial differences between firms that were at times catalyzed by region-specific (Swiss) effects. This finding highlights that not all family firms seek high levels of control for the same reasons, and not all family firms are generally reluctant toward open innovation. Hence, open innovation in family firms is less a question of whether partnerships are established and more about how they are established.

5.2 Practical implications

The present study findings also offer important implications for business practice. By shedding light on the peculiarities of how family firms handle partner selection in the context of open innovation, our findings contribute to a better understanding of why many innovation collaborations fail (Du et al., 2014 ). An extensive focus on the complementarity of resources during an early phase might cause problems in later stages when firms become aware of a lack of similarity, which ultimately inhibits collaboration. Being aware of the mechanisms that are at play might help firms develop more successful open innovation partnerships, as they can deliberately take measures to address the lack of similarity by engaging in trust-building activities. In regard to putting open innovation activities into practice, the involvement of the owner family can be both a risk as well as an opportunity. Different measures might help practitioners overcome a lack of openness in the family firm or among family firm decision-makers. One possible solution is to give decision-making authority to independent teams or hiring external experts to drive open innovation. Linking the idea of openness to structures within the family firm can help avoid blind spots regarding the opportunities of such activities. Another solution is to systematically assess the firm’s strengths and weaknesses, especially those related to rigid internal rules (e.g., “we do not process fruit”), and to search for external expertise to address specific innovation-related issues. However, our data show that such partnerships would be too much of a stretch for some firms, as they go beyond their accepted level of control. Moreover, requirements of openness and perceived control differ based on the specific partners—an insight that family firm decision-makers must take into account when building up their relationships (Brinkerink et al., 2017 ). Finally, our results show that family involvement can also promote open innovation projects. Specifically, integrating members of the next generation can be a successful way to promote open innovation in family firms.

5.3 Limitations and future research

The present study has several limitations, which offer promising paths for future research. First, while we collected data from different sources, gathering various perspectives, we mostly used interview data referring to relationship building that occurred in the past. This approach might be prone to retrospective bias (Merkl-Davies et al., 2011 ). While this methodological shortcoming is common for qualitative studies using interviews (Cox & Hassard, 2007 ), a promising avenue for future research is collecting longitudinal data on partner selection and investigating how relationships between family firms and individual partners evolve over time. Such an approach could rule out biases regarding the key informants’ memories about their initial set of potential partners and hence contribute to a better understanding of the link between partner selection and project performance (Emden et al., 2006 ; Geringer, 1991 ).

Second, our sample covers a diverse set of family firms with variance in how the family firms handle their open innovation processes. However, our sample might still be biased toward firms that are particularly willing and able to deal with open innovation. One reason is survivor bias, meaning that family firms who are not willing and able to engage in (open) innovation might not reach the second generation (and thus not be chosen as a case in our study). A second reason is that our sampling process relied on statements referring to the firms’ innovativeness. Therefore, truly noninnovative family firms are unlikely to be part of the sample.

Third, our study focuses on family firms in Switzerland. The advantage of this geographical focus is that Switzerland is characterized by medium levels of corporatism. Consequently, firms acting in this institutional context likely attach importance to both cultural (similarity) as well as technological (complementarity) aspects when selecting strategic partners (see Vasudeva et al., 2013 ). Consequently, the model suggested in this study might not be fully generalizable to other institutional contexts. Therefore, we suggest replicating our study in countries such as the UK, which has a less distinctive corporatist culture, as well as Denmark or Sweden, which exhibit more pronounced corporative norms. Furthermore, the boundaries between the owner family and the family firm are moderately regulated in the Swiss context (Gupta & Levenburg, 2010 ). The interplay among fit, perceived control, and openness for innovation might hence be more pronounced compared with that in family firms in more strongly regulated institutional contexts. Thus, we suggest comparative studies be conducted contrasting the influence of the owner family on partner selection in strongly vs. weakly regulated contexts.

6 Conclusion

In an increasingly complex and dynamic world, organizations are forced to collaborate and open up their boundaries in order to innovate. This can be particularly challenging for family firms, which are often characterized by risk aversion. Therefore, the selection of a partner and the beginning of cooperation are important as they influence not only the current partnership but also future potential partnerships.

Change history

08 september 2022.

