Exploring informational versus identity-based approaches to reducing polarization
Abstract
There is growing concern about the rise of political polarization. Polarization within the ranks of the military threatens discipline, cooperation, and morale, and thus developing effective methods for depolarizing individuals is important for military readiness.Understanding polarization is also important for computational decision-making, as polarization has a major impact on decision-making and thus must be understood and incorporated into such models. Furthermore, polarization may bias the data used to train models in poorly understood ways. Prior attempts to depolarize, however, have had limited success. We argue this is because the science of polarization is widely misunderstood, with the emphasis traditionally being placed on motivated reasoning - people s tendency to interpret information in an identity-protective way, and thus to believe that accurate information has little effect on attitudes and decisions. We challenge this view and propose that reasoning processes are much less biased by motivation than typically claimed. Thuswe hypothesize that balanced, two-sided information can effectively moderate people s opinions by avoiding perceptions of bias. In Task 1, we will test this using online experiments exposing people to balanced educational content versus one-sided content favoringtheir position. We predict two-sided information will more effectively depolarize views. In Task 2, we will compare our informationintervention against a variety of identity-focused interventions in a mega-study. In Task 3, we will examine which interventions work best for whom, and how targeting of interventions can improve depolarization effects. These experiments will advance our understanding of how to effectively depolarize individuals, help inform computational models of decision-making about how to best capture the causes and effects of polarization and associated decision-making biases, and generate unique datasets that can be used to evaluate such models.This abstract is marked for public release
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 09, 2024
- Source ID
- N000142412601
Entities
People
- David G. Rand
Organizations
- Massachusetts Institute of Technology
- Office of Naval Research
- United States Navy