|Home | About | Journals | Submit | Contact Us | Français|
The present research investigated whether having out-group friends serves as a buffer for feeling misunderstood in interracial interactions. Across three experience sampling studies, we found that among ethnic minorities who have few White friends or are not interacting with White friends, daily interracial interactions are associated with feeling less understood. By contrast, we found that among ethnic minorities who have more White friends or are interacting with White friends, the relationship between daily interracial interactions and feeling understood is not significant. We did not find similar results for Whites; that is, having ethnic minority friends did not play a role in the relationship between daily interracial interactions and feeling understood. Together, these studies demonstrate the beneficial effects of intergroup friendships for ethnic minorities.
People like being in situations where they feel understood and others are responsive to their needs. Felt understanding, as defined by Oishi, Krochik, and Akimoto (2010), is “feeling that one is accurately perceived, understood, appreciated, and cared for” (p. 403) by one’s interaction partner. Felt understanding is a central factor in developing and maintaining close relationships in general (Reis & Shaver, 1988) and interracial friendships in particular (Shelton, Trail, West, & Bergsieker, 2010). According to Reis and Shaver (1988), feeling understood involves a transactional process between a target and a partner. In general, the target discloses important information and feelings to the partner, and the partner responds in a way that suggests that he or she understands, appreciates, and validates the target’s feelings. The target feels understood to the extent that he or she believes his or her true self and feelings have been validated during the interaction. It has been suggested that it is more important for building intimacy in relationships that the target feels understood than it is that the partner actually understands the target (Reis & Shaver, 1988).
There are certainly contexts that decrease the likelihood that people feel understood by their interaction partner. Indeed, one can conclude from the intergroup contact literature that Whites and ethnic minorities often fail to see eye-to-eye on many things, leaving one another feeling misunderstood. To our knowledge, however, there has not been any research showing a direct link between interracial interactions and felt misunderstanding. In the present research, we examine this relationship in interactions occurring in everyday life and examine whether having intergroup friendships moderates the relationship. Specifically, we examine whether interacting with close out-group others (Studies 1–2) and having close out-group friends in one’s friendship network (Study 3) buffer people from feeling misunderstood during daily interracial interactions. In addition, in Study 3, we examine the mental health consequences of feeling misunderstood during daily interracial interactions.
Feeling misunderstood can stem from various processes going awry during interracial interactions. First, Whites and ethnic minorities focus on different behaviors during encounters with one another, making it possible for them to feel disconnected in the moment (Dovidio, Kawakami, & Gaertner, 2002). Whites focus on their own explicit, verbal behaviors toward Blacks whereas Blacks focus on Whites’ non-verbal behaviors during interracial interactions. Given that Whites’ verbal behaviors tend to be more positive than their non-verbal behaviors toward Blacks, Dovidio et al. (2002) found that Whites believed that they behaved in a friendly and non-prejudiced way during interactions with Blacks, and that the interaction was positive and productive. However, their Black partner perceived that Whites were less friendly than they thought they were, and Blacks were less satisfied with the interaction than were Whites. It is likely that Blacks did not believe Whites were responsive to their needs in the interaction.
Second, interracial interactions tend to be associated with feeling misunderstood because people make different attributions about out-group and in-group members’ behaviors, and often these attributions are more negative toward out-group members. For example, because people have trouble reading out-group members’ emotions (Elfenbein & Ambady, 2002; Gray, Mendes, & Denny-Brown, 2008), they are more prone to construe out-group (vs. in-group) members’ neutral, emotionless facial expression as hostile and threatening (Butz & Plant, 2006). As a result, a Black person who is listening with a neutral expression to what their White partner is saying, for example, might feel misunderstood because of cues from their partner that suggests he perceives the Black person as hostile and threatening. In related work on misattributions, interracial dyads who communicated over a closed-circuit television system with a subtle 1-s temporal disruption were more likely than dyads without the delay to erroneously attribute the delay to their partner being anxious, but this negative attribution did not occur in intraracial dyads (Pearson, West, Dovidio, Powers, Buck, & Henning, 2008). Together, these findings suggest that people read more into the emotions and behaviors of out-group (vs. in-group) partners. Given that people are making attributions about their partner’s emotions and behaviors that are false, their partner’s feelings and thoughts are not being understood accurately.
Finally, people may fail to recognize that their partner is being responsive to them during interracial interactions. Research on the signal amplification bias has shown that people believe that their social overtures communicate more interest to out-group partners than what is actually conveyed (Vorauer & Sakamoto, 2006). This sets the stage for misunderstandings because people may anticipate their partners reciprocating overtures that were never detected by their partner in the first place. If Blacks overestimate the level of interest and enthusiasm they are communicating, then they will expect their White partner to reciprocate with enthusiastic responses if they, too, are interested in the interaction. White partners may respond with a reciprocal response but one that is reciprocal to the level of interest actually conveyed by Blacks, not the level that Blacks believe that they are conveying. Consequently, Whites’ responses will be less enthusiastic than expected by Blacks, setting the stage for Blacks to feel not cared for, validated, or understood.
Taken together, the aforementioned research leads us to predict that daily interracial interactions will be associated with feeling misunderstood. We predict, however, that having a social network that includes out-group friends, and interacting with those friends, will moderate this relationship. Below we discuss the advantages of having intergroup friendships.
