Chapter 5 of Interpersonal Perception: The Foundation of Social Relationships and Table 5.1 on page 119 and Table 5.2 on page and 123.
The degree to which a perceiver rates a target differently from how the perceiver sees others and differently from how others see the target.
To separate error from relationship variance, the variable must be replicated across different measures or times. The variance that correlates across measure is treated as relationship and the variance that does not is treated as error.
Uniqueness accounts for 42 percent of the variance of trait ratings and 57 percent for liking. For liking it is almost always the largest SRM component and for trait ratings it often is.
Jennifer Boldry and Deborah Kashy in 1999 discussed relational outgroup homogeneity, or the tendency for more relationship variance in members of the ingroup than for members of the outgroup. They do indeed find such an effect. In addition, they also find more relationship variance for members of high-status groups than for members of low-status groups.
In 1990, Lynn Miller examined the relationship between self-reported levels of self-disclosure and attraction in a sorority with 45 members. She found a .38 correlation between the extent to which a person especially disclosed to a particular partner and unique liking of that partner (the intrapersonal correlation) and a .41 correlation between the extent to which a person disclosed to a particular partner and the extent to which that partner liked the discloser (the interpersonal correlation). (Both of these correlations are not disattenuated.)
In 2018, Maike Salazar Kämpf, Helén Liebermann, and colleagues studied 26 same-gendered groups of four to six strangers who had a 5-minute one-on-one discussion on topics of their own choice. Mimicry was measured using judgments of three trained observers who rated how much each participant mimicked his or her conversation partner. The researchers found that 55% of the variance of mimicry was at the level of the relationship and it was highly reciprocal, r = .75. They found a positive effect of Art's mimicry of Bob on Bob's attraction toward Art, but no effect on Art's attraction toward Bob. All analyses controlled for a baseline, zero-acquaintance measure of liking.
In 2017, Karen Huang and her colleagues found that during speed dating, a person’s asking questions is associated with the partner’s liking the person. The title of their paper is “It Doesn’t Hurt to Ask: Question-Asking Increases Liking.” (See also the 2018 SRM reanalysis in the study by Avraham Kluger and Thomas Malloy.) Asking questions and active listening may foster liking.
Brain Lakey, in a series of studies, has investigated perceived social support among people who were well acquainted. In 2016, Lakey, Vander Molen, Fles, and Andrews correlated perceived social support with positive affect at .54 and with negative affect at –.44. Presumably, this construct would be strongly related to interpersonal attraction. Perceived social support is conceptually related to feelings of validation, or the extent to which a person feels that the other person validates who they are.
Perceptions made at zero acquaintance are unstable at the relationship level. In three different studies of first impressions—from zero acquaintance to interacting with the target for about 10 minutes—the average stability of the relationship effect was only .27. These low stabilities of the relationship effect stand in sharp contrast to the rather strong stabilities for the perceiver effect and for the target effects.
Bernadette Park and Charles Judd in 1989 examined the ratings of people who met for an hour 4 days in a row. The day-to-day stability for the relationship effect was .76, much less than the day-to-day stability for the target effect of .98.
What about the stability of relationship effects when people know each other fairly well? Bernadette Park and colleagues in 1997 reported that the stability averaged across the Big Five traits over a period of nearly 6 months is only .28, whereas the stability of the target effect was much stronger.
