Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Peter N.S. Hoaken


Dynamic systems theory (DST) can provide a comprehensive account for how parent-child interactions evolve over time to produce stable patterns of interacting and can result in seemingly divergent trajectories. Recent methodological advances using state space grids (SSGs) have provided a graphical means to examine real-time dyadic processes, as well as measures of dyadic flexibility, or the ability to adapt emotional and behavioural responding in response to contextual demands. Higher levels of dyadic flexibility have been associated with improvements in child behaviour problems after treatment (Granic et al., 2007), while its converse, rigidity, has been associated with increases in behaviour problems over time (Hollenstein et al., 2004). The type of task (e.g., structured versus unstructured) in which the dyad is engaged may also impact the relative importance of dyadic and parental characteristics (e.g., warmth) on the interaction. The present study examined parent-child interactions involving clinically-referred children ages 3 years, 11 months, to 6 years of age with externalizing behaviour problems (n=33 dyads). Flexibility variables as identified in previous research were examined across task types using principal components analyses and a multiple discriminant function analysis, resulting in a standardized flexibility composite. Similarly, dyadic processes identified from SSGs were replicated and examined across tasks using a repeated-measures ANCOVA. Next, the differential prediction of dyadic flexibility by dyadic processes and parental characteristics across task types was examined using regression analyses. Finally, subgroups of children with behaviour problems were examined for differences in dyadic processes across tasks. Generally, positive parenting characteristics tended to vary more across tasks relative to negative parenting processes. Predictors of flexibility by dyadic processes and parental characteristics tended to differ by task type, illustrating the importance of looking at the demands of the task and changing contexts. Somewhat counterintuitively, negative dyadic and parental variables were found to predict flexibility, possibly reflecting a transitional reorganization of dyadic interaction processes in the preschool period. Synchronous parent-child interactions did not tend to predict flexibility. Dyadic flexibility and the differential impact of parental and dyadic processes across different task types appear promising as potential targets in early intervention programs for children with behaviour problems.