Think Once, Think Again: Exploring the Harm-Made Mind Effect Through Dual-Process Theory

Abstract

This paper presents results from a pre-registered online experiment (N = 317) on the harm-made mind phenomenon (i.e., people’s tendency to perceive mind in seemingly mindless agents, such as robots, when they are harmed). To address inconsistent findings on this phenomenon, we study the harm-made mind through a dual-processing perspective. We propose that the harm-made mind of a robot is (a) positively mediated by perceived capacity for pain and (b) negatively mediated by objectifying dehumanization. By manipulating information processing, we explore how different processing types (Type 1 vs. Type 2) impact the mediation of the harm-made mind through perceived capacity for pain and dehumanization. The results of the online experiment showed that when people gave rapid answers under cognitive load (i.e., when Type 1 processing was encouraged), watching a harmful interaction with a robot (compared to a no-harm condition) resulted in (a) a positive indirect effect on mind perception through perceived capacity for pain and (b) a positive rather than the assumed negative indirect effect on mind perception through a decrease in dehumanization. While the mediation effect of harm on mind perception through perceived capacity for pain was significant across processing types, the mediation effect of harm on mind perception through dehumanization was no longer significant when people were asked to deliberate on their answers (i.e., when Type 2 thinking was encouraged).