Subject Know Thyself? Comparing Self-Reported and Observed Emotions and Their Influence on Political Attitudes

24 Pages Posted: 4 May 2016

See all articles by Cherie D Maestas

Cherie D Maestas

University of North Carolina Charlotte

JoEllen V Pope

UNC Charlotte

Date Written: January 7, 2016

Abstract

Do subjects’ self-reports of emotion correspond to observable evidence of emotion when subjects are exposed to complex emotional stimuli such as video news reports? We explore the answer to this question with a pilot study that uses software that machine codes facial expressions of subjects to detect discrete emotions. Subjects were randomized to receive a positive, situationally-framed news story about drone warfare or a negative episodically-framed news story about drone warfare followed by a series of standard emotion self-report questions. The data produced by automated facial action coding permits a set of nuanced tests to link the ordering complex stimuli to self-reports of emotion and to subsequent opinions about drone warfare. We find evidence that self-reports and observed emotions are weakly related and observed emotion predicts attitudes about drone warfare independent of self-reports. This suggests that self-reports, while useful, do not capture the full extent of emotional experiences that shape political attitudes.

Keywords: automated facial action coding, emotion, public opinion, foreign policy opinions

Suggested Citation

Maestas, Cherie D and Pope, JoEllen V, Subject Know Thyself? Comparing Self-Reported and Observed Emotions and Their Influence on Political Attitudes (January 7, 2016). Available at SSRN: https://ssrn.com/abstract=2775190 or http://dx.doi.org/10.2139/ssrn.2775190

Cherie D Maestas (Contact Author)

University of North Carolina Charlotte ( email )

Dept. of Political Science & Public Administration
9201 University City Blvd
Charlotte, NC 28223
United States

JoEllen V Pope

UNC Charlotte ( email )

Charlotte, NC
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
34
Abstract Views
369
PlumX Metrics