Intergroup friendships have been proclaimed as the antidote for reducing many of the negative consequences associated with interacting across racial lines. Intergroup friendships embody the essential ingredients—intimate contact, equal status, and cooperation between partners—posited for improving intergroup relations (Pettigrew, 1998). Intergroup friendship is one of the best predictors of positive intergroup attitudes (Davies, Tropp, Aron, Pettigrew, & Wright, 2011), increases people’s feelings that they can trust the out-group (Tropp, 2008), and results in people cognitively associating out-group members with oneself (Page-Gould, Mendoza-Denton, Alegre, & Siy, 2010). Thus, having out-group friends is associated with more positive feelings about the out-group as a whole, which may allow for more positive interactions with out-group members and fewer misunderstandings.
Indeed, research demonstrates that having out-group friends positively influences people’s experiences during intergroup interactions. People who have intergroup friendships experience less anxiety when interacting with out-group strangers (Paolini, Hewstone, Voci, Harwood, & Cairns, 2006), display hormonal activity associated with adaptive responses to stress (Page-Gould et al., 2010), and are more willing to turn to out-group members for support after intergroup conflict (Page-Gould, 2012). Having out-group friends may allow people to construe situations with out-group strangers differently than if they did not have out-group friends—with less anxiety and more willingness to see out-group strangers as interested in interacting—resulting in a better understanding of the situation.
While the aforementioned research suggests that intergroup friendships should protect people from feeling misunderstood during interracial interactions, one study suggests that might not occur. Specifically, Seder and Oishi (2009) found that Whites (but not ethnic minorities) who had an ethnically diverse friendship network, as measured by friends on their Facebook page, felt more misunderstood by others and were less satisfied with their lives compared with Whites who had a more ethnically homogeneous friendship network. Seder and Oishi (2009) could not offer empirical evidence to explain this finding, but they reasoned that one explanation could be the degree of closeness between the friends. Perhaps the out-group friends included in one’s Facebook network were acquaintances rather than close friends. As a result, in our research we ask participants explicitly about their friends (Studies 1 and 3) or how much they like the person in the interaction (Study 2). In general, we argue that being with and having close out-group friends will buffer the relationship between daily interracial interactions and feeling misunderstood.
One reason it is important to examine the relationship between interracial contact and felt understanding is because feeling understood has important implications for close relationships and general well-being. Feeling understood promotes closeness and intimacy among unacquainted strangers (Cross, Bacon, & Morris, 2000) and romantic partners (Laurenceau, Feldman-Barrett, & Rovine, 2005). Moreover, it facilitates psychological and physical well-being in one’s daily social interactions (Lun, Kesebir, & Oishi, 2008; Oishi et al., 2010). For example, over the course of a 2-week diary study, White and Asian Americans reported greater life satisfaction and fewer daily physical health problems (e.g., headaches, stomachaches) on days when they felt more understood by others during social interactions (Lun et al., 2008). Furthermore, feeling understood by one’s interaction partner carries over to influence physical health the next day; White and Asian Americans who felt more understood on 1 day reported fewer physical health problems the following day (Lun et al., 2008). Similarly, when White and African Americans felt that their personal self was understood by partners during daily social interactions, they also felt happy and pleasant during the interaction (Oishi, Koo, & Akimoto, 2008). Taken together, feeling understood by others is a key building block for developing healthy relationships and individuals.
Despite the attention that has been paid to misunderstandings between Whites and ethnic minorities, researchers have not made the explicit connection that perceived misunderstanding is an underlying mechanism for why interracial interactions tend to be associated with negative outcomes, such as negative mood. To our knowledge, there has been only one study that dealt with perceived (mis)understanding as an explanation for the dynamics of interracial interactions. Specifically, in a 10-week diary study on how interracial friendships are developed, Shelton et al. (2010) found that the more Whites and Blacks disclose to a potential out-group friend, the more they felt understood by that potential friend, which, in turn, created stronger feelings of closeness to the potential friend over time. It is not clear, however, how feeling (mis)understood is associated with subjective well-being and mood in interracial interactions. In Study 3, we examine these downstream consequences of feeling misunderstood in interracial interactions.
Based on previous work, we reason that interracial (vs. intraracial) interactions will be associated with feeling misunderstood, but this effect will be moderated by intergroup friendships. Specifically, we predict that the greater the number of out-group members present during an interaction or the more interracial interactions people have on a particular day, the less understood people will feel in the interaction or that day. This association, however, will be attenuated for people who have out-group friends and when they are interacting with those friends. We tested these predictions across three experience sampling studies using adolescents and young adults.
Adolescence can be a difficult stage in life. As adolescents are negotiating who they are and how they want to be perceived by others, social encounters can be ripe for misunderstandings, especially when these interactions occur across racial lines. Moreover, the stressful experience of having to negotiate feeling misunderstood with out-group members on a daily basis might cause adolescents to avoid unfamiliar diverse contexts altogether. This is unfortunate because interacting across racial lines is associated with thinking about one’s own and other’s ethnic identity (Gaither & Sommers, 2013). Therefore, being able to test our predictions with an adolescent sample is a major strength of our work, as researchers develop strategies to encourage more interactions across group boundaries. In Study 1, we conducted an experience sampling study in which adolescents in high school reported on their daily interactions for 1 week. We predict that adolescents will feel misunderstood during interracial interactions, but this relationship will be attenuated when they are interacting with out-group friends versus strangers.
