Degree Name

Doctor of Philosophy


School of Psychology


This thesis presents three empirical studies examining the relations between religious sentiment and psychological outcomes over the course of adolescence. Although there are numerous theories of religious development, and many studies have found positive relations between religious sentiment and psychological outcomes, comparatively little research has focused on the adolescent years, especially longitudinally. This is surprising, given that adolescence is the period of the lifespan where the most significant changes to religious sentiment occur. The sample consisted of approximately 2000 high school students participating in the Australian Character Study (ACS) over a four-year period.

Study 1 focused on obtaining an understanding of the ways in which belief in God was related to psychological functioning at time 1 (grade 8). Three groups (mean age = 13.92 years; 946 males, 979 females) were compared according to belief in God (believer, agnostic, atheists), on positive, social, and negative outcomes, using a profile analysis. On almost all outcomes, believers were found to have improved psychological functioning compared to agnostics, who in turn, tended to outscore atheists. The profile analysis also revealed that groups had distinctive psychological profiles, with believers’ profiles distinguished by elevated subjective well-being and parental support; and atheists’ by low affective empathy and high antisocial behaviour. The findings of this study provided the basis for the questions explored in studies 2 and 3.

Study 2 further explored relations between belief in God and outcomes examined in study 1, by repeating analyses for the proceeding 3 years. Results indicated that the three groups’ profiles remained distinctive over time, but that grade 11 profiles significantly diverged from previous years for all 3 groups. Differences between the groups also widened by grade 11, with believers’ functioning found to improve over time, and atheists having more impaired functioning over time. Unlike study 1 however, all differences between groups in grades 10 and 11 were removed after controlling for parental support. This indicates that improved functioning amongst older religious adolescents is likely related to parental support.

Study 3 adopted a growth mixture model approach to identify individual differences in trajectories of religious values from grade 8 to grade 11. Overall religiousness was found to decline, and 4 unique trajectories of religious values were identified: High (steep decline), Average, Low (average decline), and Low (stable). Relative to the Average class, increased parental support in grade 8 predicted an increased likelihood of having high religious values with a sharp decline and a reduced likelihood of being in the classes with religious values below the average: Low (average decline) and Low (stable). Members of the High and Average classes were found to be more likely to have improved well-being in grade 11. These findings are suggestive of the need to account for heterogeneity in domains of religiousness when examining religious effects.

This thesis extends previous research, being the first collection of findings examining relationships between religious sentiment and psychological outcomes in a single sample over the course of adolescence. Results suggest that belief in God is related to the highs and lows experienced by adolescents, but that the strength of relations between religious sentiment and psychological outcomes may differ between early and late adolescence. The identification of individual trajectories of religious values and how these are related to positive outcomes also demonstrates the importance of accounting for individual changes in religiousness during adolescence. Results also highlight the importance of parental support, which predicted variation in the development of religious values, and explained the improved adjustment found amongst older religious adolescents.