Full length article Risk assessment in e-commerce: How sellers' photos, reputation scores, and the stake of a transaction influence buyers' purchase behavior and information processing Yue (Nancy) Dai a, * , Gregory Viken b , Eunsin Joo c , Gary Bente d a Department of Communication, Michigan State University, 404 Wilson Rd., CAS554, 473 Communication Arts & Sciences Bldg., East Lansing MI 48823 USA b Department of Communication Studies, Campbell University, PO Box 567, Buies Creek, NC 27506 USA c Department of Advertising and Public Relations, Michigan State University, 404 Wilson Rd., Communication Arts & Sciences Bldg., East Lansing MI 48823 USA d Department of Communication, Michigan State University, 404 Wilson Rd., CAS478, Communication Arts & Sciences Bldg., East Lansing MI 48823, USA article info Article history: Received 9 October 2017 Received in revised form 19 February 2018 Accepted 24 February 2018 Available online 25 February 2018 Keywords: Stake Trust Ecommerce Risk assessment Eye movement abstract Guided by Nickel's (2009) model of risk assessment and the literature on facial trustworthiness, this study investigates how the stake of a transaction interacts with information on buyers' profiles in influencing buyers' purchase decisions and information processing. Participants played buyers in a trust game and made purchase decisions based on a series of seller profiles while their eye movements on the stimuli were recorded. Results revealed that the three factors examined exerted influences on buyers' decision-making in a hierarchical fashion: Sellers' reputation exerted a primary influence on buyers' decision-making, followed by sellers' profile photos, which is further followed by the stake of a trans- action. The results confirm Nickel's (2009) model of risk assessment and inform e-marketing strategies in terms of building consumers' trust. © 2018 Elsevier Ltd. All rights reserved. 1. Introduction The internet has transformed the way business is performed. One important change has been the emergence of C2C (consumer- to-consumer) transactions, where one individual directly sells products or services to another individual on ecommerce plat- forms. In such commercial encounters, buyers face a series of risks regarding finance, product quality, delivery time, information se- curity, and so forth (Harridge-March 2006; Masoud, 2013). Before committing to a transaction, the buyers need to estimate a seller's likelihood of fulfilling the order and make a decision on whether or not to make a purchase (Flanagin, 2007). Although lacking the non-verbal cues typically available in face- to-face interactions for forming interpersonal impressions (Short, Williams, & Christie, 1976; Siegel, Dubrovsky, Kiesler, & McGuire, 1986; Sproull & Kiesler, 1986), text-based ecommerce environments still offer a variety of cues for buyers (i.e., the trust- ors) to form impressions of sellers (i.e., the trustees). These cues range from user- or system-generated reputation scores to visual representations of a seller, such as photos or avatars. Previous research showed that these static cues had a great impact on buyers' decision-making (e.g., Bente, Baptist, & Leuschner, 2012; Bente, Dratsch, Rehbach, Reyl, & Lushaj, 2014; Flanagin, 2007). How individuals make decisions in risky situations has been the focus of much research in the fields of communication and media psychology (e.g., Bente et al., 2014; Bente et al., 2012), management (e.g., Jøsang & Presti, 2004), business (e.g., Thaw, Mahmood, & Dominic, 2009), and economics (e.g., Johansson-Stenman, Mahmud, & Martinsson, 2005). In the ecommerce literature, it has been widely recognized that perceived risk of online shopping constitutes a major factor that deters sales, and that creating situ- ations where trust is stronger than perceived risk is imperative to increasing sales (Thakur & Srivastava, 2015; Vos et al., 2014). To this end, much scholarly attention has been devoted to examining how to build sufficient trust from customers in order to mitigate the negative effects of perceived risk on purchases (see for a review, Harridge-March 2006). * Corresponding author. Michigan State University, 404 Wilson Rd., Communi- cation Arts & Sciences Bldg., East Lansing, MI 48823, USA. E-mail address: daiyue@msu.edu (Y. Dai). Contents lists available at ScienceDirect Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh https://doi.org/10.1016/j.chb.2018.02.038 0747-5632/© 2018 Elsevier Ltd. All rights reserved. Computers in Human Behavior 84 (2018) 342e351