Our sample included 314 adolescents (Mage = 15.10, SD = 0.50, 162 females).1 The majority of the participants reported being Asian or Asian American (37%, n = 116), with the remaining being Hispanic or Latino (26%, n = 82), White (23%, n = 72), Black, African American, or West Indian (10%, n = 31) or another race/ethnicity (e.g., American Indian, Native Hawaiian, non-specified; 4%, n = 13). Some adolescents were born outside of the United States (12%, n = 37). For these adolescents, age of immigration ranged from 6 months to 13 years, and the largest proportion emigrated from China (5%, n = 13). Most adolescents reported that they either did not know the highest level of their parent’s education (34%, n = 107) or that their parents completed high school (22%, n = 68). Data from the current study were drawn from the second wave of a longitudinal study; 10 participants were excluded from analyses due to attrition, resulting in a final sample size of 304 (68 Whites, 236 minorities) for all analyses.
We recruited adolescents for the Youth Experience Study—a 3-year longitudinal study on ethnic identity development.2 We selected students from four similar-sized and academically comparable schools, including a predominantly Asian school, a predominantly White school, a predominantly Hispanic school, and a racially heterogeneous school. “Predominantly” is defined by a single group representing at least 40% of the school’s student population. In the racially heterogeneous schools, no group represented more than 40% of the school’s population. After the school administrator agreed to participate in the study, we sent parental consent and youth assent letters home to all ninth graders in the fall of 2008 and 2009, drawing in two cohorts. See Table 1 for sample size by school and cohort. Participation for the longitudinal study began in 9th grade, but we used data for the current study from the experience sampling portion of the 10th grade because that is when we included the items relevant to the current research questions.
Although the current research was part of a larger study, here we describe the procedure and measures relevant to this article. We met with groups of 5 to 25 participants to describe the study. We gave each participant a cellular phone to complete experience sampling reports for the next 7 days by accessing a web browser. We randomly prompted participants five times per day after school at times unknown to them for a week for a total of 35 prompts. On average, participants completed 24 (range from 1 to 35) surveys over the course of the week.
On the cellular phone at every prompt, participants responded to 40 questions that varied in content, including items that did not make reference to interactions, ethnicity of interaction partners, or emotions (e.g., What activity are you engaged in?). Means and standard deviations for the items used in this study are reported using the person-level average of the 35 experience sampling measures.
Participants were first prompted to “Look around and think about the people in your current setting” and then asked to respond to the question “How many people are you with right now?” Responses included “I am alone (0),” and then ranged from 1 to 5+ (M = 1.75, SD = 0.93). If participants indicated that they were alone, they were routed to questions not relevant to the present research. Participants indicated that they were alone an average of 23% of the time over the course of the week. If participants indicated that they were not alone, they responded to a series of questions about who they were interacting with and how they felt at that time.
To assess the amount of contact with Whites in a given situation, participants responded to the question, “How many of the people who you are interacting with are White?” (M = 1.21, SD = 0.33). To assess the amount of contact with minorities in a given situation, participants responded to the question “How many of the people who you are interacting with are members of a racial/ethnic minority group that is not your own? (M = 0.86, SD = 0.80). Responses for both questions included 0 = “none: 0%,” 1 = “about 25%,” 2 = “about 50%,” 3 = “about 75%,” 4 = “all: 100%.” The intraclass correlation (ICC) for interacting with Whites was .71, indicating that a significant proportion of the variance in contact with Whites (71%) was due to the person, p < .001. The ICC for interacting with minorities was .31, indicating that a significant proportion of the variance in contact with minorities (31%) was due to the person, p < .001.
To assess how misunderstood adolescents felt, participants responded to the question “How much do you feel misunderstood by the people you are interacting with?” (M = 2.62, SD = 0.62). Responses ranged from 0 = “not at all” to 6 = “a lot.” The ICC was .37, indicating that a significant proportion of the variance in perceptions of being misunderstood (37%) was due to the person, p < .001.
To assess whether participants were with friends or not, participants responded to the question “Are you with friends?” (M = 0.27, SD = 0.26). Responses included “no” (0) or “yes” (1). The ICC was .24, indicating that a significant proportion of the variance in the presence of friends (24%) was due to the person, p < .001.
Because variability in the situation-level variables were at least partially accounted for by person-level differences, we used multilevel modeling in MPlus 7.1 (Muthen & Muthen, 1998–2012) to model the nested nature of the data (situations at Level 1 within individuals at Level 2), with all of the relationships of interest at Level 1. To control for potential artifacts of the repeated measures nature of the study, we controlled for the order of the random prompts (1–35) at Level 1 in all analyses. In addition, we controlled for gender and school of recruitment at Level 2 in all analyses. School of recruitment was controlled for using three dummy variables, with the predominantly White school serving as the comparison.
We included the variance in the participants’ intercepts and time slope as well as the covariance between slopes and intercepts as random effects. Both minority and White participants varied significantly in their intercepts of perceived misunderstanding ( , p < .001, and , p < .001, respectively).
As predicted, results indicated that for minority adolescents, the more White people present in the interaction the more minorities felt misunderstood (b = .10, SE = 0.03, p = .002). There was no main effect for the presence of friends on feeling misunderstood (b = −.06, SE = 0.07, p = .46), feeling misunderstood did not change over time (b = .001, SE = 0.003, p = .70), and there was no significant effect for school of recruitment (b = .06, SE = 0.23, p = .80; b = .00, SE = 0.22, p = .99; b = .12, SE = 0.21, p = .59). As predicted, however, results revealed a significant two-way interaction between interracial contact and having friends present in the interaction, b = −.12, SE = 0.05, p = .01. Additional analyses revealed that when minority adolescents interacted with people who were not friends, the more White people present the more minorities reported feeling misunderstood (b = .15, SE = 0.04, p < .001), whereas when minority adolescents interacted with friends, there was no relationship between Whites being present and feeling misunderstood (b = .03, SE = 0.04, p = .52; Figure 1). We also analyzed the data when friends were present or not in the interactions that were homogeneous (e.g., when none or all of the other people in the interaction were White). When none (0%) of the people in the interaction was White, the presence of friends or not was unrelated to ethnic minorities feeling misunderstood (b = −.02, SD = 0.08, p = .78). When all of the people (100%) in the interaction were White, however, minority adolescents felt less misunderstood in the presence of friends (b = −.43, SD = 0.15, p = .01).
