Degree Name

Doctor of Philosophy (PhD)


Department of Psychology - Faculty of Health & Behavioural Sciences


Although there has been substantial investigation of the association between general attributional style and depression in adolescence, little research has examined the relationship between attributional tendencies specific to the peer-social domain and socio-emotional adjustment. A handful of studies suggest that adolescent peer-social attributional tendencies are related to certain aspects of socio-emotional adjustment, namely, loneliness, depression, and self-reported victimisation. However, this research has been of a concurrent nature and thus demonstrated no causal relationship. The studies to be reported here examined the nature of attributions for peer-social events in early adolescence and the concurrent and longitudinal associations of individual differences in peer-social attributional tendencies with indicators of socio-emotional adjustment. Two longitudinal studies were undertaken; both using similar instruments to assess attributional style for peer-related events. The first study took place over a shorter period (12 months) and used a smaller sample (n = 36 completing the study) than the second study (24 months, n = 82 completing the study). The first study assessed attributions via a personal interview protocol, which allowed respondents to articulate their attributions for the 12 hypothetical scenarios presented to them. Respondents rated these attributions on three scales representing causal internality, stability and globality. The second study presented 14 scenarios and required respondents only to rate the cause of each on the same rating scales. The three primary dependent variables in both studies were self-reported feelings of victimisation, loneliness, and depression assessed via an emotional adjustment questionnaire largely based on the Birleson Depression Scale and the Illinois Loneliness Questionnaire. The second study also involved a peer-reported measure of social status as an additional dependent variable, namely, social centrality derived from the SCM procedure. Results of Study 1 showed a predominance of attributions to others' feelings and motives. Other prominent attributional factors were relationship factors, which were particularly used to explain positive events. Attributions that equate with common attributions made for achievement events (such as personal characteristics and actions) were used relatively infrequently. Mean ratings of these attributions on the three attributional dimensions was also examined. In Study 1, attributional style was operationalised both in terms of the relative usage of attributional types as well as in terms of ratings on the attributional dimensions. Few associations between relative usage of attributional types and socio-emotional adjustment were evident. There were more associations between attributional ratings and socio-emotional adjustment. In particular, significant concurrent and longitudinal associations were found between poorer adjustment and causal stability for negative events, a composite of stable and global attributions for negative events and a bivalent composite for stable and global attributions. Study 2 found yet more associations between ratings on attributional dimensions and emotional adjustment. In contrast to Study 1, causal internality for positive events and a composite measure indicating the relative absence of self-serving attributional bias, was concurrently associated with loneliness and victimisation. There were also concurrent associations between stable and global attributions for negative events and both depression and victimisation. Longitudinal associations between these measures were evident in the correlational analysis. However, the major composite variables were not significant predictors of emotional adjustment when Time 1 adjustment was controlled in a multivariate analysis. In both studies, there was significant intercorrelation between the stability and globality scales independent of event valence. Post hoc analyses showed that a total generality composite measuring stable and global attribution for both positive and negative events predicted both loneliness and depression despite not being concurrently associated with these variables. Moreover, a comparison of groups based on Time 1 total generality showed that the high generality group was significantly higher in Time 2 maladjustment despite not differing from lower generality groups at Time 1. This group also showed increased emotional difficulties at Time 2 whereas the average and low generality groups remained stable or showed slight decreases. A possible explanation for the increased emotional maladjustment associated with the generality style is advanced. It is suggested that the unstable and unique nature of the peer-social domain in early adolescence means that stable and global attributions are generally inaccurate and lead to unrealistically low and/or high expectations. The mismatch of expectancies that emanates from this attributional style and ensuing events contributes to negative affect. The association between social centrality and attributional style was limited. There were no concurrent differences between social centrality (or status) groups. Some longitudinal associations were evident, but the most maladaptive attributional tendencies at Time 1 were found amongst individuals who attained secondary (or average) social centrality at Time 2 rather than those of lower status (peripheral/isolate). This was inconsistent with findings indicating an association between peer-social attributional tendencies and lower sociometric status in children and preadolescents. Both studies are evaluated and directions for future research are suggested.

02Whole.pdf (1425 kB)



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.