Year

2010

Degree Name

Doctor of Philosophy

Department

University of Wollongong. School of Psychology

Abstract

This thesis involved a psychometric investigation of several dysfunctional processes that have been widely hypothesized to mediate psychological disturbance and change in CBT. They included dysfunctional attitudes, poor self-esteem, thought suppression, difficulty identifying and describing emotions, ineffective problem orientation and experiential avoidance.

Much has already been written about these dysfunctional process variables and previous research provides evidence that they each significantly influence emotional wellbeing across a variety of populations and contexts. However, to date these variables have largely been studied in isolation and often in competing fields of research. This has resulted in major gaps in current understanding of the mechanisms of disturbance and change. It has also raised uncertainty as to whether the hypothesized mediators are actually separate constructs or whether they may be redundant.

The present thesis brought together the dysfunctional process variables to explore the extent of their convergence and distinction. It also examined the degree to which they uniquely predict variance in emotional wellbeing. Measures of dysfunctional processes and emotional wellbeing were completed by two groups of participants, undergraduate psychology students and NSW police recruits.

Study One involved an exploratory factor analysis of the dysfunctional processes using the student data. It produced a preliminary model of the structure of the dysfunctional process measures that indicated that the measures were largely related and yet minimally redundant.

Study Two entailed a confirmatory factor analysis involving structural equation modelling and multiple stepwise regression analyses. It explored the replicability of the initial model using the police data set. It also examined whether the initial model was optimal, by exploring the fit of alternative models on both the student and police data sets.

Study Two produced a more parsimonious model of the latent structure of the dysfunctional process variables than the initial model. This model, named The Brief Dysfunctional Processes Measure was unexpectedly brief despite its scope. It consisted of six dysfunctional process variables, namely Dysfunctional Achievement Concerns, Poor Self-Esteem, Thought Avoidance, Difficulty Identifying Emotions, Avoidant Problem Orientation and Experiential Avoidance. These were linked by an overarching factor, Negative Avoidant Dysfunctional Processes.

Study Two went on to explore the predictive relationships between variables in The Brief Dysfunctional Processes Measure with aspects of emotional wellbeing. It demonstrated that all dysfunctional process variables significantly contributed to predicting some aspects of emotional wellbeing even after controlling for gender and age. There was some variation in the pattern of predictors across samples and across the measures of emotional wellbeing. However, three variables were notably consistent and pervasive predictors. These were Experiential Avoidance, Avoidant Problem Orientation and Difficult Identifying Emotions.

The results of this investigation reveal that the dysfunctional process variables appear related yet largely distinct. Moreover, the variables in The Brief Dysfunctional Processes Measure predict unique variance in wellbeing. These findings point to the value of synthesizing existing research on the hypothesized mediators. This stands to increase understanding of their interrelationships, and may ultimately help illuminate the question of‘what works’ in psychotherapy interventions.

Share

COinS
 

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.