As predicted, for White adolescents, the more minorities in the interaction the more Whites felt misunderstood (b = .17, SE = 0.08, p = .03). There was no main effect for the presence of friends on feeling misunderstood (b = −.04, SE = 0.16, p = .82), feeling misunderstood did not change over time (b = .007, SE = 0.005, p = .22).3 In contrast to the results for minorities, the two-way interaction between interracial contact and having friends present in the situation was not significant, (b = .18, SE = 0.11, p = .17). Therefore, for White adolescents, interacting with minorities was related to feeling misunderstood regardless of whether the minorities present were friends or not.
Friendships are important for the healthy development of adolescents. Friendships provide a context for social, emotional, and cognitive growth, including fostering a sense of well-being and learning to regulate one’s emotions (Hartup & Stevens, 1997; Newcomb & Bagwell, 1995). Consistent with the notion that friendships are advantageous, our findings offer initial support that for minority adolescents, friendships with Whites mitigate the stress associated with interracial interactions. We found that for minority adolescents, the presence of White friends buffers the relationship between interacting with Whites and feeling misunderstood. Given that we only asked participants whether they were with friends or not, there could have been in-group friends present as well. However, when we examined instances in which all interaction partners were White, the presence of friends still mitigated feelings of misunderstanding, whereas instances in which none of the interaction partners was White, the presence of friends did not make a difference for feelings of misunderstanding. Together, these results suggest that for minorities, intergroup friendships, rather than just friendships in general, are particularly meaningful for felt understanding.
For White adolescents, the more minorities present in the interaction the more they felt misunderstood. Unlike the findings for minorities, however, for White adolescent, intergroup friendship was not a protective factor. We return to this null finding in “General Discussion” section.
In Study 2, we address several limitations of the methodological design of Study 1. First, we focus on one interaction partner—participants’ roommate—so that the partner’s race/ethnicity could be known. In addition, in Study 2, we focus on how much participants liked their partners rather than simply asking them to indicate whether they were friends or not, acknowledging that in an interracial context, people may refer to someone as a friend, but the person is merely an acquaintance, which might be an explanation for the null findings for Whites in Study 1.
Interracial roommate relationships have become the focus of attention for studying the antecedents and consequences of intergroup contact. Research on interracial roommate relationships has produced mixed results; these relationships are often categorized by negative short-term outcomes but have long-term benefits. For example, with respect to short-term or in the moment outcomes, interracial (vs. intraracial) roommate dyads feel less close to one another, experience more anxiety, prefer not to live with one another again, and are generally less satisfied about the relationship to the point that they are more willing to ask authorities to move to a new room (e.g., Shook & Fazio, 2008; Towles-Schwen & Fazio, 2006). However, these same relationships improve racial attitudes in the long term (van Laar, Levin, Sinclair, & Sidanius, 2005), increase the number of out-group friends people have over time (Mark & Harris, 2012), and improve interracial interactions with strangers in subsequent diverse settings (Gaither & Sommers, 2013). We examine whether being friends with one’s roommate will alleviate short-term negative consequences—feeling misunderstood—of living with an out-group roommate.
In Study 2, we conducted an 8-week diary study with students who had been randomly assigned to live with an in-group or out-group roommate. We predict that participants will feel more misunderstood with an out-group (vs. in-group) roommate, and that liking (a proxy for friendship) one’s roommate will moderate this relationship. Liking one’s roommate, and thus feeling as if you are friends, should matter more for felt understanding among interracial roommate dyads because these dyads are more likely to struggle with differences that often create barriers and obstacles in interactions.
Seventy-two same-sex undergraduate roommate dyads (45 female dyads) participated in the study for $50. Participants had been randomly assigned to be roommates by the university. The sample included 54 ethnic minority and 90 White participants, forming 38 same-race dyads (10 minority–minority, 28 White–White) and 34 cross-race roommate dyads. The ethnic minority roommates in the same-race dyads were members of the same ethnic group. There was little missing data (0.87% for roommate understanding, 10 out of 1,152 possible observations), and a total of 1,142 observations were used in the final analyses.
We recruited roommate pairs at the beginning of the academic year to participate in the Weekly Roommate Relationship Study. We did not inform participants of the intergroup nature of the study. Participants attended an orientation session to complete several pre-diary measures and received instructions on how to complete the weekly diary entries. At the end of the week for the next 8 weeks participants received an online questionnaire about their roommate interactions that week. After 8 weeks, participants received compensation and were debriefed.
Participants indicated how much they felt their roommate understood them that week using a one-item measure on a 7-point scale from strongly disagree (1) to strongly agree (7; “During the past week, my roommate understood who I really am.”). Similarly, participants indicated how much they liked their roommate that week (“During the past week, how much did you like your roommate?”) using a 7-point scale (1 = not at all, 7 = very much). We used the liking measure as a proxy for friendship.