A Correction to this paper has been published: https://doi.org/10.1007/s11187-022-00683-w

While in strategic alliance literature in particular a number of different concepts have been put forward, including (among others) compatibility (Prashant and Harbir 2009 ), configurational fit (Swoboda et al., 2011 ), or organizational fit (Douma et al., 2000 ), no consensus has been reached regarding terminology and definitions. This is particularly true of the strategic alliance literature, but it also applies to related literature streams (De Groote et al. 2019 ).

References 

Aguinis, H., & Solarino, A. M. (2019). Transparency and replicability in qualitative research: The case of interviews with elite informants. Strategic Management Journal, 40 (8), 1291–1315. https://doi.org/10.1002/smj.3015

Article   Google Scholar  

Basco, R., & Calabrò, A. (2016). Open innovation search strategies in family and non-family SMEs: Evidence from a natural resource-based cluster in Chile. Academia Revista Latinoamericana De Administracion, 29 (3), 279–302. https://doi.org/10.1108/ARLA-07-2015-0188

Bierly, P. E., III., & Gallagher, S. (2007). Explaining alliance partner selection: Fit, trust and strategic expediency. Long Range Planning, 40 (2), 134–153. https://doi.org/10.1016/j.lrp.2007.03.001

Bogner, A., Littig, B., & Menz, W. (2009). Interviewing experts . Palgrave Macmillan. https://doi.org/10.1057/9780230244276

Book   Google Scholar  

Brinkerink, J. (2018). Broad search, deep search, and the absorptive capacity performance of family and nonfamily firm R&D. Family Business Review, 31 (3), 295–317. https://doi.org/10.1177/0894486518775187

Brinkerink, J., van Gils, A., Bammens, Y., & Carree, M. (2017). Open innovation: A literature review and recommendations for family business research. In F. W. Kellermans & F. Hoy (Eds.), The Routledge companion to family business (pp. 241–266). Routledge.

Google Scholar  

Broekaert, W., Andries, P., & Debackere, K. (2016). Innovation processes in family firms: The relevance of organizational flexibility. Small Business Economics, 47 (3), 771–785. https://doi.org/10.1007/s11187-016-9760-7

Calabrò, A., Vecchiarini, M., Gast, J., Campopiano, G., De Massis, A., & Kraus, S. (2019). Innovation in family firms: A systematic literature review and guidance for future research. International Journal of Management Reviews, 21 (3), 317–355. https://doi.org/10.1111/ijmr.12192

Cardinal, L. B., Kreutzer, M., & Miller, C. C. (2017). An aspirational view of organizational control research: Re-invigorating empirical work to better meet the challenges of 21st century organizations. Academy of Management Annals, 11 (2), 559–592. https://doi.org/10.5465/annals.2014.0086

Casprini, E., De Massis, A., Di Minin, A., Frattini, F., & Piccaluga, A. (2017). How family firms execute open innovation strategies: The Loccioni case. Journal of Knowledge Management, 21 (6), 1459–1485. https://doi.org/10.1108/JKM-11-2016-0515

Cassia, L., De Massis, A., & Pizzurno, E. (2012). Strategic innovation and new product development in family firms. International Journal of Entrepreneurial Behavior & Research, 18 (2), 198–232. https://doi.org/10.1108/13552551211204229

Chesbrough, H. W. (2003). Open Innovation: The new imperative for creating and profiting from technology . Harvard Business School Press.

Chesbrough, H. W., & Crowther, A. K. (2006). Beyond high tech: Early adopters of open innovation in other industries. R&D Management, 36 (3), 229–236. https://doi.org/10.1111/j.1467-9310.2006.00428.x

Chrisman, J. J., Chua, J. H., De Massis, A., Frattini, F., & Wright, M. (2015). The ability and willingness paradox in family firm innovation. Journal of Product Innovation Management, 32 (3), 310–318. https://doi.org/10.1111/jpim.12207

Chrisman, J. J., Chua, J. H., De Massis, A., Minola, T., & Vismara, S. (2016). Management processes and strategy execution in family firms: From “what” to “how.” Small Business Economics, 47 (3), 719–734. https://doi.org/10.1007/s11187-016-9772-3

Chua, J. H., Chrisman, J. J., & Sharma, P. (1999). Defining the family business by behavior. Entrepreneurship Theory and Practice, 23 (4), 19–39. https://doi.org/10.1177/104225879902300402

