Abstract
This article describes the coding portion of a study to test the effectiveness of a motivational interviewing (MI) training program for probation officers. We describe some of the challenges with using the Motivational Interview Treatment Integrity (MITI) instrument to code interactions between probation officers and clients. Our team of raters was able to obtain adequate inter-rater reliability on most MITI scales, though reliability ratings on some of the specific behavior counts such as Giving Information, MI Adherent, and MI Non-adherent fell considerably lower than the original MITI norming study. Our results suggest that the MITI is a mostly reliable instrument for coding criminal justice interactions, though there were exceptions to this rule. Based on our experiences, we discuss some of the ways that probation interactions might be different from traditional counseling interactions, and identify some rules of thumb that helped us to code interactions. We end with suggestions for how MITI feedback can be used effectively in training and supervision in criminal justice and other non-traditional settings.
