Effects of Communicating Issue Priority for Preference Tradeoffs in Agent-human Negotiations

This study explores the consequences of a software agent volunteering one of its issue priorities and inviting its human counterpart to reciprocate in order to obtain preference tradeoffs in a multi-issue e-commerce negotiation scenario. Results indicated that while the agent followed the same negotiation strategy algorithm, agent-human dyads achieved better agreement rate and joint outcomes, and the human negotiators experienced more positive post-negotiation affect when the agent volunteered to communicate an issue priority than when it didn’t. The findings have implications for automated negotiation research and practice.