Chua, J. H., Chrisman, J. J., Steier, L. P., & Rau, S. B. (2012). Sources of heterogeneity in family firms: An introduction. Family Business Review, 36 (6), 1103–1113. https://doi.org/10.1111/j.1540-6520.2012.00540.x

Chung, S., Singh, H., & Lee, K. (2000). Complementarity, status similarity and social capital as drivers of alliance formation. Strategic Management Journal, 21 (1), 1–22. https://doi.org/10.1002/(SICI)1097-0266(200001)21:1%3c1::AID-SMJ63%3e3.0.CO;2-P

Classen, N., van Gils, A., Bammens, Y., & Carree, M. (2012). Accessing resources from innovation partners: The search breadth of family SMEs. Journal of Small Business Management, 50 (2), 191–215. https://doi.org/10.1111/j.1540-627X.2012.00350.x

Cox, J. W., & Hassard, J. (2007). Ties to the past in organization research: A comparative analysis of retrospective methods. Organization, 14 (4), 475–497. https://doi.org/10.1177/1350508407078049

Dacin, M. T., Oliver, C., & Roy, J.-P. (2007). The legitimacy of strategic alliances: An institutional perspective. Strategic Management Journal, 28 (2), 169–187. https://doi.org/10.1002/smj.577

Das, T. K., & Teng, B.-S. (2003). Partner analysis and alliance performance. Scandinavian Journal of Management, 19 (3), 279–308. https://doi.org/10.1016/S0956-5221(03)00003-4

De Groote, J. K., and Backmann, J. (2020) Initiating open innovation collaborations between incumbents and startups: How can David and Goliath get along? International Journal of Innovation Management.  https://doi.org/10.1142/S1363919620500115

De Groote, J. K., Kleindienst, I., Hoegl, M., Schweizer, D., and Laamanen, T. (2021) Similarity perceptions in investor reactions to acquisition announcements. Long Range Planning, 54 (1). https://doi.org/10.1016/j.lrp.2019.101946

De Massis, A., Kotlar, J., Chua, J. H., & Chrisman, J. J. (2014). Ability and willingness as sufficiency conditions for family-oriented particularistic behavior: Implications for theory and empirical studies. Journal of Small Business Management, 52 (2), 344–364. https://doi.org/10.1111/jsbm.12102

De Massis, A., Audretsch, D., Uhlaner, L., & Kammerlander, N. (2018). Innovation with limited resources: Management lessons from the German Mittelstand. Journal of Product Innovation Management, 35 (1), 125–146. https://doi.org/10.1111/jpim.12373

Decker, C., & Günther, C. (2017). The impact of family ownership on innovation: Evidence from the German machine tool industry. Small Business Economics, 48 (1), 199–212. https://doi.org/10.1007/s11187-016-9775-0

Diaz-Moriana, V., Hogan, T., Clinton, E., and Brophy, M. (2019). Defining family business: A closer look at definitional heterogeneity. In E. Memili, & C. Dibrell (Eds.), The Palgrave handbook of heterogeneity among family firms (pp. 333–374). New York, NY: Palgrave Macmillan. https://doi.org/10.1007/978-3-319-77676-7_13

Douma, M. U., Bilderbeek, J., Idenburg, P. J., & Looise, J. K. (2000). Strategic alliances: Managing the dynamics of fit. Long Range Planning, 33 (4), 579–598. https://doi.org/10.1016/S0024-6301(00)00062-5

Du, J., Leten, B., & Vanhaverbeke, W. (2014). Managing open innovation projects with science-based and market-based partners. Research Policy, 43 (5), 828–840. https://doi.org/10.1016/j.respol.2013.12.008

Dyer, J. H., & Singh, H. (1998). The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23 (4), 660–679. https://doi.org/10.5465/amr.1998.1255632

Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14 (4), 532–550. https://doi.org/10.2307/258557

Eisenhardt, K. M., & Graebner, M. E. (2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50 (1), 25–32. https://doi.org/10.5465/amj.2007.24160888

Emden, Z., Calantone, R. J., & Droge, C. (2006). Collaborating for new product development: Selecting the partner with maximum potential to create value. Journal of Product Innovation Management, 23 (4), 330–341. https://doi.org/10.1111/j.1540-5885.2006.00205.x

