An ERP-study on the extent to which partisanship conditions the early processing of politicians' faces.

Abstract

Partisanship has been associated with various cognitive biases. These findings are primarily based on self-reports and task performance and less on measures of neural activity. We reviewed the literature on in-group vs. out-group bias that employs face-viewing paradigms and ERP methodology to investigate unconscious bias in politics. We subsequently preregistered hypotheses about the extent to which partisanship is associated with early neural processing of political leaders' faces. Our lab experiment was conducted in the Netherlands (N = 51), a multi-party democracy, and sufficiently powered to pick up modest effect sizes for in-party vs. out-party comparisons. As expected, we find that politicians' faces elicit a stronger N170 ERP response than strangers' faces, but we did not find the same pattern for the N250 component. Contrary to our hypotheses, we did not find statistically significant differences in the P200 and N200 components for the in-party vs. out-party comparison. These findings, supported by our cluster-based permutation analysis, indicate that seeing faces of political leaders enhances attention during facial processing, regardless of party affiliation, possibly due to their frequent and affectively salient presence in media. Since in-party vs. out-party differences did not emerge early on, implications for partisanship are discussed relative to racial and minimal group bias findings.