Using Recommendation Agents to Cope with Information Overload
Muhammad Aljukhadar, Sylvain Senecal, and Charles-Etienne Daoust
International Journal of Electronic Commerce,
Volume 17 Number 2, Winter 2012-13, p. 41-70.
Abstract: Integrating the traditional and structural approaches to information load measurement, this research investigates the impact of information overload on decision strategy. Decision strategy refers to whether a consumer uses a recommendation agent and the consumer’s reactance behavior to the agent advice (whether the chosen product was the same or different from the recommended product). The research further shows the effects of information overload and decision strategy on choice quality, choice confidence, and e-store interactivity. The experiment, which involved 466 consumers, had three levels for the number of alternatives (6, 18, and 30), three levels for the number of attributes (15, 25, and 35), and two different attribute distributions across alternatives (proportional vs. disproportional). The results contribute to the literature of information overload and decision support systems by underscoring that (1) the relationship between information load and perceived overload is curvilinear, (2) information overload augments recommendation agent use and conformance to the recommendation, (3) the positive impact of using a recommendation agent on choice quality increases with information overload, and (4) consumers become more confident in their choices and perceive higher e-store interactivity when they conform to product recommendations. As such, the results help to explain some conflicting findings in the information overload literature and contribute to practice by highlighting the importance of decision aid tools in information-intensive environments.
Key Words and Phrases: Choice quality, decision strategy, heuristics, information overload, product recommendations, reactance behavior, recommendation agent.