Faems, D., van Looy, B., & Debackere, K. (2005). Interorganizational collaboration and innovation: Toward a portfolio approach. Journal of Product Innovation Management, 22 (3), 238–250. https://doi.org/10.1111/j.0737-6782.2005.00120.x

Feranita, F., Kotlar, J., & De Massis, A. (2017). Collaborative innovation in family firms: Past research, current debates and agenda for future research. Journal of Family Business Strategy, 8 (3), 137–156. https://doi.org/10.1016/j.jfbs.2017.07.001

Flick, U. (2004). Triangulation in qualitative research. In U. Flick, E. von Kardoff, & I. Steinke (Eds.), A companion to qualitative research (pp. 178–183). London, England: SAGE. https://doi.org/10.1007/978-3-322-97512-6

Gehman, J., Glaser, V. L., Eisenhardt, K. M., Gioia, D., Langley, A., & Corley, K. G. (2018). Finding theory-method fit: A comparison of three qualitative approaches to theory building. Journal of Management Inquiry, 27 (3), 284–300. https://doi.org/10.1177/1056492617706029

Geringer, J. M. (1991). Strategic determinants of partner selection criteria in international joint ventures. Journal of International Business Studies, 22 (1), 41–62. https://doi.org/10.1057/palgrave.jibs.8490291

Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking qualitative rigor in inductive research: Notes on the Gioia methodology. Organizational Research Methods, 16 (1), 15–31. https://doi.org/10.1177/1094428112452151

Gjergji, R., Lazzarotti, V., Visconti, F., & García-Marco, T. (2019). Open innovation in family firms: A systematic literature review. Management Research: Journal of the Iberoamerican Academy of Management, 17 (3), 304–332. https://doi.org/10.1108/MRJIAM-03-2019-0913

Golafshani, N. (2003). Understanding reliability and validity in qualitative research. The Qualitative Report, 8 (4), 597–607.

Gupta, V., & Levenburg, N. (2010). A thematic analysis of cultural variations in family businesses: The CASE project. Family Business Review, 23 (2), 155–169. https://doi.org/10.1177/089448651002300205

Hatak, I., & Hyslop, K. (2015). Cooperation between family businesses of different size: A case study. Journal of Co-Operative Organization and Management, 3 (2), 52–59. https://doi.org/10.1016/j.jcom.2015.11.003

Hernández-Linares, R., Sarkar, S., & Cobo, M. J. (2018). Inspecting the Achilles heel: A quantitative analysis of 50 years of family business definitions. Scientometrics, 115 (2), 929–951. https://doi.org/10.1007/s11192-018-2702-1

Hitt, M. A., Dacin, M. T., Levitas, E., Arregle, J.-L., & Borza, A. (2000). Partner selection in emerging and developed market contexts: Resource-bases and organizational learning perspectives. Academy of Management Journal, 43 (3), 449–467. https://doi.org/10.5465/1556404

Horton, J., Macve, R., and Struyven, G. (2004). Qualitative research: Experiences in using semi-structured interviews. In C. Humphrey, & B. Lee (Eds.), The real life guide to accounting research (pp. 339–357). Oxford, England: Elsevier. https://doi.org/10.1016/B978-008043972-3/50022-0

Howorth, C., Rose, M., Hamilton, E., & Westhead, P. (2010). Family firm diversity and development: An introduction. International Small Business Journal, 28 (5), 437–451. https://doi.org/10.1177/0266242610373685

Kammerlander, N., and De Massis, A. (2020). Frequently asked questions in qualitative family business research and some guidelines to avoid risky paths. In A. De Massis & N. Kammerlander (Eds.) Handbook of Qualitative Research Methods for Family Business (pp. 1–23). Cheltenham, England: Edward Elgar Publishing. https://doi.org/10.4337/9781788116459.00007

Kotlar, J., De Massis, A., Frattini, F., Bianchi, M., & Fang, H. (2013). Technology acquisition in family and nonfamily firms: A longitudinal analysis of Spanish manufacturing firms. Journal of Product Innovation Management, 30 (6), 1073–1088. https://doi.org/10.1111/jpim.12046

