What Makes a Good Counselor? Learning to Distinguish between High-quality and Low-quality Counseling Conversations

39873

Authors

nan

Abstract

AbstractThe quality of a counseling intervention relies highly on the active collaboration between clients and counselors. In this paper, we explore several linguistic aspects of the collaboration process occurring during counseling conversations. Specifically, we address the differences between high-quality and low-quality counseling. Our approach examines participants’ turn-by-turn interaction, their linguistic alignment, the sentiment expressed by speakers during the conversation, as well as the different topics being discussed. Our results suggest important language differences in low- and high-quality counseling, which we further use to derive linguistic features able to capture the differences between the two groups. These features are then used to build automatic classifiers that can predict counseling quality with accuracies of up to 88%.

Keywords

nan

Full Content

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 926–935 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 926 What Makes a Good Counselor? Learning to Distinguish between High-quality and Low-quality Counseling Conversations Ver´onica P´erez-Rosas1, Xinyi Wu1, Kenneth Resnicow2, Rada Mihalcea1 1Computer Science & Engineering, University of Michigan, USA 2School of Public Health, University of Michigan, USA {vrncapr,wuxinyi,kresnic,mihalcea}@umich.edu Abstract The quality of a counseling intervention re- lies highly on the active collaboration between clients and counselors. In this paper, we ex- plore several linguistic aspects of the collab- oration process occurring during counseling conversations. Specifically, we address the dif- ferences between high-quality and low-quality counseling. Our approach examines partici- pants’ turn-by-turn interaction, their linguistic alignment, the sentiment expressed by speak- ers during the conversation, as well as the dif- ferent topics being discussed. Our results sug- gest important language differences in low- and high-quality counseling, which we further use to derive linguistic features able to capture the differences between the two groups. These features are then used to build automatic clas- sifiers that can predict counseling quality with accuracies of up to 88%. 1 Introduction An increasing number of people are suffering from behavioral health problems, including substance abuse, smoking cessation, or eating disorders. Be- cause behavior is something that is generally seen as changeable, behavioral counseling has emerged as an important strategy for helping people to iden- tify and potentially change self-destructive or un- healthy behaviors (Rollnick et al., 2008). Despite its practical benefits, such as combating addiction and providing broader disease preven- tion and management, the process behind success- ful behavioral counseling has not been fully elu- cidated (Moyers et al., 2009), which in turn raises the question of what makes a good counselor. Seeking to answer this question, our paper ex- plores differences between high- and low-quality counseling conversations. Since the quality of a counseling intervention relies highly on the active collaboration between clients and coun- selors (Gaume et al., 2009; Vader et al., 2010), we explore which aspects of the collaboration pro- cess occurring between counseling participants are related to counseling quality. Our categorization of counseling quality relies on general counsel- ing principles taken from the literature on client- centered counseling (Miller and Rollnick, 2013). Thus, conversations where the counselor centers on the client and expresses empathy are consid- ered as high-quality interactions. In contrast, con- versations where the counselor mainly provides instruction and advice, and the client complies are regarded as low-quality interactions. Our work makes three important contributions. First, we explore the use of noisy counseling data obtained from public sources for the analysis of counseling quality. Second, we conduct exten- sive analyses on conversational aspects such as turn-by-turn interaction, the sentiment expressed during the interaction, linguistic alignment, and salient topics during the conversation to obtain in- sights into what are the patterns of high-quality counseling. Hence, identifying specific conversa- tional strategies used by good counselors. Third, we show that features derived from our analyses, …

Content truncated to 500 words

Labeled Text

AbstractThe quality of a counseling intervention relies highly on the active collaboration between clients and counselors. In this paper, we explore several linguistic aspects of the collaboration process occurring during counseling conversations. Specifically, we address the differences between high-quality and low-quality counseling. Our approach examines participants’ turn-by-turn interaction, their linguistic alignment, the sentiment expressed by speakers during the conversation, as well as the different topics being discussed. Our results suggest important language differences in low- and high-quality counseling, which we further use to derive linguistic features able to capture the differences between the two groups. These features are then used to build automatic classifiers that can predict counseling quality with accuracies of up to 88%.

Publication Details

Paper ID:
39873
Year:
2019
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan