Ageism under the COVID-19 pandemic and its cross-cultural variations

A computational social science approach

Ageism under the COVID-19 pandemic and its cross-cultural variations

A computational social science approach

This project is funded by the Interdisciplinary Research Seed Funding of the Faculty of Social Science at The Chinese University of Hong Kong.

Project Aims

The COVID-19 pandemic has made beliefs and attitudes toward older adults a central element of public discourse. In this context, this project is a timely initiative to understand the formation and evolution of aging stereotypes from a cross-cultural perspective. A long body of literature has discussed that agentic and communal values are crucial factors shaping aging stereotypes. Nevertheless, few studies have simultaneously used large scale cross-national data, and data collected outside of the laboratory, in examining this issue. This project uses novel computational methods, by combining automated text analysis of Twitter messages (study 1) and online factorial experiments (study 2), to examine the public discourse about older adults and “open the black box” of aging stereotypes. In study 1, we observe the evolution of agentic and communal content in Twitter messages across time and context. In study 2, we use crowdsourcing platforms to conduct online experiments in China and the United States. They seek to identify and untangle the elements contributing to the formation of stereotypes, and how individual-level agentic and communal values influence the consideration of these elements in the stereotype formation process.

Significance to Interdisciplinary Research

The project contributes to interdisciplinary research because it combines two research tools from computational social sciences (automated text analysis and an online factorial survey using crowdsourcing platforms) to answer two research questions: 1) How aging stereotypes develop in twitter before and during the COVID-19 pandemic and how these aging beliefs are shaped by cultural contexts (i.e. individualism vs. collectivism) through personal values (agency vs. communion)? 2) What sources contribute to the formation of aging stereotypes and how they are related to cultural values (individualism vs. collectivism) and personal values (agency vs. communion)?

Study 1 will use data from Twitter to examine associations between valence of aging stereotypes (warmth vs. competence), cultures (individualism vs. collectivism), and values (agentic vs. communal) based on unsupervised and supervised models. On one hand, with former method, we analyze the text to identify the latent structure of textual corpora, which is a method increasingly used in cultural sociology. On the other hand, the study of emotions and values in psychology have mainly relied on supervised models based on machine learning. We combine both traditions for a more comprehensive understanding of the text. Moreover, study 1 has an interdisciplinary theoretical contribution to examine the complexity of aging views, considering its sociological and psychological underpinnings. From a theoretical perspective, the discourses contained in the texts are aggregated to understand beliefs about older adults in what cultural sociologists have framed as public culture. By tracking public discourse, we examine how a moral shock (COVID-19 pandemic) could have affected shared beliefs about a vulnerable population. Thus, we directly speak to the literature that considers cultural change as a mainly local phenomenon and driven by cohort replacement. Our study will contribute to a more nuanced and balanced view of possibilities for cultural change. In addition, public disclosures of beliefs towards older adults during the pandemic can be explained from the perspective of psychology. Specifically, individuals’ beliefs toward outgroups can be assessed in terms of the warmth and competence aspects according to the stereotype content model (Fiske et al., 2002). Since older age groups are regarded as high in warmth but low in competence than younger age groups (Cuddy & Fiske, 2002), we can identify changes of ageism on Twitter across time and contexts. Moreover, we will assess how cultural values (individualism vs. collectivism) and personal values (agency vs. communion) impact contents of aging stereotypes.

Study 2 relies on psychological and sociological rationales to “open the black box” of aging stereotypes. We will further examine how different domains of life in old age lead to subjective social status, a common variable used in stratification beliefs, which is entailed by stereotypes. In study 2, targeting on participants from two cultures (individualism in the United States and collectivism in China), we will use an online factorial survey to experimentally manipulate cultural, social, economic, and physical elements and estimate their contributions to aging stereotypes formation, and how this process varies across individuals when considering agentic/communal values. With this study, the association between individual values and aging stereotype is examined through the lens of social psychology; moreover, cultural sociology provides insights to understand how beliefs about certain groups are formed and shaped by cultural values as rule-alike-structures. Using a factorial survey approach, we aim to provide further evidence in understanding how different elements suggested by study 1 are combined to explain stereotypes toward older people. In addition, we also move forward the literatures by incorporating the agentic and communal individual values into the understanding of how older age groups are represented in the social media discourse in times of COVID-19.


Chunyan May, Department of Psychology, The Chinese University of Hong Kong.

Xuqian Chen, Department of Psychology, The Chinese University of Hong Kong.

Chen Dan, Department of Sociology, The Chinese University of Hong Kong.

Francisco Olivos
Research Assistant Professor

My research interests include cultural sociology, social inequality, sociology of education and computational social sciences.