2014 SITAR Wiggins Award Winner - Kelly McDonald
Socially Anxious Behavior Disrupts Patterns of Interpersonal Calibration
Kelly McDonald & Pamela Sadler
Wilfrid Laurier University
Erik Woody
University of Waterloo
Abstract: Because neuroticism is not represented on the Interpersonal Circumplex, instances of it in social interactions, such as socially anxious behavior, might be presumed to be irrelevant to complementarity. In contrast, the present work evaluates the possibility that complementarity is impacted if one member of the dyad shows socially anxious behavior. A total of 95 participants interacted with a confederate who displayed high or low levels of agreeableness and high or low levels of socially anxious behavior. Participants indicated the types of interpersonal responses they had towards the confederate, and independent raters coded the behavioral responses participants actually displayed. Although the confederate's socially anxious behaviors did not impact the extent to which reciprocity occurred on the dominance dimension, they did have a substantial impact on the extent to which correspondence occurred on the affiliation dimension. Participants' affiliative reactions were significantly less correspondent when the confederates displayed socially anxious behaviors than when they did not. Interestingly, participants responded with relatively high levels of affiliation towards a socially anxious confederate regardless of whether the confederate was agreeable or disagreeable. This pattern of results suggests both important qualifications to the principle of interpersonal complementarity and possible merits of incorporating a neuroticism dimension in conceptions of interpersonal style.
The interpersonal circumplex (IPC) dimensions of dominance and affiliation relate to other theories of personality, such as the Big Five (Costa & Widiger, 1994; Pincus & Gurtman, 1995; Wiggins & Pincus, 1994). More specifically, dominance and affiliation can be regarded as rotational variants of the two Big Five factors of agreeableness and extraversion. Because the IPC is two-dimensional, and the Big Five are theorized to be orthogonal to each other, it seems that there is no room on the IPC for the remaining Big Five traits of conscientiousness, openness, and neuroticism. Although the theoretical structures of the Big Five Factor model and the IPC seem to imply that the remaining Big Five traits do not have an interpersonal role, conceptually it seems that these traits should influence individuals’ interpersonal relations. In the current work, we focused on how the Big Five trait of neuroticism, exhibited through socially anxious behavior, and the Big Five trait of agreeableness impacts interpersonal complementarity.
There is a substantial amount of literature to support the idea that neuroticism is an interpersonally relevant trait. In relationships, individuals who are high in trait neuroticism experience higher degrees of conflict, higher dissolution rates, and lower levels of support (Ozer & Benet-Martinez, 2006; Widiger, 2009). Research on the impact of social anxiety, one type of neurotic behavior, also lends support to the idea that neuroticism plays a significant role in interpersonally relevant situations. For example, people with social anxiety develop fewer social relationships, and when they do develop social relationships, these relationships are less intimate, functional, and satisfying than relationships of people without social anxiety (Alden & Taylor, 2004). Even in brief interactions between unacquainted dyads, others are less likely to desire future interactions with socially anxious individuals (Meleshko & Alden, 1993).
If neuroticism is indeed an interpersonally relevant trait, as the literature suggests, then instances of it, such as social anxiety, should influence interpersonal processes. In the current research, we examined whether anxious behavior impacts the interpersonal process of complementarity, defined by oppositeness (or reciprocity) on the dominance dimension and sameness (or correspondence) on the affiliation dimension (Kiesler, 1996).
To investigate this idea, we examined whether individuals responded differently toward agreeable and disagreeable behavior when accompanied by a presence or absence of anxiety. In other words, does anxious behavior moderate the way people respond to agreeableness? We chose to investigate agreeable and disagreeable behaviors because of the interpersonal nature of the trait of agreeableness. For example, highly agreeable individuals are often quite successful in interpersonal relationships. They tend to have rapport with their interaction partners and in turn agreeable individuals are often met with invitations to form future relationships from their interaction partners (Cuperman & Ickes, 2009). Another reason why we chose to investigate responses towards agreeable and disagreeable behaviors expressed either anxiously or calmly is because correspondent and reciprocal reactions towards agreeable and disagreeable behaviors can be predicted by the IPC. The Big Five factor of agreeableness can be represented as an axis on the IPC, with agreeable interpersonal behavior falling in the friendly-submissive quadrant, and disagreeable interpersonal behavior falling in the hostile-dominant quadrant. According to the principles of reciprocity and correspondence, agreeable interpersonal behavior should pull for more friendly and dominant responses from interaction partners, whereas disagreeable interpersonal behavior should pull for more hostile and submissive responses from interaction partners (see Figure 1).
