Prior research has shown a robust effect of personalized product recommendations on user preference judgments for items. Specifically, the display of system-predicted preference ratings as item recommendations has been shown in multiple studies to bias users’ preference ratings after item consumption in the direction of the predicted rating. Top-N lists represent another common approach for presenting item recommendations in recommender systems. Through three controlled laboratory experiments, we show that top-N lists do not induce a discernible bias in user preference judgments. This result is robust, holding for both lists of personalized item recommendations and lists of items that are top-rated based on averages of aggregate user ratings. Adding numerical ratings to the list items does generate a bias, consistent with earlier studies. Thus, in contexts where preference biases are of concern to an online retailer or platform, top-N lists, without numerical predicted ratings, would be a promising format for displaying item recommendations.

Despite the popularity of badges in gamification applications, there is a lack of research on how to design badges to increase target behaviour. Motivated by this gap, we conduct a large-scale field experiment in a commuting-by-bicycle programme to explore efficacies of different badge designs in motivating ridership. We systematically vary the rewards, signifiers, and completion logic components of badges. We find adding an option for sharing a badge on Facebook, as a reward for badge attainment, increases ridership. Changing the badge signifier from a self-interested frame to a pro-environmental frame does not make a difference. Changing completion logic from a fixed to a relative goal increases ridership only among frequent riders. These findings have direct implications for gamification design and provide useful directions for research into the motivations behind the design elements.

This paper explores how explanations within an online recommender agent (RA) affect consumers’ trust in the RA and their willingness to adopt its recommendations. The paper proposes a theoretical research model which employs a multidimensional notion of trust and considers the effects of varying two features of the RA: explanation availability, i.e., providing, or not, an explanation of the underlying reasoning process of an RA; and explanation mode, i.e., using graphs or text as the means of providing explanations. In addition, we investigate the impact of a consumer’s perceived level of personalization of the RA as a measured variable. Two dimensions of trust—trust in integrity and trust in competence—along with intention to accept the recommendations serve as the dependent variables within the model. A within-subject experiment is conducted with four RAs. Empirical data provide evidence that explanation availability, explanation mode, and perceived personalization all have significant influence on consumers’ trust beliefs in an RA and willingness to adopt its recommendations. Effects of both the availability and mode of explanations on consumers’ trust beliefs are found to be mediated by consumers’ perceived personalization of the RA that, in turn, mediate the effects on intention to use.