Exploring Public Benefit Recipients' Information-Seeking Behavior during COVID-19: A Latent Class Analysis

Background/Purpose:

Public benefits are a vital form of assistance, designed to support individuals or families in need. Programs like SNAP, TANF, and Medicaid are crucial for providing healthcare, food, and income support to those in need. The COVID-19 pandemic has made these recipients more vulnerable, highlighting the need for accessible resources for both policymakers and service providers. However, research on how public benefit recipients seek information is scarce. This study examines their information-seeking behaviors and characteristics to improve communication strategies during health emergencies.

Methods:

Using survey data from the San Francisco Human Services Agency (SFHSA) targeting households receiving public benefits in May 2020 (n=10,089), we first conducted a latent class analysis of eleven dummy variables to identify whether respondents received information through various channels including television, radio, printed newspapers, online news, the SFHSA website, social media, family and friends, text, email, traditional mail, or other sources. Using the classification identified in the latent class analysis, a multinomial logistic regression analysis was then conducted to explore whether different classes of media users have different needs during COVID-19. The primary outcome variable was the participants’ most pressing need at the time of completing the survey. Covariates included age, race, education level, and whether English was the primary language.

Results:

Findings from latent class analysis revealed three distinct subgroups of participants, each exhibiting different behaviors in accessing information during the COVID-19 pandemic. Class 1 is the largest (N=6,931, 68.7%), followed by Class 3 (N=1,661, 16.47%), and Class 2 (N=1,497, 14.83%), indicating that the majority of the sample falls into the "Diverse Media User" category. Class 1, “Diverse Media Users,” used both traditional and digital media, preferring online news (54.08%) over radio (12.55%) and printed news (5.71%). Class 2, “Exclusive TV Users,” completely relied on television, reflecting a deep commitment to traditional media. Class 3, “Digital and Personal Network Users,” favor digital sources (online news at 84.35%, social media at 71.64%) and personal networks (81.73%), with less use of traditional media. Older individuals and English speakers prefer digital sources (Class 3) over traditional TV-focused media (Class 2), highlighting a shift towards digital and varied information channels. The Chi-square test shows a significant link between latent classes and immediate needs (χ² = 60.728, df = 12, p < 0.001), highlighting that specific needs vary significantly by class, underlining the importance of class membership in addressing the needs of public benefit recipients during the pandemic.

Conclusions/Implications:

Understanding media usage preferences and the impact of demographic characteristics provides essential insights for strategic communication and policy design. By identifying three distinct classes—from those exclusively focused on television, to versatile media users, and individuals centered on digital and network reliance—it highlights the varied landscape of information consumption across the population. Demographics shape media use, reflecting broader cultural and access disparities. This underscores the importance of considering age, primary language, education, and race in communication planning and community outreach.