Predicting Counselor Behaviors in Motivational Interviewing Encounters

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 1128–1137, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Predicting Counselor Behaviors in Motivational Interviewing Encounters Ver´onica P´erez-Rosas1, Rada Mihalcea1, Kenneth Resnicow2, Satinder Singh1, Lawrence An3, Kathy J. Goggin4 and Delwyn Catley5 1Computer Science and Engineering, University of Michigan 2School of Public Health. University of Michigan 3Center for Health Communications Research, University of Michigan 4,5Department of Pediatrics. Children’s Mercy Kansas City, University of Missouri–Kansas City 1,2,3{vrncapr,mihalcea,kresnic,baveja,lcan}@umich 4,5{kjgoggin,dcatley}@cmh.edu Abstract As the number of people receiving psycho- therapeutic treatment increases, the au- tomatic evaluation of counseling practice arises as an important challenge in the clinical domain. In this paper, we address the automatic evaluation of counseling performance by analyzing counselors’ lan- guage during their interaction with clients. In particular, we present a model towards the automation of Motivational Interview- ing (MI) coding, which is the current gold standard to evaluate MI counseling. First, we build a dataset of hand labeled MI en- counters; second, we use text-based meth- ods to extract and analyze linguistic pat- terns associated with counselor behaviors; and third, we develop an automatic sys- tem to predict these behaviors. We intro- duce a new set of features based on seman- tic information and syntactic patterns, and show that they lead to accuracy figures of up to 90%, which represent a significant improvement with respect to features used in the past. 1 Introduction Effective behavioral counseling is an essential el- ement in combating public health issues such as mental health, substance abuse, and nutri- tion among others. A key component in train- ing addiction counselors and other health care providers is providing detailed clinical feedback and evaluation. In clinical psychotherapy, this is done through behavioral coding, a labor in- tensive and time consuming process that requires highly trained practitioners who observe the coun- seling interactions via audio/video or reading tran- scripts and, then provide detailed evaluative feed- back based on a set of predefined behaviors. Recently, research efforts have been made to- wards implementing automatic means to assist this process and provide clinicians with tools to code and analyze counseling narratives (Atkins et al., 2014; Xiao et al., 2014; Klonek et al., 2015). Such tools can enable analyzes at larger scale by provid- ing faster, cheaper, and more reliable methods for coding and data summarizing tasks. Following this line of work, this paper presents a text-based approach for the automatic coding of counselor’s verbal behaviors during counseling encounters. We focus our analysis on counseling conducted using Motivational Interviewing (MI), a well established evidence-based psychotherapy style, and the Motivational Interviewing Treat- ment Integrity (MITI) coding scheme. 2 Background on Motivational Interviewing Miller and Rollnick define MI as a collaborative, goal-oriented style of psychotherapy with partic- ular attention to the language of change (Miller and Rollnick, 2013). MI has been widely used as treatment method in clinical trials on psychother- apy research to address addictive behaviors such as alcohol, tobacco …

Content truncated to 500 words