Besides consistency of the relationship effect across time, it can also be examined whether the relationship effect is consistent across the Big Five. For instance, if Jane sees Tarzan as especially Extraverted, does Jane also see him as especially Agreeable? Park and Judd in 1989 examined the correlation of the relationship effect between Dominance and Friendliness on each of the 4 days, and the average correlation across the 4 days was .20. Thus a target who was rated by a given perceiver as especially high on Dominance was somewhat likely to be rated by that same perceiver as especially high on Friendliness. Overall, ratings on the two factors were not very highly correlated. The data from the 1997 study by Bernadette Park, Sue Kraus, and Carey Ryan were reanalyzed to determine the correlation of the target effects between each pair of Big Five factors at each of the three waves. For example, the correlation between the target effects for Agreeableness and Extraversion were computed for the three waves (.33, .05, and .25), and the median of the three values (.25) was used as the correlation between the two. The average of all of the 10 possible pairs of Big Five traits is only .29, a level that is generally considered a small correlation. (Note these are disattenuated correlations, so they could well be one and are, therefore, larger than the more familiar attenuated correlations.) The largest correlation, .55, is between Agreeableness and Emotional Stability factors. In 2017, Elizabeth Fles and Brian Lakey studied 16 groups, each with four service providers who knew each other well. The average correlation of the 10 Big Five relationship correlations is just .16. Their largest correlation was between Agreeableness and Openness and was .47. The correlation between Agreeableness and Emotional Stability was just .26.
In sum, though based on only three studies, relationship effects for traits are fairly differentiated and quite clearly are not unidimensional. However, the research is relatively limited, and so the conclusion remains provisional.
A major question is, What is the strength of correlation between personality judgments and attraction for each of the three SRM components? In the 1994 Interpersonal Perception book, I reported the correlation between attraction and trait ratings to be about .65 at the level of the relationship. The most detailed investigation of the link is the study by Bernadette Park and Cheryl Flink in 1989. Based on factor analyses, they did not use the Big Five factors, but rather two factors, what was referred to as the Big Two in Chapter 1. As they describe it, “the first factor consists primarily of traits relevant to extraversion or sociability. The second factor . . . includes traits relevant to conscientiousness, honesty, and intelligence” (p. 510). I refer to their first factor as Dominance and to the second one as Friendliness. Park and Flink show a contrasting pattern of correlations of these two factors with attraction. Dominance correlates strongly with the target effect for attraction, r = .75, but much weaker with the relationship effect, r = .53, whereas Friendliness correlates strongly with the relationship effect for attraction, r = .80, but much weaker with the target effect, r = .26. Elizabeth Fles and Brian Lakey found that perceived support, a construct closely related to attraction, correlates much more strongly with the Big Five factors at the level of target, r = .43, than at the level of the relationship, r = .28.
Although the database is limited, some conclusions can be drawn. First, attraction and trait judgments are correlated. However, the causal direction is unclear. It is not known whether cognitive appraisals lead to affect (you do bad things and I do not like you) or whether affect colors our cognitive appraisals (I do not like you and I think you do bad things).
The relationship effect dominates both the perceptions of targets’ personalities and the attraction of targets. As seen in Tables 5.1 and 5.2, most of the time the relationship effect explains most of the variance in trait rating and attraction judgments. In fact, for long-term attraction, relationship explains nearly twice as much variance as do perceiver and target combined. In 2016, Lakey reviewed studies not only of personality judgment but also of anxiety, perceived social support, leadership, and task performance to argue for the strength of relationship effects.
Consider four examples that further illustrate the importance of relationship effect in areas where one would think individual differences would dominate. The first example was a study by Benjamin Meagher and myself in 2013 in which we had active Christian church members make ratings of fellow group members. We were particularly interested in the extent to which a target was viewed as a "spiritual model," that is, someone whom the perceiver would want to emulate. In some sense, a spiritual model is like a living saint. One might think that target variance would dominate the ratings of this variable. Would we not all agree that Francis of Assisi and Mahatma Gandhi are saints, whereas Adolf Hitler and Joseph Stalin are sinners? However, what we found is that target explained only 27% of the total variance, but relationship effect explained 60%, more than twice as much variance. Moreover, these relationship effects were reciprocal: If Mary thought that Joseph was an exceptionally good spiritual model, Joseph tended to feel the same way about Mary.
A second example is aggression, an area of research that typically focuses on individual differences. Someone who is generally aggressive toward others is often labeled a bully, and someone who consistently receives aggression is labeled a victim. However, research by John Coie, Toon Cillessen, and colleagues in 1999 found that relationships were as important as whether an individual is a bully or a victim. Variance due to being a bully accounted for 36%, being a victim accounted for 28%, and the relationship accounted for 38%. The title of their paper makes this point very clearly: "It Takes Two to Fight."