We expected, consistent with the principles of reciprocity and correspondence that participants would respond with more dominance and affiliation towards agreeable behavior and more submissiveness and hostility towards disagreeable behavior. However, we hypothesized that by infusing the expressed agreeable and disagreeable interpersonal behaviors with anxiety, patterns of reciprocity and correspondence towards agreeable and disagreeable behaviors would be altered.
In this study, a total of 95 undergraduate students (63 females, 32 males) were randomly assigned to interact with a confederate. The cover story of the experiment required the confederate to deliver a presentation and the participant to evaluate the presentation. In addition, after the confederate delivered the presentation, the confederate and participant were left alone in a room for 45 seconds. All of the research sessions were video and audio recorded. The way in which the confederate delivered the presentation and acted throughout the research session served as the experimental manipulation. More specifically, the confederate’s behaviors differed in agreeableness (High vs. Low) and social anxiety (High vs. Low). The behaviors displayed by the confederate in each of the four conditions were consistent with personality, interpersonal theory, and social anxiety literature (e.g., Alden & Taylor, 2004; Kiesler, 1996; Widiger, 2009). For example, in the “high agreeable–high anxiety” condition, the confederate smiled and engaged the participant, while also displaying socially anxious mannerisms (such as trembling and long conversational pauses). In contrast, in the “disagreeable–low anxiety” condition, the confederate behaved in an abrupt and impolite manner, while also displaying behaviors consistent with someone who is at ease (such as fluid speech and a relaxed body posture). Prior to beginning data collection, videotapes of the confederates’ performances in each of the four conditions were rated by senior graduate students to ensure that the manipulations were salient and the two confederates involved in the study did not differ on how believable or natural they appeared in their roles.
After interacting with the confederate, participants completed a 22-item scale indicating their interpersonal responses to the confederate. Relevant items were drawn from a variety of established interpersonal scales (such as the Social Behavior Inventory and the Checklist of Interpersonal Transactions), and modified to reflect the context of the experiment. For example, the item, “I expressed an opinion to someone” (from the dominance pole of the SBI) was modified to reflect a state-like dominant response of the participant towards the confederate (“I expressed an opinion to the speaker about his/her presentation”). These items were used to generate two dimension scores: dominant versus submissive responses (a = .62), and affiliative versus hostile responses (a = .78). Example items from the dominance and affiliative dimension scores, respectively, are: “I waited for the speaker to talk or act first” and “I nodded at the speaker.”
Three independent coders watched the videotapes of the participant-confederate interactions (including the delivery of the speech by the confederate, and the 45-second period in which the participant and confederate were left alone in the room) and used a modified version of the Social Behavior Inventory (SBI) to rate participants’ behavior in response to the confederate. The original items on the SBI were modified so that coders could rate how often the participant engaged in certain behaviors throughout the research session. For example, the original SBI item “I did not respond to another’s questions or comments” was altered to “the participant did not respond to the confederate’s questions or comments.” Although coders completed modified versions of all 46 items on the SBI, 6 of the modified items (such as “the participant spoke favorably of someone who was not present”) were not endorsed by coders at all, and thus were not included when calculating subscale scores. The remaining 40 items were used to generate two dimension scores: dominant versus submissive responses (a = .83) and affiliative versus hostile responses (a = .68). Example items from the observer-coded dominance and affiliative dimension scores, respectively, are: “this person expressed an opinion to the confederate” and “this person complimented or praised the confederate.”
