Covering Scientific Uncertainty in Ongoing Research

Natalie Herbert, University of Pennsylvania


What should experts say to inform public health decision making when the available scientific evidence is uncertain? In this dissertation, I focus on this question through the lens of electronic cigarettes (e-cigarettes), a topic of ongoing conversations in public health around risk reduction despite uncertain long-term health effects. This dissertation measures two kinds of messages about scientific uncertainty in public dialogue: 1) conflict messages, which present contradictory conclusions from experts and emphasize disagreement, and 2) limited evidence messages, those presenting uncertainty in terms of what scientists are still investigating without emphasizing disagreement. Study 1 demonstrates the prevalence of these distinct uncertainty messages in a population of newspaper articles about e-cigarettes (N=376 articles from 2017 through mid-2018). The results of this content analysis compose the stimuli of the subsequent experimental study (Study 2, N=457 current and former smokers, 4 condition mixed design). Results demonstrates these two kinds of uncertainty messages produce divergent perceptions of experts as hypothesized: conflict increases perceptions of expert disagreement, and limited evidence increases perceptions of expert uncertainty. Based on this validation, a final experimental study was undertaken to examine the hypothesis that exposure to high levels of disagreement about e-cigarettes will spillover to decrease intentions to perform recommended healthy behaviors—a so-called “spillover” hypothesized in the previous literature. Study 3 focused on N=765 current and former smokers in 5 conditions: 2 (perceived disagreement: high or low) x 2 (perceived uncertainty: high or low) + 1 (no message). Results do not reveal spillover towards more distantly related recommended health behaviors (e.g., exercise, nutrition). Results do support backlash effects after exposure to conflict, including decreased intentions to try e-cigarettes for smoking cessation (among current smokers). Conclusions from this series of studies merit further investigation of the messages available to experts communicating about scientific uncertainty, especially preceding public health crisis. Results of this work support further inquiry evaluating the cumulative impact of exposure to conflict messages over time, particularly as these cumulative effects may diverge from other ways of presenting scientific uncertainty.

Subject Area

Communication|Public Health Education|Information science|Public health

Recommended Citation

Herbert, Natalie, "Covering Scientific Uncertainty in Ongoing Research" (2020). Dissertations available from ProQuest. AAI27997436.