Modeling motivation three ways: effects of MI metrics on treatment outcomes among adolescents.

The purpose of this study was to determine how three different measures of motivation (cognitive motivation, taking steps, and self-efficacy for change and maintenance) predict substance use outcomes after engaging in a Motivational Interviewing intervention. Participants were 225 high school students enrolled in Project Reducing the Effects of Alcohol and Drugs on Youth (Project READY), a NIDA-funded intervention initially developed with Motivational Interviewing (MI) principles for adolescents identified by schools as having problems with alcohol or other drug use. We measured motivation at multiple time points during the intervention in multiple methods. Cognitive motivation was assessed using a Decisional Balance matrix at Session 3 of treatment. We measured self-efficacy with the Situational Confidence Questionnaire, administered at 4-, 8-, 12-, and 16-week follow-ups. A measure of taking steps (SOCRATES, v. 8) was administered at intake and Session 8. We hypothesized that motivation would follow the Transtheoretical Model (TTM) pathway, and we proposed a model where cognitive motivation would predict self-confidence for change and taking steps toward change, and self-confidence and taking steps would predict substance use outcomes. We tested our model using path analysis in AMOS and found support for a motivational continuum predicting percent days abstinent at 16-week follow-up [χ(2) = 2.75, df = 7, p = .90, CFI = 1, RMSEA (90% confidence interval) = .00 - .03]. This model demonstrates that motivational metrics predict unique outcomes at different time points and serve as important components of intervention.