The impact of interpersonal closeness cues in text-based healthcare chatbots on attachment bond and the desire to continue interacting: an experimental design
Year: 2018
Type: conference-paper
Abstract: Working alliance describes an important relationship quality between health professionals and patients and is robustly linked to treatment success. However, due to limited resources of health professionals, working alliance cannot always be promoted just-in-time in a ubiquitous fashion. To address this scalability problem, we investigate the direct effect of interpersonal closeness cues of text-based healthcare chatbots (THCBs) on attachment bond from the working alliance con-struct and the indirect effect on the desire to continue interacting with THCBs. The underlying research model and hypotheses are informed by counselling psychology and research on conver-sational agents. In order to investigate the hypothesized effects, we first develop a THCB codebook with 12 design dimensions on interpersonal closeness cues that are categorized into visual cues (i.e. avatar), verbal cues (i.e. greetings, address, jargon, T-V-distinction), quasi-nonverbal cues (i.e. emoticons) and relational cues (i.e. small talk, self-disclosure, empathy, humor, meta-relational talk and continuity). In a second step, four distinct THCB designs are developed along the continuum of interpersonal closeness (i.e. institutional-like, expert-like, peer-like and myself-like THCBs) and a corresponding study design for an interactive THCB-based online experiment is presented to test our hypotheses. We conclude this work-in-progress by outlining our future work.
DOI: 10.5167/uzh-158352
Authors: Kowatsch, Tobias, Nißen, Marcia, Rüegger, Dominik, Stieger, Mirjam Nadine, Flückiger, Christoph, Allemand, Mathias, von Wangenheim, Florian
Institutions: University of St.Gallen, ETH Zurich, University of Zurich
Language: English
Subfield: Social Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W2888240163
