Date of Award
Master of Arts in Psychology
Lisa L. Harlow
This study attempts to identify how women naturally progress through the Transtheoretical Model Stages of Condom Use over a one year period, as well as identifies the psycho-social and behavioral predictors of these changes. Three separate types of analyses (Latent Transition Analysis, Discriminant Function Analysis, and Logistic Regression) were used in an effort to assess how they complement each other in longitudinal studies such as this. A total of 545 women participated in the one year study of HIV risk in women. Latent Transition Analysis identified the best fitting model of change of stage, which included both forward and backward movement. Precontemplation and Maintenance were found to be the most stable stages. Results from the MANOVA/DFA and Logistic Regression procedures showed that women with high Condom Pros at baseline were more likely to move either forward or backward at least one stage than they were to remain in the same Stage of Condom Use over one year. In addition, Positive Psychosexual Attitudes were negatively associated with forward stage movement. The Logistic Regression model for progression showed that those women who were non-white, had more than a high school education, or were single were more likely to progress than to remain stable. The model on backward movement identified age and race as important predictors. Strengths and limitations for all three types of analyses were apparent. Based on the results of this study, interventions can be developed focusing on these findings in an effort to increase condom use among women. In addition, these results suggest that a combination of methods be employed when analyzing longitudinal data such as in this study.
Evers, Kerry E., "Longitudinal Prediction of Women in the Stages of Change for Condom Use: A Comparison of Methods" (1996). Open Access Master's Theses. Paper 1658.