Assessing Motivational Interviewing 2.0: An illustration of software-supported coding schemes
Scholars in Motivational Interviewing (MI) have developed several coding schemes to assess treatment adherence, client language, and sequential dynamics. Traditionally, these coding schemes have been administered with paper and pencil. The presentation introduces implementations of software-supported MI coding schemes and discusses how software-implementation facilitates coding work, while still capturing the sequential timed-event data of the dyadic interaction. Furthermore, it is presented how coding instruments can be economized by means of a thin behavior slicing procedure. Data originated from a study with MI trained interviewers who discussed sustainable use of environmental resources as a target behavior with their respective clients. First, it is presented how branchedchain coding can be used to familiarize inexperienced observers with the MI Skill Code. Second, it is calculated two observer agreement measures of fourteen double-coded sessions for the MI Treatment Integrity Code (MITI): Time-unit kappa and Intraclass correlation. Third, the researchers extracted thin behavior slices (ten minutes) and compared their MITI code statistics with the entire session. Results show that Kappa indices are more conservative reliability estimates than Intraclass correlations. Furthermore, thin behavior slicing revealed that only 10-minutes can provide accurate estimates for MITI verbal behavior codes. The researchers discuss costs and benefits of software-supported coding schemes. Presentation at the ICMI International Conference on Motivational Interviewing, Amsterdam, June 16-18, 2014.