論文ID: TJSKE-D-23-00004
Agent systems that support humans by providing useful information are becoming widespread. However, the agents cannot always provide the most useful information due to the limit of technology. The effect of provision of “neither-good-nor-bad” information on the perceived trustworthiness of agents has been investigated in previous studies. The results show that providing “neither-good-nor-bad” information is effective when users act to gain profits. These studies have focused on agents’ positive advice, which recommends a user to choose a specific option. In this paper, by contrast, we investigate its effect on agents’ negative advice, which recommends a user not to choose a specific option, on trustworthiness of the agent. Experimental results show that providing “neither-good-nor-bad” information in the style of negative advice is effective when users act to gain profits and to avoid losses of already obtained profits.
