-
Coding
What is coding?
Coding involves evaluating statements in a conversation using a reliable, standardized manual. Coding of the counselor's statements in a conversation is based on the Motivational Interviewing Treatment Integrity Manual (MITI 4.2.1) and answers the question: 'To what extent is MI used during the conversation?'
what is coding?how does coding work?how are the coders trained?Quality assurance of operationsIMPORTANT TO CONSIDERtnt applicationConversations with actorsConfidentiality
What is coding?
Coding involves evaluating the counselor's statements reliably using a standardized manual.
Coding involves evaluating statements in a conversation using a reliable, standardized manual. Coding of the counselor's statements in a conversation is based on the Motivational Interviewing Treatment Integrity Manual (MITI 4.2.1) and answers the question: 'To what extent is MI used during the conversation?'
Coding of the client's statements in a conversation is based on the Client Language Easy Rating (CLEAR) and answers which client statements are neutral and are not judged to affect the client's future change, as well as which ones represent change talk (towards a change) and which ones represent sustain talk (against a change).
How does coding work?
Coding is done using coding manuals. Counselor statements are coded based on the 'Motivational Interviewing Treatment Integrity Coding Manual 4.2.1' (MITI). Client statements are coded based on the CLEAR manual. The client's level of self-disclosure can also be coded on a 7-point scale.
SELF-DISCLOSURE
The client's level of self-disclosure in a conversation is evaluated using a 7-point Likert scale. The client's coded level of self-disclosure has been found to be related to the client's future change in research.
CLEAR
With CLEAR ('Client Language Easy Rating'), two variables are coded with seven subcategories each. The two types of client statements are those that indicate the client will change (Change Talk) and those that indicate the client will not change (Sustain Talk).
Statements for and against change are defined in six subcategories and are coded as frequency counts. The coding is summarized in the percentage of change talk (frequency of change talk / frequency of change talk + frequency of sustain talk).
MITI
With MITI, 14 variables are coded, defining the counselor's statements in the conversation and the overall performance of the conversation. The four overarching variables are summarized into two components: the Relationship variable - engagement (Empathy and Partnership) and the Technical variable - directive (Evoking Change Talk and Softening Sustain Talk). Ten types of conversation statements are defined and frequency counted (Information, Persuade, Persuade with Permission, Question, Simple Reflection, Complex Reflection, Affirm, Seek Collaboration, Autonomy Support, Confront).
The first version of the 'Motivational Interviewing Treatment Integrity Coding Manual' was published in 2003 (MITI 2.0). It was updated based on new research to versions MITI 3.0 and MITI 3.1. In 2015, MIC Lab published the latest update (MITI 4.2.1) in Swedish.
Manuals and protocols
-
Training the coders
Coding at MIC Lab is done in the same way as in scientific trials of MI. Regardless of who codes at MIC Lab, the same assessment is made, and the assessment should not vary from year to year. Continuous training and reliability testing are required to work as a professional coder at MIC Lab.
Training of coders
Coding at MIC Lab is done in the same way as in scientific trials of MI. Regardless of who codes at MIC Lab, the same assessment is made. The assessment should not vary from year to year Continuous training and reliability testing are required to work as a professional coder at MIC Lab.
Training history
To become a professional MITI coder at MIC Lab, a basic training is required and is evaluated with a reliability control. Continuous training is then required. In 2005, when MIC Lab was established as a research group at Karolinska Institutet, the first coders were trained at the University of New Mexico, USA, by Dr. Theresa Moyers and colleagues who developed MITI. In 2005, Dr. Moyers and Dr. Ernst trained the coders at MIC Lab. In 2006, Dr. Denise Ernst and Dr. Carolina Yahne provided training, and in 2007, MIC Lab coders were trained in MITI and CLAMI (a precursor to CLEAR) by Dr. Ernst. In 2015, Dr. Ernst trained MIC Lab coders in MITI 4.2.1.
Training meetings
The coders at MIC Lab train together for 2 hours every week. During the meetings, the same conversations are coded independently of each other. Difficulties in coding are discussed. In the case of difficult assessments not covered by the MITI manual, addenda are developed as supplements to the manual. During the training meetings, the pedagogy of providing feedback on coding results to the person who had the conversation is developed and improved.
Inter-coder reliability
A measure of reliability for each of the coders is obtained each year by coding randomly selected conversations received by MIC Lab. The reliability between coders for each variable in MITI has been found to be between 'good' and 'excellent' according to the statistical guidelines provided by Cicchetti (1994).
Quality assurance of operations
Different organizations, such as institutions, treatment centers, health centers, and the local environmental office, can have their MI practice quality assured. Quality assurance is done at two different levels: 'Scientific assessment of MI competence' and 'Monitoring of MI competence'. Please contact MIC Lab for more information on quality assurance of operations. kontakt@miclab.se
Scientific assessment of MI competence
For 'Scientific assessment of MI competence', the same procedures as in scientific trials of MI are required. This type of quality assurance is used to demonstrate that the MI being performed is evidence-based. The organization records all conversations where MI is used during a certain period of time. Three randomly selected conversations per quarter and per counselor are then coded by MIC Lab.
After the quality assurance is completed, a certificate is issued and MIC Lab produces a brief report in which the coding results are anonymized and summarized so that no individual counselor can be identified. The MI competence of the counselor group is presented and related to recommended competence levels for basic competency and proficiency in MI according to MITI.
The quality assurance can be combined with counselors receiving feedback on their performances either through the organization providing supervision based on the coding protocols or by MIC Lab providing counselors with feedback over the phone on one or more coded conversations. The counselor is given the opportunity to discuss the conversation and the coding directly with the coder.
Monitoring of MI competence
For 'Monitoring of MI competence', one conversation per counselor and per quarter is sent in for coding. The counselor or organization chooses which conversation to submit to MIC Lab. After the quality assurance is completed, MIC Lab provides a certificate that the organization is continuously receiving independent feedback on its MI practice.
Monitoring of MI can be combined with counselors receiving feedback on their performances either through the organization providing supervision based on the coding protocols or by MIC Lab providing counselors with feedback over the phone on one or more coded conversations. The counselor is given the opportunity to discuss the conversation and the coding directly with the coder.