A manipulation check revealed that participants rated the confederate as warmer when in the agreeable condition than the disagreeable condition, t(93) = 7.63, p < .001, η² = .39 and more anxious when in the high anxiety condition than the low anxiety condition, t(93) = 8.45, p < .001, η² = .43
To
examine whether anxious behavior impacted patterns of reciprocity in dominance,
we performed 2 (agreeable vs. disagreeable) X 2 (high vs. low social anxiety)
ANOVAs with participants’ self-reported dominant responses and observer-coded
dominant responses as the dependent variables. Recall that a reciprocal
response towards agreeable interpersonal behavior is dominance, and a
reciprocal response towards disagreeable interpersonal behavior is submissiveness.
For self-reported dominant responses towards the confederate (Figure 2), there
was a main effect of agreeableness, F(1,93)
= 17.13, p < .001, η² = .16. Participant
reports suggested that they responded in a reciprocal way towards agreeable and
disagreeable behaviors from the confederate. That is, participants reported
more submissive responses to disagreeable behavior from the confederate and
more dominant responses towards agreeable behavior from the confederate,
regardless of confederate level of anxious behavior. Surprisingly,
patterns of reciprocity towards agreeable and disagreeable interpersonal
behaviors were not impacted by the confederate’s anxiety level. That is, the
similarity in slopes of the red and blue lines in Figure 2 shows that participants
responded in similarly reciprocal ways to the high-anxiety confederate as to
the low-anxiety confederate.
For observer-coded dominant responses (Figure 3), the expected patterns of reciprocity towards agreeable and disagreeable behaviors were observed. Participants were coded as responding with more dominance towards agreeable behavior from the confederate and more submissiveness towards disagreeable behavior from the confederate, F(1, 83) = 3.91, p = .05, η² = .05. Again, the anxious behavior of the confederate did not impact patterns of reciprocity. However, unlike the results for participants’ self-reported dominant responses, there was a main effect of_
For observer-coded dominant responses (Figure 3), the expected patterns of reciprocity towards agreeable and disagreeable behaviors were observed. Participants were coded as responding with more dominance towards agreeable behavior from the confederate and more submissiveness towards disagreeable behavior from the confederate, F(1, 83) = 3.91, p = .05, η² = .05. Again, the anxious behavior of the confederate did not impact patterns of reciprocity. However, unlike the results for participants’ self-reported dominant responses, there was a main effect of_
To
examine whether anxious behavior impacted patterns of affiliative
correspondence, we performed 2 (agreeable vs. disagreeable) X 2 (high vs. low
social anxiety) ANOVAs with participants’ self-reported affiliative responses
and observer-coded affiliative responses as the dependent variables. Recall
that a correspondent response towards agreeable interpersonal behavior is
affiliation, and a correspondent response towards disagreeable interpersonal
behavior is hostility. For self-reported affiliative responses towards the
confederate (Figure 4), there was a main effect of agreeableness, F(1, 93) = 8.85, p = .004, η² = .09: Participant
reports suggested that they responded in a correspondent way towards agreeable
and disagreeable behaviors from the confederate. That is, participants reported
more affiliative responses towards agreeable behavior from the confederate and
more hostile responses towards disagreeable behavior from the confederate,
regardless of confederate’s level of anxious behavior. Additionally, there was
a main effect of confederate anxiety level, F(1,
93) = 47.14, p < .001, η² = .34: Participant
reports indicated that they responded with more affiliative responses towards
the high-anxiety confederate than the low-anxiety confederate. Both
of these main effects were qualified by a significant interaction between the
agreeableness level of the confederate and the anxiety level of the
confederate, F(1,93)
= 7.09, p = .009, η² = .07. As shown by the
steeper slope of the blue line in Figure 4, participants had a much stronger
correspondent reaction towards the low-anxiety confederate than the
high-anxiety confederate, shown by the red line with a much flatter slope.
For
observer-coded affiliative responses (Figure 5), participants were perceived as
responding in a correspondent way towards agreeable and disagreeable behaviors
from the confederate. That is, participants were seen as responding with more
affiliation when the confederate was engaging in agreeable behaviors and more
hostility when the confederate was engaging in disagreeable behaviors, F(1, 83) = 33.34, p <.001, η² = .29. Similar to what participants reported,
there was a significant interaction between the agreeableness level of the
confederate and the anxiety level of the confederate, F(1,83) = 6.92, p = .010,
η² = .08. As shown by the steeper slope of the
blue line in Figure 5, participants were coded as having a more correspondent
reaction towards the low-anxiety confederate than the high-anxiety confederate,
shown by the slope of the red line.
