The perception of dramatic risks: Biased media, but unbiased minds

Area of Science:

• Cognitive Psychology
• Risk Perception Research
• Media Studies

Background:

• A foundational 1978 study suggested people overestimate dramatic risks (e.g., homicide) and underestimate common risks (e.g., heart disease).
• This bias was linked to media overrepresentation of dramatic events, a conclusion widely accepted in social sciences.
• The reliability and replicability of these findings have been questioned due to a lack of rigorous statistical evaluation in the original study.

Purpose of the Study:

• To systematically evaluate the replicability of the original findings on risk perception and media coverage.
• To statistically analyze both original and replication datasets concerning dramatic versus nondramatic risk judgments and media portrayals.
• To investigate the sources influencing people's risk frequency judgments, specifically media versus personal social networks.

Main Methods:

• Conducted a systematic literature search to identify replications of the Lichtenstein et al. (1978) study.
• Performed Bayesian statistical analysis on the original dataset and data from identified replications.
• Analyzed media coverage and compared it with people's risk frequency judgments across datasets.

Main Results:

• All analyzed datasets showed strong evidence for media overrepresentation of dramatic risks and underrepresentation of nondramatic risks.
• A significant overestimation of dramatic risks and underestimation of nondramatic risks in people's judgments was found only in the original dataset, not in replications.
• Aggregated data across all studies indicated no reliable differential distortion in people's risk frequency judgments, suggesting the absence of a media-driven bias.

Conclusions:

• The common belief that public risk perception distortions mirror media coverage distortions has a limited empirical basis.
• People's risk frequency judgments may rely more heavily on samples from personal social networks than on media portrayals.
• Risk judgment processes involve metacognitive mechanisms sensitive to the source of mentally available information, such as social networks.