Modelling Dynamic Affective and Decision Making Processes during Approach and Avoidance Goal Striving

Abstract

This thesis examined the cognitive processes by which individuals engage in goal level choice and the affective processes that individuals experience during approach and avoidance goal striving. The aim was to develop a set of formal theories that can explain and predict the trajectories in goal level choice and affective states as individuals strive towards goals framed as desirable or undesirable end-states. These formal theories provide an explanation of processes at multiple levels of analysis, and incorporate constructs from the work motivation, self-regulation, emotion, personality, and decision making literatures. Data from three experimental studies were analysed using multi-level techniques and computational modelling in order to test the proposed theories. The findings are reported within three manuscripts that form the body of this thesis. Manuscript 1 investigated goal level choice as a cognitive process involving an interaction of decision making, risk preferences, personality, and learning. A computational model was developed in order to describe and predict how individuals make goal level choices over time. Drawing from the decision making literature, the model specifies that goal level choice at the within-person level can be explained by a dynamic anchoring and adjustment process and a risk preference bias. The model further specifies that between-person variability in risk preferences can be partly explained by an interaction of goal framing and trait neuroticism. An experimental study was conducted to evaluate this model. Participants were randomly allocated to strive towards approach or avoidance framed goals as they performed a medium-fidelity Air Traffic Control (ATC) simulation task across 10 trials. At the beginning of each trial, participants chose among five goal levels. The findings of this study indicated that our hypothesized model was able to account for the trajectories of goal level choices over time. Manuscript 1 contributed to the goal striving literature by identifying relevant cognitive processes in goal level choice. Manuscripts 2 and 3 investigated the affective processes that emerge as individuals strive towards a chosen goal level in approach and avoidance contexts. Manuscript 2 presented a scale validation study for the momentary affect scale (MAS), which was developed in order to measure the dynamics of affective experiences during goal striving. As part of the scale validation, it was necessary to address the conceptual question of whether the affect scale should consist of unipolar or bipolar items. Participants performed a medium-fidelity ATC simulation task and were repeatedly measured on a set of unipolar and bipolar affect items. Multi-level confirmatory factor analyses were used to evaluate the problem of unipolarity versus bipolarity and to validate the MAS against the latent constructs being measured. The findings of this study provided evidence of bipolarity in the structure of affect and indicated that the MAS can be used to measure rapid changes in affective states during goal striving. Manuscript 3 presented a formal model to explain how affective states change during approach and avoidance goal striving. The model specifies that during goal striving, indicators of goal progress are associated with affective reactions on the valence dimension, whilst indicators of task demand are associated with affective reactions on the arousal dimension. An empirical study was conducted to evaluate this model and identify whether it could account for the trajectories in affective states as participants strove towards approach or avoidance framed goals. Participants performed a medium-fidelity ATC simulation task derived from the study in Manuscript 1, but modified with longer goal striving episodes and greater variability in workload in order to observe greater range and duration of affective experiences. Affect ratings were collected on the MAS at one minute intervals as participants engaged in goal striving across five trials. The hypothesized model and a series of nested alternative models were fitted to the data. The findings indicated that the hypothesized model provided a parsimonious and accurate account of the changes in valence and arousal in relation to the indicators of goal progress and task demand, respectively. Manuscript 3 contributed to the goal striving literature by providing a formal account of how affective states can change over time as individuals engage in approach or avoidance goal striving. The overall contribution of this thesis is to further the theoretical and empirical understanding of cognitive and affective processes in goal striving. The theoretical frameworks used in these papers emphasize the role of multi-disciplinary and integrative approaches for understanding complex phenomena. Specifically, the integration of multiple disciplines and constructs enable a more precise account of how goal level choice and affective states vary over time. A key methodological advantage of this research is the use of multi-level research designs and computational modelling, which are appropriate for testing dynamic constructs and process-based theories. In terms of practical implications, the models in these studies offer insight into patterns of human behaviour and affective experiences during goal striving, which can be applied in the domain of human factors.