Answers are grounded in this paper and cite their sources.
Reflexive thematic analysis uncovers how generative AI supports autonomy and problem-solving in older adults, highlighting cognitive rather than emotional pathways to healthy ageing.
Purpose This paper aims to offer a worked methodological demonstration of an integrated socioemotional selectivity theory (SST)–self-determination theory (SDT) reflexive thematic analysis (RTA), examining how urban, community-dwelling older adults in India are illustrated as using and emotionally experiencing generative AI and how this relates to autonomy, competence, relatedness and healthy ageing. Design/methodology/approach An RTA, organised using a Gioia-style structure, was conducted on an illustrative data set of 16 urban older adults (aged 61–83, mean 70.6) across 10 Indian cities: 5 full transcripts and 11 condensed summaries. As the data set was constructed for teaching purposes rather than collected under ethics-approved fieldwork, findings are a hypothesis-generating exercise rather than generalisable evidence. Findings Six aggregate dimensions were derived: everyday instrumental and cognitive integration; emotional relief primarily through reduction of uncertainty rather than direct companionship; differentiated relational meaning ranging from simulated companionship to explicitly non-relational, intellectual partnership; autonomy and competence gained through self-directed problem-solving; a near-universal positioning of AI as bounded against human relationships; and countercurrents of privacy unease, calibrated trust, literacy barriers and fear of overdependence. The integrated SST–SDT chain fit unevenly: several accounts were better explained by an informational, cognitive-appraisal pathway than an emotional one, and relatedness, unlike autonomy and competence, was not consistently reported. Originality/value This paper provides a transparent, reflexive worked example of applying an integrated SST–SDT framework with Gioia-structured RTA to a novel, under-researched population, and specifies what a genuine empirical study in this domain would require.
No takes yet. Share an insight, caveat, or question.
Godse et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: