What Causes the Differences in Communication Styles? A Multicultural Study on Directness and Elaborateness

What Causes the Differences in Communication Styles? A Multicultural Study on Directness and Elaborateness Juliana Miehle1, Wolfgang Minker1, Stefan Ultes2 1Institute of Communications Engineering, Ulm University, Albert-Einstein-Allee 43, 89081 Ulm, Germany 2Department of Engineering, University of Cambridge, Trumpington Street, Cambridge CB2 1PZ, UK juliana.miehle@uni-ulm.de, wolfgang.minker@uni-ulm.de, su259@cam.ac.uk Abstract With the aim of designing a Spoken Dialogue System which adapts to the user’s communication idiosyncrasies, we present a multi- cultural study to investigate the causes of differences in the communication styles elaborateness and directness in Human-Computer Interaction. By adapting the system’s behaviour to the user, the conversation agent may appear more familiar and trustworthy. 339 persons from Germany, Russia, Poland, Spain and the United Kingdom participated in this web-based study. The participants had to imagine that they are talking to a digital agent. For every dialogue turn, they had to read four different variants of the system output and indicate their preference. With the results of this study, we could demonstrate the influence of the user’s culture and gender, the frequency of use of speech based assistants as well as the system’s role on the user’s preference concerning the system’s communication style in terms of its elaborateness and its directness. Keywords: Spoken Dialogue System, User Adaptation, Communication Idiosyncrasies 1. Introduction For humans, speech is the most intuitive and most natu- ral way to communicate. Therefore, scientists and engi- neers aim to realise methods and systems that enable not only interpersonal communication but also interaction with machines through natural spoken language. Today, we are able to communicate with various computer applications via speech. However, the usability and acceptance of Spo- ken Dialogue Systems is rather low and in public opinion, such systems often fall into disrepute (Hempel, 2008). For Human-Human Interaction, it has been shown that peo- ple adapt their interaction styles to one another across many levels of utterance production when they communicate, e.g. by matching each other’s behaviour or synchronising the timing of behaviour (Burgoon et al., 2007; Niederhoffer and Pennebaker, 2002; Brennan, 1996; Pickering and Gar- rod, 2004; Nenkova et al., 2008). Moreover, various stud- ies suggest to adapt Spoken Dialogue Systems to the user in a similar way (Cassell and Bickmore, 2003; Forbes- Riley et al., 2008; Stenchikova and Stent, 2007; Reitter et al., 2006; Mairesse and Walker, 2010). By adapting the system’s behaviour to the user, the conversation agent may appear more familiar and trustworthy and the dialogue may be more effective. Therefore, current research focuses on user-adaptive Spoken Dialogue Systems, e.g. (Honold et al., 2014; Ultes et al., 2015; Casanueva et al., 2015). Pragst et al. (2015) specifically focus on the adaptiveness of Dialogue Management to the cultural background and the emotional state of the user. Our aim is to design a Spoken Dialogue System which adapts to the user’s communication idiosyncrasies. Accord- ing to various cultural models for Human-Human Interac- tion (Hofstede, 2009; Elliott et al., 2016; Kaplan, 1966; Lewis, 2010), different cultures prefer different communi- cation styles. Moreover, Burleson (2003) presents a study of culture …

Content truncated to 500 words