Falling behind again? Characterizing and assessing older adults' algorithm literacy in interactions with video recommendations

Abstract

Algorithms play a significant role in shaping our experiences of interacting with intelligent information systems but also inherit and amplify data biases, potentially leading to unfair decisions or discriminatory outcomes. This motivates us to investigate users' algorithm literacy, which covers the awareness and knowledge of algorithms and the skills to intervene in the operations of personalization algorithms when interacting with recommendation systems. Since vulnerable groups are more likely to suffer from the negative consequences of algorithmic decision-making, investigating algorithm literacy among such groups is critical. This study aims to examine older adults' algorithm literacy, who are often considered a vulnerable group and labeled as digital laggards in contemporary information society. The empirical evidence collected from 21 participants in in-depth interviews and cognitive mapping studies demonstrated that almost all participants are algorithm-aware to some extent and identified (1) three types of information and sources collected by algorithms in user understanding, (2) two paradigms of how respondents understand personalized recommendations, and (3) two sets of strategies they develop to employ algorithms for improving user experience. The findings shed light on designing human-centered intelligent information systems for unbiased personalization and developing a more inclusive AI-assisted society that equally benefits people of all ages.