Using normative language when describing scientific findings: Study protocol for a randomized, controlled trial of effects on trust and credibility.
Academic Article
Overview
abstract
BACKGROUND: Trust in science and scientists has received renewed attention because of the 'infodemic' occurring alongside COVID-19. A robust evidence basis shows that such trust is associated with belief in misinformation and willingness to engage in public and personal health behaviors. At the same time, trust, and the associated construct of credibility, are complex meta-cognitive concepts that often are over-simplified in quantitative research. Discussion of research often includes both normative language (what one ought to do based on a study's findings) and cognitive language (what a study found), but these types of claims are very different, since normative claims make assumptions about people's interests. Thus, this paper presents a protocol for a large randomized, controlled trial to experimentally test whether some of the variability in trust in science and scientists and perceived message credibility is attributable to the use of normative language when sharing study findings in contrast to the use of cognitive language alone. OBJECTIVE: The objective of this trial will be to examine if reading normative and cognitive claims about a scientific study, compared to cognitive claims alone, results in lower trust in science and scientists as well as lower perceived credibility of the scientist who conducted the study, perceived credibility of the research, trust in the scientific information on the post, and trust in scientific information coming from the author of the post. METHODS: We will conduct a randomized, controlled trial consisting of 2 parallel groups and a 1:1 allocation ratio. A sample of 1,500 adults ages 18 and older who represent the overall US population distribution by gender, race/ethnicity, and age will randomly be assigned to either an "intervention" arm (normative and cognitive claims) or a control arm (cognitive claims alone). In each arm, participants will view and verify understanding of an ecologically valid claim or set of claims (i.e., from a highly cited, published research study) designed to look like a social media post. Outcomes will be trust in science and scientists, perceived credibility of the scientist who conducted the study, perceived credibility of the research, trust in the scientific information on the post, and trust in scientific information coming from the author of the post. Analyses will incorporate nine covariates. RESULTS: This study was approved by the Indiana University Institutional Review Board on August 2, 2022 and will be conducted without using any external funding mechanisms. CONCLUSIONS: If there is a measurable effect attributable to the inclusion of normative language when writing about scientific findings, it should generate discussion about how such findings are presented and disseminated. CLINICALTRIAL: This study was preregistered using the Open Science Framework (direct link to registration: https://osf.io/n7yfc). INTERNATIONAL REGISTERED REPORT: PRR1-10.2196/41747.
