Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Psychology


Behavioral Sciences



First Advisor

Lisa Harlow


While the efficacy of Transtheoretical Model (TTM)-based interventions has been demonstrated time and time again, little has been done to understand the causal mechanisms responsible for the change resulting from exposure to a TTM-based intervention. The proposed study is to leverage sophisticated statistical techniques to examine the causal relationship among TTM constructs to better understand the direction of influence TTM-based interventions have. To accomplish this, a secondary data analysis using longitudinal data collected during a trial evaluating the efficacy of a TTM-based alcohol intervention targeting adults whose drinking exceeds national recommended guidelines was completed.

This study used a 2 x 5 factorial design to evaluate the efficacy of a computer-delivered alcohol intervention relative to assessment-only control. While the original study design involves participants randomly assigned to two conditions, the focus of these analyses will be on those who received the intervention (N=492), for whom data on TTM-related constructs was collected in addition to outcomes of interest. Participants in the intervention group were assessed on decisional balance and temptations at baseline, 3-, and 6-months. They were also assessed on stage of change to adhere to low-risk drinking guidelines and alcohol consumption outcomes (i.e., drinks per week, drinks per drinking day, and binge frequency) at five time points (baseline, 3-, 6-, 12-, and 18-months).

Preliminary analyses yielded no associations between decisional balance and drinking outcome variables, but there was a strong association between temptations and all measures of alcohol consumption. As such, the main outcome analyses focused curve modeling (LGCM) was used to explore these relationships using the lavaan package in R. Parallel process models, simultaneously modeling temptations and each of the alcohol variables, were conducted with the inclusion of stage of change as a time-varying covariate. Models allowing for covariances to be freely estimated had significantly better fit than models that fixed these covariances at zero. These models were further extended with the inclusion of regression paths between the two slope factors in each model. Both potential directions of influence between these two factors were considered in separate models, but their addition negligibly changed model fit, and none of the six models examined yielded a significant regression path. While nonsignificant, the regression path regressing the slope of drinks per week on the slope of temptations approached significance, suggesting that change in drinks per week my precede change in temptations.

The relationships observed between cognitive constructs and drinking outcome behavior deviated from previous literature, in that there was no association between decisional balance and drinking, and there was some evidence that the direction of influence between temptations and drinking was not as expected. Future work is needed to replicate and elucidate these findings, especially given the many methodological issues inherent to these data.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.