Kotlar, J., De Massis, A., Frattini, F., & Kammerlander, N. (2020). Motivation gaps and implementation traps: The paradoxical and time-varying effects of family ownership on firm absorptive capacity. Journal of Product Innovation Management, 37 (1), 2–25. https://doi.org/10.1111/jpim.12503

Lambrechts, F., Voordeckers, W., Roijakkers, N., & Vanhaverbeke, W. (2017). Exploring open innovation in entrepreneurial private family firms in low- and medium-technology industries. Organizational Dynamics, 46 (4), 244–261. https://doi.org/10.1016/j.orgdyn.2017.05.001

Langley, A., & Abdallah, C. (2011). Templates and turns in qualitative studies of strategy and management. In D. J. Ketchen, & D. D. Bergh (Eds.), Building Methodological Bridges. Research Methodology in Strategy and Management (Vol. 6, pp. 201–235). Bingley, England: Emerald Group Publishing Limited. https://doi.org/10.1108/S1479-8387(2011)0000006007

Lazzarotti, V., & Pellegrini, L. (2015). An explorative study on family firms and open innovation breadth: Do non-family managers make the difference? European Journal of International Management, 9 (2), 179–200. https://doi.org/10.1504/EJIM.2015.067854

Lee, K. S., Lim, G. H., & Tan, S. J. (1999). Dealing with resource disadvantage: Generic strategies for SMEs. Small Business Economics, 12 (4), 299–311. https://doi.org/10.1023/A:1008085310245

Leppäaho, T., Plakoyiannaki, E., & Dimitratos, P. (2016). The case study in family business: An analysis of current research practices and recommendations. Family Business Review, 29 (2), 159–173. https://doi.org/10.1177/0894486515614157

Manotungvorapun, N., & Gerdsri, N. (2016). Complementarity vs. compatibility: What really matters for partner selection in open innovation? International Journal of Transitions and Innovation Systems, 5 (2), 122–139. https://doi.org/10.1504/IJTIS.2016.082063

Merkl-Davies, D. M., Brennan, N. M., & McLeay, S. J. (2011). Impression management and retrospective sense-making in corporate narratives. Accounting, Auditing & Accountability Journal, 24 (3), 315–344. https://doi.org/10.1108/09513571111124036

Miller, D., Le Breton-Miller, I., & Scholnick, B. (2008). Stewardship vs stagnation: An empirical comparison of small family and non-family businesses. Journal of Management Studies, 45 (1), 51–78.

Morse, J. M., Barrett, M., Mayan, M., Olson, K., & Spiers, J. (2002). Verification strategies for establishing reliability and validity in qualitative research. International Journal of Qualitative Methods, 1 (2), 13–22. https://doi.org/10.1177/160940690200100202

Neubaum, D. O., Kammerlander, N., & Brigham, K. H. (2019). Capturing family firm heterogeneity: How taxonomies and typologies can help the field move forward. Family Business Review, 32 (2), 106–130. https://doi.org/10.1177/0894486519848512

Nieto, M. J., Santamaria, L., & Fernandez, Z. (2015). Understanding the innovation behavior of family firms. Journal of Small Business Management, 53 (2), 382–399. https://doi.org/10.1111/jsbm.12075

Onwuegbuzie, A. J., & Leech, N. L. (2007). Validity and qualitative research: An oxymoron? Quality & Quantity, 41 (2), 233–249. https://doi.org/10.1007/s11135-006-9000-3

O’Reilly, M., & Parker, N. (2013). “Unsatisfactory saturation”: A critical exploration of the notion of saturated sample sizes in qualitative research. Qualitative Research, 13 (2), 190–197. https://doi.org/10.1177/1468794112446106

Parida, V., Westerberg, M., & Frishammar, J. (2012). Inbound open innovation activities in high-tech SMEs: The impact on innovation performance. Journal of Small Business Management, 50 (2), 283–309. https://doi.org/10.1111/j.1540-627X.2012.00354.x

Prashant, K., & Harbir, S. (2009). Managing strategic alliances: What do we know now, and where do we go from here? Academy of Management Perspectives, 23 (3), 45–62. https://doi.org/10.5465/amp.2009.43479263

Rabionet, S. E. (2011). How I learned to design and conduct semi-structured interviews: An ongoing and continuous journey. Qualitative Report, 16 (2), 563–566.