To adjust for the non-independence due to roommates and the fact that we measured the same people over time, we used multilevel modeling. In addition to the correlation between dyad members’ residuals, the random effects also included the variance in participants’ intercepts, variance in participants’ time slopes, the within-person covariance of intercepts and slopes, the between-person covariance of intercepts, the between-person covariance of slopes, and the between-person covariance of intercepts and slopes. Given that we are not interested in the change in feeling understood over time, we included time in the model as a main effect only.4 In addition to time, the model included the fixed effects of participant’s race (coded −1 for minority and 1 for White), roommate’s race (coded −1 for minority and 1 for White), liking (grand-mean centered), the three-way interaction of participant’s race, roommate’s race and liking, and all lower order interactions.
There was no significant main effect of participant’s race, b = 0.11, SE = 0.09, p = .24; no main effect for roommate’s race, b = −0.03, SE = 0.09, p = .73; and no two-way interaction of participant’s race and roommate’s race, b = −0.07, SE = 0.10, p = .47 on understanding. There was a significant increase in perceived understanding over time, b = 0.08, SE = 0.01, p < .001. There was a significant main effect of liking such that the more participants liked their roommate the more they felt understood, b = 0.45, SE = 0.03, p < .001. There was no significant two-way interaction of participant race and liking, b = 0.03, SE = 0.03, p = .45; and no significant two-way interaction of roommate race and liking, b = 0.03, SE = 0.03, p = .33, on understanding.
More important, analyses revealed a significant three-way interaction of participant’s race, roommate’s race, and liking on understanding, b = −0.08, SE = 0.03, p = .02. Similar to Study 1, for Whites, the two-way interaction of roommate’s race and liking on understanding was not significant, b = −0.05, SE = 0.04, p = .29. As predicted, for minorities, however, there was a significant two-way interaction of roommate’s race and liking, b = 0.11, SE = 0.05, p = .04. Further analyses revealed that for minorities with White roommates, the relationship between liking and understanding was significant and large, b = 0.53, SE = 0.07, p < .001, and for minorities with same-race roommates, this relationship was smaller, b = 0.31, SE = 0.08, p < .001 (Figure 2).
Study 2 shows that the more minority college students liked their White roommates the more they felt understood during weekly interactions with them. We found a similar pattern for minorities with a minority roommate, but the relationship was weaker. This finding is particularly interesting in the broader context of previous research on interracial roommate relationships that has shown that students in interracial roommate dyads (vs. intraracial) are generally incompatible and experience more stress during their interactions. We found that being friends—as measured by high levels of liking one’s roommate—is associated with feeling understood for minorities in interracial roommate dyads. Our findings suggest that one way to alleviate some of the stress experienced by randomly assigned interracial roommate dyads is to foster friendship between them. Consistent with Study 1, this relationship was not significant for Whites, which we return to in the “General Discussion” section.
In Study 3, we addressed several limitations in our previous studies. First, we focused on minorities’ social network more generally rather than on the relationship of the person in the interaction. This allows us to examine whether having out-group friends has benefits on minorities’ interactions with out-group strangers. Indeed, it has been shown that people who have intergroup friendships experience less anxiety when interacting with out-group strangers (Paolini et al., 2006). The positive experiences with out-group friends may spill over to interactions with strangers; having out-group friends may allow people to construe ambiguous situations with out-group strangers in a more positive way. We also asked participants to report how many of their close friends were White rather than using a measure of liking as a way to operationalize friendship. These changes allow us to make a more compelling claim that intergroup friendship is a resource that is related to feeling understood during interracial interactions.
Second, we used a multiple item composite of understanding rather than a single-item measure. Although single-item measures of felt understanding have been used in similar experience sampling studies (Lun et al., 2008), in Study 3, we added additional items that better approximate the definition of felt understanding. Consistent with the measure used by Seder and Oishi (2009), participants indicated how much they felt validated, appreciated, and understood by their interaction partners. Finally, in Study 3, we consider how feeling misunderstood in interracial interactions is related to daily well-being and mood.
Based on previous research, we predict that interracial interactions will be associated with feeling misunderstood, lower positive mood, lower well-being, and higher negative mood. These relationships, however, will be attenuated by having out-group friends. In addition, we predict that feeling misunderstood will mediate the relationship between interacting with Whites and daily well-being and mood. In other words, one reason interacting with White strangers is associated with negative mood, for example, is because minorities do not feel understood during these interactions.
The sample consists of 129 ethnic minority participants (n = 45 Asian or Asian American; n = 24 Hispanic or Latino; n = 40 Black, African American, or West Indian; n = 3 Other; n = 13 multiracial; n = 4 did not indicate a race). We excluded participants who indicated their race as other, multiracial, or those missing race information as well as nine participants who did not have data for the critical understanding variable. These exclusions resulted in 100 participants (64 females, 25 males, and 11 missing gender scores) in the analyses.
We recruited participants from two universities (n1 = 90 and n2 = 10) to participate in the Identity and Social Interactions Study.5 Participants attended an orientation session where they completed a background questionnaire and received instructions about the remaining part of the study. For the next 21 days, participants received an online questionnaire to complete at the end of each day. At the end of the 21 days, participants completed a final questionnaire, received compensation ($50), and were debriefed. Given the difficulty of recruiting minority participants, we collected data in two rounds (i.e., completed one 21-day period, n1 = 70 and n2 = 30, waited a few weeks and then recruited the remaining participants).
Participants completed a variety of measures in the pre-diary and diary questionnaires that are not relevant to the present work; therefore, we do not describe them here. Below are the measures that are pertinent to our research questions.
