Psychometric properties of the system for coding couples' interactions in therapy--alcohol.

Few systems are available for coding in-session behaviors for couples in therapy. Alcohol Behavioral Couple Therapy (ABCT) is an empirically supported treatment, but little is known about its mechanisms of behavior change. In the current study, an adapted version of the Motivational Interviewing for Significant Others coding system was developed into the System for Coding Couples' Interactions in Therapy-Alcohol (SCCIT-A), which was used to code couples' interactions and behaviors during ABCT. Results showed good interrater reliability of the SCCIT-A and provided evidence that the SCCIT-A may be a promising measure for understanding couples in therapy. A 3-factor model of the SCCIT-A (Positive, Negative, and Change Talk/Counter-Change Talk) was examined using a confirmatory factor analysis, but model fit was poor. Because model fit was poor, ratios were computed for Positive/Negative ratings and for Change Talk/Counter-Change Talk codes based on previous research in the couples and Motivational Interviewing literature. Post hoc analyses examined correlations between specific SCCIT-A codes and baseline characteristics, and indicated some concurrent validity. Correlations were run between ratios and baseline characteristics; ratios may be an alternative to using the factors from the SCCIT-A. Reliability and validity analyses suggest that the SCCIT-A has the potential to be a useful measure for coding in-session behaviors of both partners in couples therapy and could be used to identify mechanisms of behavior change for ABCT. Additional research is needed to improve the reliability of some codes and to further develop the SCCIT-A and other measures of couples' interactions in therapy.