ABSTRACT
Social service robots have been used in hospitals as a technological solution for routine patient tasks, such as medicine delivery. In such application, robot speech capability has been found to be a critical interface feature in the effectiveness of patient robot interaction (PRI). The objective of this study was to review a human linguistic etiquette model in the literature (by Brown and Levinson) and investigate its applicability to PRI. The mediating effect of robot physical humanoid features on user perceptions of linguistic etiquette strategies (derived from the model) was also assessed. Results revealed the etiquette model could be partially extended to PRI. Subjects interpreted and understood robot negative utterances consistent with the original model; however, consistency was not found for positive etiquette strategies. With respect to user overall perception of robot etiquette (PE), negative and positive strategies resulted in the highest and lowest PE scores, respectively. However, the effects of strategy on PE were not mediated by simple humanoid feature changes in robot appearance. Overall, this study provides a basis for
detennining appropriate robot etiquette strategies to enhance/improve user experiences in collaborative task scenarios with robots in a healthcare context.
