Postdoctoral Fellowship: SPRF: Explaining Behavior at Different Levels of Abstraction

This award was provided as part of NSF’s Social, Behavioral and Economic Sciences Postdoctoral Research Fellowships (SPRF) program and SBE's Perception, Action and Cognition program. The goal of the SPRF program is to prepare promising, early career doctoral-level scientists for scientific careers in academia, industry or private sector, and government.

SPRF awards involve two years of training under the sponsorship of established scientists and encourage Postdoctoral Fellows to perform independent research. NSF seeks to promote the participation of scientists from all segments of the scientific community, including those from underrepresented groups, in its research programs and activities; the postdoctoral period is considered to be an important level of professional development in attaining this goal.

Each Postdoctoral Fellow must address important scientific questions that advance their respective disciplinary fields. Under the sponsorship of Tobias Gerstenberg at Stanford University, this postdoctoral fellowship award supports an early career scientist investigating how adults learn about and explain the causes of others’ behavior. Every time we explain somebody’s actions, we solve a basic puzzle; there are often many different kinds of explanations available.

We routinely draw on knowledge of other people’s beliefs and goals, as well as more stable traits and features of the situation to account for their actions. This complex machinery for making sense of others is foundational to everyday behavior and is at the heart of collaboration and communication. However, the processes by which people learn about mental states, traits, and situational pressures from others’ behavior and use this information to explain why they acted as they did remains poorly understood.

A better understanding of the conditions under which people draw particular inferences or prefer certain explanations will enhance clinical efforts to support people with impairments in social reasoning, will improve the development of human-understandable artificial intelligence, and will have applications in legal settings where explanations of others’ motives and agency play a key role in judicial outcomes.

To understand others’ actions, we need to flexibly combine information about the external environment with what we know about a person’s goals, beliefs, and dispositions. Longstanding debates in psychology have centered on the role that each of these causes plays in action explanation. In the current project, we propose that a single mechanism governs how people choose the best explanation of others’ actions: people update their beliefs about each candidate cause based on the evidence they observed and then mentally simulate what would have happened in relevant counterfactual situations in which the candidate cause had been different.

Participants favor causes as explanations for which they are more certain that the relevant counterfactual outcome would have differed from what actually happened. We develop and test this computational model of action explanation in experimental paradigms that allow precise control over an actor’s mental states, traits, and situational constraints.

We evaluate our model’s predictions against participants’ judgments in two kinds of experiments. We first examine when traits versus situational factors are chosen as explanations, and subsequently pit traits against mental states. By developing and testing a flexible computational framework of action explanation, our work sheds new light on important theoretical questions about how people understand the behavior of others.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.