RESUMO
OBJECTIVE: We investigated the technical model of motivational interviewing (MI) in a dual-outcome intervention (i.e., alcohol, sexual risk; N = 164; 57% female). METHOD: We identified latent classes of client change statements, based on the proportion of change talk (CT) over the session. We then examined whether outcomes were related to CT class, and whether the relations between MI skill and outcomes varied by CT class. RESULTS: We found three classes of alcohol-CT and two classes of sexual risk-CT. While CT class membership did not predict outcomes directly, greater therapist MI-consistent skill was associated with fewer heavy drinking days in the increasing alcohol-CT class. For sexual risk outcomes, therapist MI-consistent skill was associated with reduced odds of condomless sex for the low sexual risk-CT class. CONCLUSIONS: The relation of therapist MI consistency to outcomes appears to be a function of client CT during the session.