Participants answered the question, “how many of your closest friends are White/European American?,” using a response scale of 1 = none, 2 = few, 3 = many, 4 = most, and 5 = all (M = 2.95, SD = 0.91). A few of the participants indicated that they had zero (4%) or all (2%) White friends. The rest of the participants indicated they had a few (26%) White friends or that many (34%) or most (25%) of their friends were White.
Participants answered two questions about their contact with others: “How frequently did you interact with White people today?” (M = 3.39, SD = 1.12), and “how frequently did you interact with members of your own racial/ethnic group today?” (M = 3.07, SD = 1.15), using 5-point scales (1 = not at all, 5 = very frequently). We computed a difference score between these items (M = 0.32, SD = 1.86), capturing the extent to which participants were interacting with Whites more than with members of their own group, and thus it works as a control for effects due to simply interacting with more people throughout the day.6
The number of White friends participants had was positively correlated with frequency of daily contact with Whites, r = .41, p < .01, and negatively correlated with daily contact with their own racial/ethnic group, r = −.47, p < .01. Although minorities are interacting with their White friends throughout the day, they are also interacting with Whites who are not their friends.
Participants completed three items to assess how understood they felt by their interaction partners that day using a 7-point Likert-type scale (1 = not at all, 7 = very much so). The specific items are as follows: “During your interactions today, to what extent did you “… understood by others?, “… feel validated by others?,” and “… appreciated by others?” (α = .86). The mean was 4.91 (SD = 1.20), and the ICC was .49, p < .001.
Daily well-being was assessed with two items on 7-point scales. The first item, “how was today?,” ranged from 1 = horrible to 7 = excellent, and the second item, “how satisfied are you with your life today?,” ranged from 1 = extremely dissatisfied to 7 = extremely satisfied. A composite of the items made a daily well-being measure (α = .90). The mean was 5.11 (SD = 1.11), and the ICC was .29, p < .001.
Participants indicated how happy, calm, joyful, and excited they felt each day, and these items were combined for a positive mood composite (α = .77, M = 4.19, SD = 1.24). Similarly, participants indicated how sad, nervous, anxious, discouraged, on edge, hopeless, blue, disappointed, dissatisfied, disillusioned, frustrated, and defeated they felt each day, and these items were combined for a negative mood composite (α = .94, M = 2.08, SD = 1.15). All items were rated on 7-point scales (1 = not at all, 7 = extremely). The ICC was .39 for positive mood and .51 for negative mood (ps < .001)
We used multilevel modeling to test the hypothesis that having White friends buffers minorities from the negative effects of contact with Whites on felt understanding. Participants were modeled at Level 2 and day at Level 1. Control variables were included in the analysis as fixed effects, most at Level 2 (i.e., participant’s gender, year in school, school, and the round of data collection) but also a Level 1 variable that captures linear changes in felt understanding across the 21 days (i.e., time). Time was included only as a main effect (no interactions with time) because change overtime in felt understanding was not our primary interest—our hypotheses are about how events on a particular day influence perceptions on that same day. These variables had no statistically significant relationship with understanding and will not be discussed further. In addition to the control variables, we included contact with Whites (Level 1), grand-mean centered White friends (Level 2), and the cross-level interaction of contact with Whites and White friends.
We included the variance in the participants’ intercepts and time slope as well as the covariance between slopes and intercepts as random effects. Participants varied significantly in their intercepts of felt understanding, , p < .01; and there was significant variance in the amount of change in felt understanding people experienced overtime, , p < .01. There was no covariance between intercepts and slopes, σint,time=0.01, p = .06, suggesting that initial levels of felt understanding were marginally related to the amount of change in felt understanding over time. Next, we report the results of the fixed effects that test our main hypothesis.
For the fixed effects, there was a statistically significant main effect of contact with Whites on perceptions of felt understanding, b = −0.03, SE = 0.02, p = .04. There was no statistically significant main effect of White friends on felt understanding, b = 0.01, SE = 0.10, p = .91. We hypothesized that the negative relationship between contact with Whites and felt understanding would be reduced (be closer to zero) the more White friends the minority participants reported having. The interaction of contact with Whites and White friends tests this hypothesis. As expected, there was a significant two-way interaction between contact with Whites and White friends on understanding, b = 0.05, SE = 0.02, p = .01 (see Figure 3). Additional analyses revealed that when participants have a low number of White friends (1 SD below the mean), more contact with Whites is associated with lower levels of felt understanding in daily interactions, b = −0.07, SE = 0.02, p < .01, whereas when participants have more White friends (1 SD above the mean), contact with Whites is not associated with felt understanding, b = 0.01, SE = 0.02, p = .69.
Next, we used multilevel modeling, with the same model as described above, to estimate the relationship between contact with Whites and having White friends on daily well-being. There was no significant main effect of having White friends on well-being, b = 0.05, SE = 0.08, p = .55. As expected, there was a significant effect of contact with Whites on well-being such that more contact with Whites was related to lower daily well-being, b = −0.06, SE = 0.02, p < .01. The number of White friends did not moderate this effect, b = 0.01, SE = 0.02, p = .57.
The same model was also used to examine mood. For negative mood, contact with Whites was significantly associated with increased daily negative mood, b = 0.04, SE = 0.01, p = .02. There was no main effect of number of White friends, b = −0.08, SE = 0.10, p = .41, and no interaction of number of White friends and contact with Whites, b = 0.02, SE = 0.02, p = .30.