A third example comes from coworker relationships. In her 2013 dissertation, Eliza Byington studied 874 business students who worked in 178 teams over 10 weeks as consultants to an organization. Afterward, they rated their satisfaction with their coworkers. A total of 57% of the variance was at the level of the relationship.
A fourth example also comes from coworker relationships. In 2021, Shannon Taylor, Lauren Locklear, Donald Kluemper, and Xinxin L in two studies with replications studied workplace incivility, both experienced and instigated. Individual characteristics were responsible for 31% and 24% of the variation in experienced and instigated incivility, respectively, whereas the dyad was responsible for 28% and 35%. The dyad is responsible for a considerable amount of variation in workplace incivility.
It might be asked why it is that the relationship effect dominates the other two individual-level components, but almost all theories are theories of individuals and not of relationships. Much of the variance in perceptions of others is highly idiosyncratic. Echoing Brian Lakey's call in 2016, social science needs a better understanding of why person perception is so idiosyncratic.
Boldry, J. G., & Kashy, D. A. (1999). Intergroup perception in naturally occurring groups of differential status: A social relations perspective. Journal of Personality and Social Psychology, 77, 1200–1212.
Byington, E. (2013). Exploring coworker relationships: Antecedents and dimensions of interpersonal fit, coworker satisfaction, and relational models. Unpublished doctoral dissertation, Erasmus University Rotterdam, Rotterdam, The Netherlands.
Coie, J. D., Cillessen, A. H. N., Dodge, K. A., Hubbard, J. A., Schwartz, D., Lemerise, E. A., et al. (1999). It takes two to fight: A test of relational factors and a method for assessing aggressive dyads. Developmental Psychology, 35, 1179–1188.
Fles, E., & Lakey, B. (2017). The personality characteristics of supportive providers. Personality and Individual Differences, 104, 87–91.
Huang, K., Yeomans, M., Brooks, A. W., Minson, J., & Gino, F. (2017). It doesn’t hurt to ask: Question-asking increases liking. Journal of Personality and Social Psychology, 113, 430–452.
Kluger, A. N., & Malloy, T. E. (2018). Question asking as a dyadic behavior. Journal of Personality and Social Psychology, 87, 88–104.
Lakey, B. (2016). Understanding the P × S aspect of within-person variation: A variance partitioning approach. Frontiers of Psychology, 7, 2004.
Meagher, B. R., & Kenny, D. A. (2013). Judge, that ye shall be judged: Interpersonal judgments of religious characteristics within faith communities. Psychology of Religion and Spirituality, 5, 116–128.
Miller, L. C. (1990). Intimacy and liking: Mutual influence and the role of unique relationships. Journal of Personality and Social Psychology, 59, 50–60.
Park, B., & Flink, C. (1989). A social relations analysis of agreement in liking judgments. Journal of Personality and Social Psychology, 56, 506–518.
Park, B., & Judd, C. M. (1989). Agreement on initial impressions: Differences due to perceivers, trait dimensions, and target behaviors. Journal of Personality and Social Psychology, 56, 493–505.
Park, B., Kraus, S., & Ryan, C. S. (1997). Longitudinal changes in consensus as a function of acquaintance and agreement in liking. Journal of Personality and Social Psychology, 72, 604–616.
Salazar Kämpf, M., Liebermann, H., Kerschreiter, R., Krause, S., Nestler, S., & Schmukle, S. C. (2018). Disentangling the sources of mimicry: Social relations analyses of the link between mimicry and liking. Psychological Science, 29, 131–138.
Taylor, S. G., Locklear, L. R., Kluemper, D. H., & Lu, X. (2021). Beyond targets and instigators: Examining workplace incivility in dyads and the moderating role of perceived incivility norms. Journal of Applied Psychology. Advance online publication. https://doi.org/10.1037/apl0000910
Go to the Interpersonal Perception menu.