Inaccuracy Blindness in Collaboration Persists, even with an Evaluation Prompt

The tendency to believe and act on others' misinformation is documented in much prior work. This paper focuses on inaccuracy blindness, the tendency to take a collaborator's poor information at face value, which reduces problem-solving success. We draw on social psychological research from the 1970s showing that evaluative rating scales can prompt a change in perspective. In a series of studies, we prototyped and tested an evaluation prompt meant to encourage skepticism in participant detectives trying to identify a serial killer. In tests of the prototype, the prompt was partially successful in inducing skepticism (Exp. 1), but a larger study (Exp. 2) showed that, despite the evaluation prompt, participants' inaccuracy blindness persisted. This work, and the literature more generally, shows that the tendency to be misled by collaborators' inaccurate information remains a strong phenomenon that is hard to counteract and remains a significant challenge for the CHI community.