This study, published in Frontiers in Psychology, utilizes Latent Profile Analysis to segment the Singaporean public into three distinct groups—the less-concerned, the risk-majority, and the risk-anxious—based on their COVID-19 risk perceptions and preventive behaviors. By characterizing these groups through personality traits and information preferences, the researchers provide a roadmap for moving beyond "one-size-fits-all" health messaging toward more targeted, effective interventions.
What Problem Does This Paper Address?
Effective crisis communication isn't just about broadcasting facts; it's about ensuring the public accepts and acts on them. However, "the public" is not a monolith. Traditional health communication research often relies on "variable-centered" approaches that assume a homogeneous population (e.g., assuming a rise in "perceived risk" increases "mask-wearing" equally for everyone).
This paper addresses the need for audience segmentation—identifying subgroups with shared attitudes and behaviors—to understand why some people ignore guidelines while others are paralyzed by anxiety. Specifically, it looks at the Singaporean context during the Delta variant wave to see how knowledge, emotion, and risk perception cluster together to drive preventive actions.
Methods
The researchers employed Latent Profile Analysis (LPA), a data-driven, "person-centered" analytical method. Unlike standard regression, which looks at how one variable affects another across an entire sample, LPA identifies hidden (latent) subgroups based on similar patterns across multiple indicators.
The authors used nine "determinant" measures for the segmentation:
- Preventive behaviors (e.g., handwashing, social distancing).
- Perceived knowledge of the virus.
- Risk perception (perceived severity and vulnerability).
- Perceived preventive efficacy.
- Emotional responses (positive, anxious, and sympathetic).
Following the identification of these profiles, the authors used ANOVA and Chi-square tests to characterize the groups across secondary variables, including the "Big Five" personality traits and information-processing styles.
Datasets
The study utilized a dataset of 2,033 Singaporean citizens recruited through local market research firms.
- Timing: Data was collected via a web-based questionnaire in August 2021 (during the Delta variant wave).
- Representativeness: The researchers used quota sampling based on Singapore’s 2020 population census to ensure the sample matched the national distribution for gender, age, and ethnicity.
Key Insights and Ideas
- The "Less-Concerned" Challenge: Roughly 32% of the sample fell into the "less-concerned" group. These individuals had the lowest risk perception and the lowest engagement in preventive behaviors. Crucially, they scored low on both Need for Cognition (NFC)—the tendency to enjoy effortful thinking—and Need for Affect (NFA). This suggests they are difficult to reach because they are less likely to be swayed by either purely factual or purely emotional messaging.
- Personality as a Compliance Predictor: The "risk-anxious" group (7%) wasn't just worried; they were the most conscientious, extraverted, and open to experience. This suggests that high compliance often correlates with specific psychological profiles that are more adaptable to changing regulations.
- Information Preferences and Fatigue: While all groups preferred official government sources, the "less-concerned" group had a significantly higher preference for informal sources like online forums and independent blogs. They were also more likely to read "conflict" stories (clashes in opinion), which might indicate a higher susceptibility to message fatigue.
- Social Contract vs. Individual Responsibility: The "risk-majority" (61%) placed the highest burden of responsibility for reducing transmission on fellow citizens, whereas the "risk-anxious" attributed high responsibility to all stakeholders (government, citizens, and healthcare workers) equally.
Key Results
The LPA identified a three-class solution as the most statistically robust model:
- The Less-Concerned (32%): Scored lowest across all risk and behavior metrics. They were more likely to be male and younger (ages 25–34).
- The Risk-Majority (61%): Represented the "compliant middle." Their scores hovered near the mean for most variables.
- The Risk-Anxious (7%): Scored highest in knowledge, perceived severity, and preventive action. This group was older (ages 55–74) and scored significantly higher on Openness ($M=35.18$) and Conscientiousness ($M=38.70$) than the other groups ($p < 0.001$).
Quantitatively, the "risk-anxious" group showed significantly higher intensity in information seeking ($M=4.05$ on a 5-point scale) compared to the "less-concerned" ($M=3.07$).
Limitations and Future Work
- Geographic Specificity: The study was conducted in Singapore, a country with high public trust and compliance. The "less-concerned" group in Singapore might still be more compliant than the average citizen in other regions, limiting the generalizability of the findings to more polarized societies.
- Cross-Sectional Design: The data is a snapshot in time. As the pandemic moved from the Delta wave to the "living with COVID" (Omicron) phase, these profiles likely shifted.
- Future Directions: The authors suggest that future research should investigate message fatigue and cyberchondria (anxiety caused by excessive online health searching), particularly for the "risk-anxious" segment, and explore how these different profiles respond to misinformation and fake news.
Summary updated