Abstract

The present research examines the long-hypothesized association between variability of self-esteem and emotional reactivity within the context of peer interactions. Self-esteem variability is defined as short-term fluctuations in self-esteem and operationalized as the standard deviation of multiple assessments. Emotional reactivity is defined as the association between the perception of an event and the emotional response to it. 102 male adolescents (ages 13 to 15) participated in the research. Multiple data on self-perceived social competence were collected using Time Sampling. These repeated data were averaged into a measure of level of perceived social competence and the standard deviation used as a measure of variability of perceived social competence. Using the Experience Sampling Method, repeated data were collected in situ on the perceptions of the quality of participants' peer interactions (1144 distinct peer events) and on the emotional reactions following them as measured by scales of sadness, anxiety, anger and positive affect. Individual difference data (e.g., popularity) were collected during group testing. Hierarchical Linear Modeling was used for analyses. The predictive power of (1) level and (2) variability of perceived social competence, (3) emotional distress (measured with depression and anxiety scales), (4) popularity (measured by peer rated friendship nominations and liking ratings), (5) scores on Stability of Self scale and (6) the interaction between level and variability of perceived social competence were tested concurrently. Consistent with hypotheses, variability was associated with increased emotional reactivity to events, that is, it was a significant moderator of the association between perceptions of interaction quality and feelings of sadness, anxiety and positive affect. For example, participants with higher variability reported more sadness when interaction was perceived as negative than did their peers with less variable perceived social competence; no differences between participants were noted when the interaction was perceived as positive. Emotional distress was a significant predictor of the negative association between interaction quality and feelings of anger. Popularity acted as a 'buffer' to temper the association between interaction quality and sadness. Consistent with hypotheses, variability of perceived social competence was a significant predictor of higher levels of sadness, anxiety and anger. Unexpectedly, average level of perceived social competence was a significant predictor of higher levels of sadness, anxiety and anger. Results are discussed in terms of the nature and measurement of self-esteem variability, its role as a vulnerability factor for the experience of negative affect, and popularity's role as a protective factor.