Assessing the Design Choices for Online Recommendation Agents for Older Adults: Older Does Not Always Mean Simpler Information Technology1

Grounded in the aging and complexity literature, this experimental study investigated the moderating role of individuals’ cognitive age on the impact of recommendation agent (RA) comprehensiveness (i.e., amount of detail involved in using an RA) on users’ perceptions regarding RA complexity and RA usefulness. An experiment involving 140 online shoppers was conducted to understand the experiences of cognitively younger and older adults while using low or high comprehensiveness RAs designed for this study. Results reveal the tension that exists for older adults when using highly comprehensive RAs, as they perceive them to be more complex but also more useful in providing recommended products. The finding that cognitively older adults perceive high comprehensiveness RAs to be more useful compared to low comprehensiveness RAs provides a novel insight to the information systems literature, as it is contrary to the prevalent belief that “the older the user, the simpler the information technology should be.” Theoretically, this study improves our understanding of how increasing levels of RA comprehensiveness differentially affects the perceptions of RA complexity and RA usefulness of users of different cognitive ages. For practitioners, the results provide important guidelines about the kind of RA that is appropriate for consumers with different cognitive ages.