For positive mood, there was no significant main effect of number of White friends, b = −0.01, SE = 0.10, p = .93, but more contact with Whites was significantly associated with lower levels of daily positive mood, b = −0.07, SE = 0.02, p < .01. There was a marginally significant two-way interaction of contact and White friends, b = 0.03, SE = 0.02, p = .09. When minorities have a low number of White friends, more contact with Whites was associated with lower levels of daily positive mood, b = −0.10, SE = 0.03, p < .01; however, when minorities have more White friends, the relationship between contact with Whites and positive mood was smaller and marginally significant, b = −0.04, SE = 0.02, p = .10.
Given that intergroup friendship did not moderate the relationships between contact with Whites and well-being and mood (the interaction was only marginally significant for positive mood), we tested whether felt understanding mediates the relationships between contact with Whites and well-being, positive mood and negative mood controlling for intergroup friendship. Essentially, this examines whether the reason why daily interracial interactions is associated with lower well-being, higher negative mood, and lower positive mood is because minorities feel misunderstood.
First, we examined misunderstanding as a mediator of the relationship between daily interracial contact and well-being. There was a significant overall effect of contact with Whites on felt understanding, b = −0.03, SE = 0.02, p < .05. Furthermore, the effect of felt understanding on well-being was significant, b = 0.49, SE = 0.02, p < .01. Because the direct effect of contact on well-being remains significant, b = −0.05, SE = 0.02, p < .01, this is evidence of partial mediation if the indirect effect is significant. We also tested whether the indirect effect was statistically different from zero by estimating its 95% confidence interval (CI) using the Monte Carlo Method for Assessing Mediation (MCMAM; Selig & Preacher, 2008). The indirect effect of contact with Whites on well-being through felt understanding was significantly different from zero based on the 95% CI = [−0.03, <0.00]. In sum, there is evidence that 27.8% of the effect of contact with Whites on well-being is mediated by felt understanding.
Next, similar mediation analyses were conducted for negative mood. The effect of felt understanding on negative mood was significant, b = −0.26, SE = 0.02, p < .01, with a significant direct effect of contact, b = 0.03, SE = 0.01, p = .04. This indirect effect was also significant, CI = [>0.00, 0.02], indicating that 19.2% of the effect of contact with Whites on negative mood is mediated by felt understanding.
Finally, similar mediation analyses were conducted for positive mood. The effect of felt understanding on positive mood was significant, b = 0.46, SE = 0.02, p < .01, with a significant direct effect of contact, b = −0.05, SE = 0.02, p < .01. The indirect effect of contact on positive mood through felt understanding was significant, CI = [−0.03, <0.00]—evidence that 20.3% of the contact effect on positive mood is mediated by felt understanding.
Consistent with our previous studies, the findings from Study 3 show that intergroup friendships protect minorities from feeling misunderstood during daily interracial interactions. This work shows that positive intergroup experiences in one context, in this case with friends, can generalize to intergroup experiences with strangers in a different context (e.g., Paolini et al., 2006). However, intergroup friendships do not protect minorities from experiencing negative well-being, more negative mood, and less positive mood. We discuss these findings in more detail below.
Across three experience sampling studies using adolescents and young adults, we found that intergroup friendships play a positive role in interracial interactions for minorities in their everyday life. Although minorities feel misunderstood during interracial interactions, if that White interaction partner is a friend or if minorities have White friends in their social network, they are less likely to feel misunderstood. Whites, however, did not benefit from having intergroup friendships in the same manner; instead, interracial contact was associated with more misunderstanding regardless of whether Whites were interacting with out-group friends.
When people feel that their partners understand them and are responsive to their needs, trust and relationship satisfaction increase (Reis & Shaver, 1988). If they do not feel this is the case, they may eventually avoid interactions with that person and perhaps with people who resemble that person. In an intergroup context, this may mean avoiding interactions with out-group members in general and may cause some minorities to be especially wary of forming friendships with out-group members (Mendoza-Denton, Downey, Purdie, Davis, & Pietrzak, 2002). Reducing the amount of intergroup contact may, in turn, foster (or at least not reduce) negative beliefs about the out-group. As a result, getting a handle on how perceived understanding influences interracial interactions is likely to be useful for improving intergroup relations in the long run.
The present research contributes to a growing body of research revealing the benefits of intergroup friendships. Intergroup friendship alleviates the stress that is often associated with interacting across racial lines. Whereas the majority of research on intergroup friendship has focused on the benefits for Whites, our work shows the benefits for minorities; in fact, the findings are not significant for Whites. These effects are particularly interesting given the potential benefits of intergroup friendships for minority students in predominantly White institutions—a group highly likely to experience alienation and discrimination (Mendoza-Denton & Page-Gould, 2008). Friendship is a reciprocal relationship that provides a safe context for people to learn, expand the self, and practice important skills, such as perspective taking (Hartup & Stevens, 1997). When friendships cross racial lines, these interpersonal skills may directly help a person navigate and understand the world and experiences of another group. It may also mean realizing that not all out-group members hold negative racial beliefs and learning to appreciate that challenges in interracial interactions may not be a reflection of the other person’s prejudice but rather are challenges that occur in all types of interpersonal interactions.
