The affective and relational bases of adolescent mathematics anxiety

According to previous research, mathematics anxiety tends to increase in early adolescence, in addition to general decreases in academic motivation. Concurrent with these trends, adolescence is also the time when peers begin to play a larger part in the social development of the individual. The objective of this thesis was to explore adolescent mathematics anxiety from two broad perspectives. Two studies were conducted that examined affective (Study 1) and relational (Study 2) factors associated with the mathematics anxiety of Year 8 students from two co-educational schools in Victoria, Australia. A two-wave, two- school design allowed extensive analyses of these factors from both cross-sectional and longitudinal positions. Part one of the thesis focuses on the affective basis of adolescent mathematics anxiety. Study 1 explored trait mathematics anxiety in addition to the state or on- task anxiety that arose as participants completed a mathematics problem- solving task. Assessment of on-task anxiety was achieved via the use of the computer program Between the Lines (BTL). The two dimensions of emotionality and worry were found within trait responses but not state responses. Trait mathematics anxiety was also found to be distinct from general test anxiety. Consistent with research trends, girls reported higher levels of mathematics anxiety, and anxiety was negatively correlated with achievement. Person-centred analysis of on-task engagement produced profiles that showed a unique association of anxiety and interest with performance patterns on the BTL tasks; in turn, these profiles were related to particular trends of regulatory compatibility. Finally, relationships between anxiety and motivation, namely value and competence beliefs, were examined according to the control-value theory of achievement emotions. These associations varied for trait and state (pre- and post-task) models and also according to school. Part two of the thesis offers a relational perspective on adolescent mathematics anxiety by focusing on the role that peer networks play in its aetiology. While the dyadic effects of peer relationships on motivation and learning have been investigated, less attention has been directed to the influence of different types of peer networks. Study 2 used a social network approach and contemporary modelling techniques to explore peer interactions and their effects on mathematics anxiety and motivation. This thesis provided the first application of the two modelling techniques in the area. A cross- sectional analysis extended trends observed within Study 1 data to emphasize the impact of contextual factors, with peer influence and peer culture varying according to school, type of peer relationship, and the specific attribute considered, namely, mathematics anxiety, value or competence beliefs. Social influence modelling allowed for the exploration of different mechanisms of social influence – broadly, contagion and activity processes – within time, help-seeking and close- friend networks. Longitudinal modelling of these three networks, along with mathematics anxiety and gender, again revealed between school variation and differences across peer relationships. No evidence of social influence was found, however, social selection effects were observed in varying networks across the two schools.