Theses Bachelor's

The Effects of Political Attack Rhetoric on Public Trust in Scientists

Sayegh, Helen

Throughout the COVID-19 pandemic, the U.S. has experienced relatively sluggish public adoption of spread-reduction measures. There is a growing narrative that political rhetoric undercutting the scientific community and its recommendations is the very reason compliance with public health recommendations has been so challenging to secure. Further, there is concern that this brand of rhetoric hurts public trust in scientists on a general level. This thesis represents an attempt to test the causal link identified in the statements above.

Research questions: Does political rhetoric attacking scientists and their recommendations truly affect public perceptions of competence, beneficence, and trustworthiness? And does that rhetoric indeed reduce compliance with recommendations?

The predominant hypotheses of this thesis are as follows:
1) Rhetoric that attacks and/or undercuts scientific sources will negatively impact public perception of both scientists—the “attacked”—and politicians—the “attackers”—on dimensions of competence, beneficence, and trustworthiness.

2) When receiving public health messaging, attitude expressed toward science is consequential—i.e., it matters—to listeners.

These questions and hypotheses were explored and tested via a two-part survey experiment launched on Amazon’s Mechanical Turk. The first part of the survey leverages a 23 factorial design, in which the factors are: 1) factual consensus; 2) speaker party; and 3) ad hominem praise/attack behavior. After exposure to stimuli varying along these dimensions, participants rated their perceptions of the trustworthiness, beneficence, and competence of a scientist and politician. The second part of the survey leverages a conjoint design, in which the attributes are 1) Source; 2) Justification; and 3) Attitude toward science. Participants were presented with a public health recommendation followed by speaker profiles varying randomly along the attributes above. They were tasked with selecting which speaker they would be most willing to listen to if deciding whether or not to follow the public health recommendation.

Through the factorial experiment, Hypothesis (1) was partially supported. Attack rhetoric negatively impacted participant perceptions of politicians (“attackers”), but had minimal effect on participant perceptions of scientists. However, subgroup analysis revealed that certain characteristics make participants more susceptible to be swayed in their opinions by attack rhetoric.

Through the conjoint experiment, Hypothesis (2) was supported. The attitude speakers expressed toward scientists and their role in public health policy-making was the most important relative factor for respondents when choosing between profiles. Respondents preferred scientific speakers to political ones, and had a strong preference for speakers who felt either neutral or very positive toward scientists in the policy-making arena.


  • thumnail for Sayegh_Helen_FINALTHESIS.pdf Sayegh_Helen_FINALTHESIS.pdf application/pdf 2.86 MB Download File

More About This Work

Academic Units
Political Science
Thesis Advisors
Shapiro, Robert Y.
B.A., Columbia University
Published Here
April 19, 2021