In our work, we focused on intergroup friendships rather than on the amount of intergroup contact. We did so for two reasons. First, given that two of our studies were conducted at predominantly White institutions, we reasoned that our minority participants were likely to have had a lot of contact with Whites, leaving little variability for our models. Second, and more importantly, the positive effects from mere contact with out-group members may not be enough for minorities, considering their status and experiences in our society. When one is a member of a devalued group, having many superficial interactions with dominant groups may only serve to enhance perceived intergroup differences and may even reinforce minorities’ concerns about not feeling appreciated, valued, and that their perspective is understood. McLaughlin-Volpe, Mendoza-Denton, and Shelton (2005), for example, found that superficial interactions with large numbers of out-group acquaintances were not related to attitude change for minorities; in fact, ethnic minority students with more superficial contact had more negative feelings toward the out-group. Therefore, we would expect that positive consequences of intergroup contact—in our work, reduced misunderstandings—to be limited to minorities who have a close and supportive relationship with a White friend.
Although our work shows one benefit of having out-group friends for minorities, other research has shown unintended negative consequences of positive intergroup contact—the kind that would occur among friends—for minorities. Specifically, positive contact with out-group members leads minorities to be less likely to perceive discrimination against their group and challenge social inequalities (Dixon, Tropp, Durrheim, & Tredoux, 2010; Saguy, Tausch, Dovidio, & Pratto, 2009; Tropp, Hawi, Van Laar, & Levin, 2012). Positive contact with the out-group causes minorities to see out-group members as more on their side and not as oppositional to their group. As a result, minorities are less inclined to think they could be the target of discrimination and that there is a need to engage in collective action. These findings suggest that intergroup friendship may not always be the antidote for improving race relations. Indeed, our own findings revealed that although having out-group friends helps minorities feel understood by out-group strangers, it is not related to their general well-being and affect. Although it is not clear why this is the case, our findings contribute to the growing body of literature showing that intergroup friendships are not the antidote for solving all problems associated with interracial interactions.
There are several limitations in the present research that need to be addressed in future work. First, the correlational nature of experience sampling studies makes it difficult to determine causal relationships. There could be variables not included in the models that are driving the relationships we found. As a result, future experimental work is necessary. Second, although we found that interracial contact was associated with feeling misunderstood for Whites, this relationship was not buffered by having out-group friends. One possible explanation is that, despite our attempt to have participants focus on close out-group friends, the friends Whites selected may not have been considered their close friends. Whites and minorities may have different ideas about what is considered an out-group friend. Given concerns about appearing prejudiced, Whites might claim a Black person as a friend but that Black friend is not considered in the same light as a White friend. As such, Whites might have indicated that the person in the interaction was a friend (Study 1) or that they were close to their roommate (Study 2), but in reality, they were not. Minorities, however, may feel less obligated to make claims about having White friends so their report of friends were more accurate, and, thus, the benefits of having out-group friends materialized in our studies. Future work is needed to address this issue. Third, although we have established that intergroup friendship protects minorities from feeling misunderstood in interracial interactions, we do not know the underlying mechanisms for this relationship. One possibility is that because people who have intergroup friendships consider the out-groups as part of the self (Page-Gould et al., 2010), they afford the person the benefit of the doubt when difficulties or miscommunications occur in interactions with an out-group stranger; the out-group stranger is now perceived as the self. Finally, we examined participants’ perceived understanding, but we did not examine the extent to which their partners felt as if they understood them. We did so because it has been suggested that perceived understanding is more important for building intimacy in relationships than partners actually understanding (Reis & Shaver, 1988). Nonetheless, people may not accurately perceive how much their partners understand them. The extent to which people underestimate, overestimate, or are accurate about this could be influential interracial interactions. As a result, it would be worthwhile to examine this empathic accuracy in interracial interactions.
Interpersonal interactions can be challenging. People present themselves to their partners and hope that their partners respond in a way that suggests that they are understood—accurately perceived, cared for, and appreciated. Unfortunately, when interpersonal interactions occur across racial lines, feeling understood by one’s partner does not always occur, opening the door not only for psychological and relationship problems but also for potentially intergroup problems. The present research shows that for minorities, having close intergroup friendships can attenuate these problems, adding to the growing body of work on the power of intergroup friendships.
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was supported by grants from the National Institute of Child and Human Development (5R01HD055436-02) and National Science Foundation (BCS-0742390).
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
1We excluded multiethnics (n = 55) because it is difficult to define intergroup contact for these individuals.
2Findings from this data set have been published (Yip, Douglass, & Shelton, 2013), but there is no overlap with this article.
3Unexpectedly, we found a significant difference in misunderstanding by schools such that Whites in the predominantly Latino school felt lower levels of misunderstanding than Whites in the predominantly White (b = −1.81, SE = 0.23, p < .001), predominantly Asian (b = −1.61, SE = 0.38, p < .001), and heterogeneous school (b = −1.72, SE = 0.23, p < .001).
4Models that estimate the change overtime in the outcome, the random variance in the time slope, and the covariance between slopes and intercepts are technically referred to as growth curve models; however, because we are not interested in how change in understanding overtime is moderated by our study variables, we do not include the interactions with time. The random effects of time were retained due to the significant variance in the slopes of time—not including random effects that are present can bias the SEs of the fixed effects.
5Findings from this data set are in an unpublished paper (Trujillo, Garcia, & Shelton, 2014), but there is no overlap with this article.
6In a model with daily contact with Whites predicting understanding controlling for average daily contact with minorities, we found a negative but non-significant effect of daily contact with Whites, b = −0.04, SE = 0.03, p = .186. In a model with daily contact with one’s own group predicting understanding controlling for average daily contact with Whites, we found a positive but non-significant effect of daily contact with one’s own group, b = 0.05, SE = 0.03, p = .117. Because the size of these effects are relatively equal, the effects of the difference score reported below are not driven by either one of these components of the differences score on their own.