Acceptability Study of Emotional Expressions of In-Vehicle Virtual Assistant

We analyzed the differences in drivers' preferences for emotional expressions from an in-vehicle virtual assistant based on data collected from 823 respondents in five countries categorized into three different cultural settings: collectivism (Korea, China and Brazil), individualism (USA) and a mixture of both (Russia). To quantify the degree of acceptance, we used a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree) on four emotion types categorized based on Thayer's two-dimensional emotion model. Analysis revealed that respondents in all five countries preferred emotion type 1 (pleased, happy, excited) and emotion type 4 (calm, peaceful, relaxed), that is, positive emotions under Thayer's two-dimensional emotion model expressions, and they showed low acceptance for emotion type 2 (nervous, angry, annoying) and emotion type 3 (sad, bored, sleepy), negative emotional expressions. In a cross-cultural comparative analysis, Brazil and China ranked highest in both positive emotion types (1 and 4), and Korea ranked highest in negative emotion types.