Support Factors for Families of Young Children in the COVID-19 Pandemic: Leveraging Open-Ended Survey Responses

• In Event: Risk and Resilience in Early Childhood During the COVID-19 Pandemic: From a Social Equity Perspective

Abstract

Introduction
The COVID-19 pandemic has unprecedently affected families with young children. Under rapidly changing social/economic environments, the lives of families with young children have become highly unpredictable due to employment insecurities, school/childcare disruptions, and socio-cultural instabilities (Liu & Fisher, 2022). These challenges largely increased caregivers’ emotional distress and disrupted child socioemotional development (Spinelli et al., 2020). In these unpredictable social contexts, it is critical to understand what actions policymakers and community leaders should take to mitigate families’ difficulties. Although traditional quantitative survey data obtained through validated measurement can provide valuable information, it may not be sufficient to track real-time experiences in an unprecedented, fast-changing environment. Open-ended, qualitative survey responses can help clarify ambiguities and identify opinions that researchers have not considered before (Roberts et al., 2014).
Research Questions
This study aimed to leverage both quantitative and qualitative survey data to understand 1) what factors families of young children have found helpful during different phases of the pandemic; 2) how these factors differ by families’ socio-demographic characteristics, and 3) which factors could effectively foster caregivers’ and children’s better well-being.
Study Population
We studied families with young children under six.
Methods
Data were drawn from a large, national, and longitudinal study that used frequent online surveys to assess pandemic's impact on early childhood development. Data were collected between April 2020 and April 2022, where 37,361 valid open-ended responses to the question “What is helping you and your family the most right now?” were obtained from 9,371 caregivers (all in English; 8.88 % Black; 16.16% Hispanic/Latino[a]; 34.15% below 200% FPL). Caregivers’ emotional distress was indicated by a composite of depression, anxiety, stress, and loneliness symptoms. Child well-being was assessed using two items from CBCL: “fussiness/defiance” and “fear/anxiety”.
A novel text-based analysis technique, Structural Topic Modeling (STM; Roberts et al., 2014), was used to extract supportive factors. STM has the unique advantage of incorporating quantitative covariates (i.e., race/ethnicity, income levels) into estimating topic prevalence using text-based data. Estimated topic prevalence obtained from the STM model was further combined with quantitative well-being data. A mediation model, then, was constructed to test which factors indirectly promoted child well-being through reducing caregivers’ emotional distress.
Results
Figure 1 presents the 23 identified supportive factors and two exemplary topic trends overs time (in correspondence with the U.S. COVID-19 infection rate and important government assistance distribution). Topics that have significant prevalence differences across race/ethnicity and income levels are presented in Figure 2. Accordingly, lower-income, Black, and Hispanic/Latino(a) families relied more on various financial assistance and religious beliefs/practices, while their higher-income and White counterparts suggested more supportive factors related to social support networks and outdoor activities. These comparisons revealed the disparities experienced by and distinctions of needed support among families from different income and racial/ethnic backgrounds. Lastly, the mediation model using extracted topic prevalence data suggested six supportive factors (indicated by * in Figure 1) that indirectly improved child well-being through reducing caregivers’ emotional distress. These factors mainly focused on financial assistance, positive attitudes, work flexibility, and social network support.

Authors

• Sihong Liu, Stanford University
  Presenting Author

• Sara Weston, University of Oregon

• Ian Shryock, University of Oregon

• Philip Fisher, Stanford University