Abstract
Rationales:
Using AI for health information seeking is a novel behavior, and as such, developing effective communication strategies to optimize AI adoption in this area presents challenges. To lay the groundwork, research is needed to map out users’ behavioral underpinnings regarding AI use, as understanding users’ needs, concerns and perspectives could inform the design of targeted and effective communication strategies in this context.
Objective:
Guided by the planned risk information seeking model and the comprehensive model of information seeking, our study examines how socio-psychological factors (i.e., attitudes, perceived descriptive and injunctive norms, self-efficacy, technological anxiety) and factors related to information carriers (i.e., trust in and perceived accuracy of AI), shape users’ latent profiles. In addition, we explore how individual differences in demographic attributes and anthropocentrism predict membership in these user profiles.
Methods:
We conducted a quota-sampled survey with 1,051 AI-experienced users in Hong Kong. Latent profile analysis was used to examine users’ profile patterns. The hierarchical multiple logistic regression was employed to examine how individual differences predict membership in these user profiles.
Results:
The latent profile analysis revealed five heterogeneous profiles, which we labeled “Discreet Approachers,” “Casual Investigators,” “Apprehensive Moderates,” “Apathetic Bystanders,” and “Anxious Explorers.” Each profile was associated with specific predictors related to individual differences in demographic attributes and/or aspects of anthropocentrism.
Conclusion:
The findings advance theoretical understandings of using AI for health information seeking, provide theory-driven strategies to empower users to make well-informed decisions, and offer insights to optimize the adoption of AI technology.
Using AI for health information seeking is a novel behavior, and as such, developing effective communication strategies to optimize AI adoption in this area presents challenges. To lay the groundwork, research is needed to map out users’ behavioral underpinnings regarding AI use, as understanding users’ needs, concerns and perspectives could inform the design of targeted and effective communication strategies in this context.
Objective:
Guided by the planned risk information seeking model and the comprehensive model of information seeking, our study examines how socio-psychological factors (i.e., attitudes, perceived descriptive and injunctive norms, self-efficacy, technological anxiety) and factors related to information carriers (i.e., trust in and perceived accuracy of AI), shape users’ latent profiles. In addition, we explore how individual differences in demographic attributes and anthropocentrism predict membership in these user profiles.
Methods:
We conducted a quota-sampled survey with 1,051 AI-experienced users in Hong Kong. Latent profile analysis was used to examine users’ profile patterns. The hierarchical multiple logistic regression was employed to examine how individual differences predict membership in these user profiles.
Results:
The latent profile analysis revealed five heterogeneous profiles, which we labeled “Discreet Approachers,” “Casual Investigators,” “Apprehensive Moderates,” “Apathetic Bystanders,” and “Anxious Explorers.” Each profile was associated with specific predictors related to individual differences in demographic attributes and/or aspects of anthropocentrism.
Conclusion:
The findings advance theoretical understandings of using AI for health information seeking, provide theory-driven strategies to empower users to make well-informed decisions, and offer insights to optimize the adoption of AI technology.
| Original language | English |
|---|---|
| Article number | 118059 |
| Number of pages | 50 |
| Journal | Social Science and Medicine |
| DOIs | |
| Publication status | E-pub ahead of print - 10 Apr 2025 |
User-Defined Keywords
- health information seeking
- planned risk information seeking model
- comprehensive information seeking model
- audience segmentation
- artificial intelligence