To summarize
the results, the anxiety level of the confederate did not impact the extent to
which reciprocity occurred on the dominance dimension: Regardless of the
confederate’s anxiety level, participant reports suggested that they responded
in a reciprocal way towards the confederate. Participants reported responding
with more dominance towards a confederate enacting agreeable behaviors and
responding with more submissiveness towards a confederate enacting disagreeable
behaviors. Furthermore, observer reports concurred with this pattern.
In contrast, the anxiety level of the confederate appeared to interfere with patterns of correspondence on the affiliation dimension: Although participants responded with more affiliation towards agreeable behaviors from the confederate and more hostility towards disagreeable behaviors from the confederate, this expected pattern of correspondence only occurred when participants were interacting with the low-anxiety confederate. Participants' affiliative responses towards the confederate were significantly less correspondent when participants were interacting with the anxious confederate.
Interesting differences also emerged between participants’ overt (i.e. observer-coded behaviors) and self-perceived responses. Although observers saw more dominant behaviors from participants in response to a high-anxiety confederate than a low-anxiety confederate, participants seemed to be unaware of this pattern, rating their dominance responses towards the confederate about the same regardless of the confederate’s anxiety level. Instead, participants tended to see themselves as responding with more affiliation towards the high-anxiety confederate than the low-anxiety confederate, even though observers saw no such difference.
This research allowed us to begin to address the question of whether the natural complementarity that occurs when two people interact is impacted by anxiety behaviors. In the present work, the degree of complementarity shown was restricted to one direction only. That is, this work investigated the degree of complementarity shown by participants in response to scripted confederate behavior. An interesting next step is to investigate the degree of complementarity exhibited between two participants (rather than between participant-confederate dyads). Accordingly, we are currently investigating if anxiety has a similar impact on patterns of complementarity in interactions in which both members of the dyad are free to act and react to each other’s behaviors and interpersonal cues.
One possible explanation for why patterns of correspondence are weakened when someone in an interaction is anxious is that both members of the dyad could be overlooking their partner’s interpersonal bids for correspondence. For example, if the unanxious partner is distracted by the anxiety of the anxious partner, the unanxious partner could miss interpersonal cues from their anxious partner, impairing the extent to which the interaction partners are calibrating their affiliation levels towards one another.
This research is the beginning of an interesting exploration into clarifying how anxiety influences the ways in which behaviors of pairs of people fit together during interpersonal interactions. The results of this research suggest that people’s anxiety behaviors, and the broader construct of neuroticism, ought to be considered when investigating interpersonal processes.
References
Alden, L. E., & Taylor, C. T. (2004). Interpersonal processes in social phobia. Clinical Psychology Review, 24(7), 857- 882.
Costa P. T., & Widiger, T. A. (1994). Personality disorders and the five-factor model of personality. Washington: American Psychological Association.
Cuperman, R., & Ickes, W. (2009). Big five predictors of behavior and perceptions in initial dyadic interactions: Personality similarity helps extraverts and introverts but hurts “disagreeables.”Journal of Personality and Social Psychology, 97(4), 667-684.
Kiesler, D. J. (1996). Contemporary interpersonal theory and research: Personality, psychopathology, and psychotherapy. New York: Wiley.
Meleshko, K. G. A., & Alden, L. E. (1993). Anxiety and self-disclosure: Toward a motivational model. Journal of Personality and Social Psychology, 64, 1000-1009.
Ozer, D. J., & Benet-Martinez, V. (2006). Personality and the prediction of consequential outcomes. Annual Review of Psychology, 57, 401- 421.
Pincus, A. L., & Gurtman, M. B. (1995). The three faces of interpersonal dependency: structural analyses of self-report dependency measures. Journal of Personality and Social Psychology, 69(4), 744.
Widiger, T. A. (2009). Neuroticism. In M. R. Leary & R. H. Hoyle (Eds.), Handbook of individual differences in social behavior (pp. 129 – 146). New York: Guilford.
Wiggins, J. S., & Pincus, A. L. (1994). Personality structure and the structure of personality disorders. In P. T. Costa & T. A. Widiger (Eds,), Personality disorders and the five- factor model of personality (pp. 73 – 93). Washington: American Psychological Association.
Correspondence concerning this article should be addressed to Kelly McDonald.
Email: [email protected]
In contrast, the anxiety level of the confederate appeared to interfere with patterns of correspondence on the affiliation dimension: Although participants responded with more affiliation towards agreeable behaviors from the confederate and more hostility towards disagreeable behaviors from the confederate, this expected pattern of correspondence only occurred when participants were interacting with the low-anxiety confederate. Participants' affiliative responses towards the confederate were significantly less correspondent when participants were interacting with the anxious confederate.
Interesting differences also emerged between participants’ overt (i.e. observer-coded behaviors) and self-perceived responses. Although observers saw more dominant behaviors from participants in response to a high-anxiety confederate than a low-anxiety confederate, participants seemed to be unaware of this pattern, rating their dominance responses towards the confederate about the same regardless of the confederate’s anxiety level. Instead, participants tended to see themselves as responding with more affiliation towards the high-anxiety confederate than the low-anxiety confederate, even though observers saw no such difference.
This research allowed us to begin to address the question of whether the natural complementarity that occurs when two people interact is impacted by anxiety behaviors. In the present work, the degree of complementarity shown was restricted to one direction only. That is, this work investigated the degree of complementarity shown by participants in response to scripted confederate behavior. An interesting next step is to investigate the degree of complementarity exhibited between two participants (rather than between participant-confederate dyads). Accordingly, we are currently investigating if anxiety has a similar impact on patterns of complementarity in interactions in which both members of the dyad are free to act and react to each other’s behaviors and interpersonal cues.
One possible explanation for why patterns of correspondence are weakened when someone in an interaction is anxious is that both members of the dyad could be overlooking their partner’s interpersonal bids for correspondence. For example, if the unanxious partner is distracted by the anxiety of the anxious partner, the unanxious partner could miss interpersonal cues from their anxious partner, impairing the extent to which the interaction partners are calibrating their affiliation levels towards one another.
This research is the beginning of an interesting exploration into clarifying how anxiety influences the ways in which behaviors of pairs of people fit together during interpersonal interactions. The results of this research suggest that people’s anxiety behaviors, and the broader construct of neuroticism, ought to be considered when investigating interpersonal processes.
References
Alden, L. E., & Taylor, C. T. (2004). Interpersonal processes in social phobia. Clinical Psychology Review, 24(7), 857- 882.
Costa P. T., & Widiger, T. A. (1994). Personality disorders and the five-factor model of personality. Washington: American Psychological Association.
Cuperman, R., & Ickes, W. (2009). Big five predictors of behavior and perceptions in initial dyadic interactions: Personality similarity helps extraverts and introverts but hurts “disagreeables.”Journal of Personality and Social Psychology, 97(4), 667-684.
Kiesler, D. J. (1996). Contemporary interpersonal theory and research: Personality, psychopathology, and psychotherapy. New York: Wiley.
Meleshko, K. G. A., & Alden, L. E. (1993). Anxiety and self-disclosure: Toward a motivational model. Journal of Personality and Social Psychology, 64, 1000-1009.
Ozer, D. J., & Benet-Martinez, V. (2006). Personality and the prediction of consequential outcomes. Annual Review of Psychology, 57, 401- 421.
Pincus, A. L., & Gurtman, M. B. (1995). The three faces of interpersonal dependency: structural analyses of self-report dependency measures. Journal of Personality and Social Psychology, 69(4), 744.
Widiger, T. A. (2009). Neuroticism. In M. R. Leary & R. H. Hoyle (Eds.), Handbook of individual differences in social behavior (pp. 129 – 146). New York: Guilford.
Wiggins, J. S., & Pincus, A. L. (1994). Personality structure and the structure of personality disorders. In P. T. Costa & T. A. Widiger (Eds,), Personality disorders and the five- factor model of personality (pp. 73 – 93). Washington: American Psychological Association.
Correspondence concerning this article should be addressed to Kelly McDonald.
Email: [email protected]