Date of Award

Winter 2008

Document Type


Degree Name

Doctor of Philosophy (PhD)


Health Services Research

Committee Director

Martha Walker

Committee Member

Stacey Plichta

Committee Member

George Maihafer


This study is an assessment of the usefulness of the Expanded Parallel Process Model in predicting health enhancing physical activity as an outcome variable. The theory is tested in the context of risk for coronary heart disease and involves secondary analyses of a dataset from a group of working adults. These individuals had elected first to participate in a health plan ‘Quality Improvement Study’ and were then randomly selected to receive an intervention program designed to get people to be more active.

Data on self-reported demographics, physical activity levels, health status characteristics and perceptions measured on a Likert-type scale known as the Risk Behavior Diagnosis Scale are analyzed. The perceptions measured cover the threat to one's health that a heart attack poses in terms of severity and one's susceptibility as well as perceptions about the effectiveness of physical activity to reduce or avert this threat and one's ability to engage in the required amount of physical activity to obtain this health benefit. These Risk Behavior Diagnosis Scale measures represent the Expanded Parallel Process Model hypothesized mediating variables which are perceived severity, perceived susceptibility, perceived response efficacy and perceived self-efficacy.

Testing of the model consists of the examination of variable relationships to assess whether the Expanded Parallel Process Model related perceptions are related to meeting health enhancing physical activity requirements. Overall, the results of data analyses offer limited and weak support for the use of the Expanded Parallel Process Model to explain differences in health enhancing physical activity behavior of working adults. Although the magnitude of the hypothesized Expanded Parallel Process Model related mediator variables observed in this dissertation study are small, this evidence may suffice in calling for further research using study designs, other than cross-sectional surveys that provide opportunity for mediator analysis. This study also concludes that the determinants of health behavior change are complex and this warrants the development and exploration of theoretical models that encompass different approaches to the study of physical activity behavior.