Rau, S. B., Schneider-Siebke, V., & Günther, C. (2019). Family firm values explaining family firm heterogeneity. Family Business Review, 32 (2), 195–215. https://doi.org/10.1177/0894486519846670

Reay, T., & Zhang, Z. (2014). Qualitative methods in family business research. In L. Melin, M. Nordqvist, & P. Sharma (Eds.), The SAGE handbook of family business (pp. 573–593). Los Angeles, CA: SAGE Publications Ltd. https://doi.org/10.4135/9781446247556.n29

Russo, M., & Cesarani, M. (2017). Strategic alliance success factors: A literature review on alliance lifecycle. Journal of Business Administration, 3 (8), 1–9. https://doi.org/10.5430/ijba.v8n3p1

Shah, R. H., & Swaminathan, V. (2008). Factors influencing partner selection in strategic alliances: The moderating role of alliance context. Strategic Management Journal, 29 (5), 471–494. https://doi.org/10.1002/smj.656

Sirmon, D. G., & Hitt, M. A. (2003). Managing resources: Linking unique resources, management, and wealth creation in family firms. Entrepreneurship Theory and Practice, 27 (4), 339–358. https://doi.org/10.1111/1540-8520.t01-1-00013

Spencer, J. W., Murtha, T. P., & Lenway, S. A. (2005). How governments matter to new industry creation. Academy of Management Review, 30 (2), 321–337. https://doi.org/10.5465/amr.2005.16387889

Swoboda, B., Meierer, M., Foscht, T., & Morschett, D. (2011). International SME alliances: The impact of alliance building and configurational fit on success. Long Range Planning, 44 (4), 271–288. https://doi.org/10.1016/j.lrp.2011.04.002

Vasudeva, G., Zaheer, A., & Hernandez, E. (2013). The embeddedness of networks: Institutions, structural holes, and innovativeness in the fuel cell industry. Organization Science, 24 (3), 645–663. https://doi.org/10.1287/orsc.1120.0780

Werner, A., Schröder, C., & Chlosta, S. (2018). Driving factors of innovation in family and non-family SMEs. Small Business Economics, 50 (1), 201–218. https://doi.org/10.1007/s11187-017-9884-4

Wray, N., Markovic, M., & Manderson, L. (2007). “Researcher saturation”: The impact of data triangulation and intensive-research practices on the researcher and qualitative research process. Qualitative Health Research, 17 (10), 1392–1402. https://doi.org/10.1177/1049732307308308

Yin, R. K. (2014). Case study research: Design and methods . Sage.

Yoon, B., & Song, B. (2014). A systematic approach of partner selection for open innovation. Industrial Management & Data Systems, 114 (7), 1068–1093. https://doi.org/10.1108/IMDS-03-2014-0086

Download references

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and affiliations.

Institute of Family Business & Mittelstand, WHU – Otto Beisheim School of Management, Vallendar, Germany

Julia K. de Groote & Nadine Kammerlander

Institute for Organization and Human Resource Management, University of Bern, Bern, Switzerland

Sabrina Schell & Andreas Hack

Institute for New Work, Bern University of Applied Sciences, Bern, Switzerland

Sabrina Schell

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Julia K. de Groote .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The original online version of this article was revised: Nadine Kammerlander is affiliated at WHU – Otto Beisheim School of Management and Andreas Hack at the University of Bern. The correct presentation of affiliations should appear as below:

J. K. de Groote (*) · N. Kammerlander Institute of Family Business & Mittelstand, WHU – Otto Beisheim School of Management, Vallendar, Germany e-mail: [email protected]

S. Schell · A. Hack Institute for Organization and Human Resource Management, University of Bern, Bern, Switzerland

S. Schell Institute for New Work, Bern University of Applied Sciences, Bern, Switzerland

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 27.5 KB)

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

de Groote, J.K., Schell, S., Kammerlander, N. et al. The role of similarity and complementarity in the selection of potential partners for open innovation projects in family firms. Small Bus Econ 60 , 1347–1367 (2023). https://doi.org/10.1007/s11187-022-00666-x

Download citation

Accepted : 27 July 2022

Published : 11 August 2022

Issue Date : April 2023

DOI : https://doi.org/10.1007/s11187-022-00666-x

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

• Open innovation
• Family firms
• Partner selection
• Complementarity

JEL Classification

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