How do Online Comments Affect Perceived Descriptive Norms of E-Cigarette Use? The Role of Quasi-Statistical Sense, Valence Perceptions, and Exposure Dosage

Journal of Computer-Mediated Communication How do Online Comments Affect Perceived Descriptive Norms of E-Cigarette Use? The Role of Quasi-Statistical Sense, Valence Perceptions, and Exposure Dosage Jiaying Liu Department of Communication Studies, University of Georgia, Athens, GA 30606, Georgia Rui Shi Department of Communication Studies, Rowan University, Glassboro, NJ 08028, New Jersey By facilitating user participatory features such as online comments, digital media expand the means through which individuals can get access to others’ behavior choices. This opens new research avenues in the pursuit of understanding how social influence operates in the virtual space. The current study examined whether anonymous others’ behavior choices within the online comment board may affect viewers’ descriptive norm perceptions in the real world. Results show that, given sufficient total exposure, viewers’ “quasi-statistical sense” allowed them to correctly identify the numerical majority through subtle individual behavior cues embedded in the online comments, which effectively influenced their estimation of the actual e-cigarette use prevalence among the U.S. population. Perceived behavior choice dominance and valence stance dominance toward e-cigarette use on the online comment board were found to mediate the relationship. Implications for the underlying mechanism of descriptive norm perception formation and future directions are discussed. Keywords: Online Comments, Social Influence, Descriptive Social Norms, Quasi-Statistical Sense, Exposure Dosage, E-Cigarette Use. doi:10.1093/jcmc/zmy021 Introduction While different subfields of social sciences take their own departure from the theoretically rich construct “social norms,” it is in general understood as shared customary rules or conventions, including values, standards, expectations, and traditions that function as a “social proof” heuristic and govern group behaviors as well as social order (Bicchieri, 2005; Fung & Scheufele, 2014; Lapinski & Rimal, Corresponding author: Jiaying Liu; e-mail: Editorial Record: First manuscript received on October 15, 2017. Revisions received on March 20, 2018, May 18, 2018 and July 24, 2018. Accepted by Mike Yao on September 10, 2018. Final manuscript received on September 16, 2018. First published online on 17 December 2018. Journal of Computer-Mediated Communication 24 (2019) 1–20 © The Author(s) 2018. Published by Oxford University Press on behalf of International Communication Association. All rights reserved. For permissions, please e-mail: 1 How do Online Comments Affect Perceived Descriptive Norms of E-Cigarette Use? J. Liu & R. Shi Individual behavior cues as normative information People frequently use two sources of information to understand descriptive social norms—systematic summary information, and scattered individual behavior cues (Tankard & Paluck, 2016). Summary information refers to prevalence statistics usually provided by census, survey results, newspaper reports, or educational campaigns, and is considered the most straightforward way to deliver descriptive norm information. In fact, most previous studies in the realm of social norms manipulate descriptive norm perceptions by directly providing summary information in the messages, such as “almost 75% of guests who are asked to participate in our new resource savings program do help by using their towels more than once” (Goldstein, Cialdini, & Griskevicius, 2008). While such descriptive norm information often reflects the actual behavior prevalence in the environment, it does not resemble the 2 Journal of Computer-Mediated Communication 24 (2019) 1–20 2005; Sherif, 1936). Cialdini, Reno, and Kallgren (1990) proffered two types of norms, injunctive and descriptive norms, that guide behavior decisions, with the former providing information about what ought or is expected to be done, and the latter describing what is actually and commonly done by social others. In other words, injunctive norms promote behavior compliance through expected social rewards or sanctions, whereas descriptive norms motivate actions through behavior prevalence information. Both norms are considered potent tools for influencing cognition and behavior changes (Fishbein & Ajzen, 2010). As social norms are unwritten rules in a society, and that individuals often have limited access and attention to a full picture of what others think and do, they tend to form subjective perceptions of social norms (i.e., perceived norms) based on their own observations. Such normative perceptions, accurate or not, matter much more than the actual norms in guiding decisionmaking (Lapinski & Rimal, 2005; Rimal & Real, 2003; Tankard & Paluck, 2016). Interestingly, while individuals can often interpret social approval or disapproval information correctly, their subjective perceptions of descriptive norms may not always match the actual behavior prevalence in the environment (Borsari & Carey, 2003; Cruz, Henningsen, & Williams, 2000; Neighbors, Dillard, Lewis, Bergstrom, & Neil, 2006; Prentice & Miller, 1993). However, it remains unclear how individuals construct their perceived behavior prevalence in their day-to-day life in the first place. Theories that explain the public opinion formation process, such as the Spiral of Silence Theory (Noelle-Neumann, 1974), may shed light on the underlying mechanism. One crucial premise of the theory stipulates that, people have a quasi-statistical sense that allows them to gauge the prevailing opinion climate through scrutinizing their social environment. In other words, people can instinctively collect opinion or behavior information through local observation and unique experiences, sense the dominance distribution, and act accordingly. The findings from the classic Asch experiments also corroborated this idea, by providing empirical evidence to suggest that individuals were able to correctly sense the majority answer in the group, and that such observations were quite precise as there was an increase in conformity with more confederates (Asch, 1955, 1956). As our media landscape continues to evolve, individuals have turned increasingly to social media and other online platforms which provide access to others’ opinions and behavior choices through user-generated content, such as online comments (Lee & Tandoc, 2017). This provides us unprecedented opportunities to examine the underlying process of descriptive norm perception formation with the new media platform where mediated and interpersonal processes intersect, extending the Asch paradigm from an off-line group setting to an online comment board setting. Therefore, in the current investigation, we aim to understand how exposure to online comments that contain anonymous others’ behavior choices may shape one’s descriptive norm perceptions, namely perceived prevalence of the target behavior in the real world, through their quasi-statistical sense. In the next sections that follow, w (...truncated)