Toward an Aggregate, Implicit, and Dynamic Model of Norm Formation: Capturing Large-Scale Media Representations of Dynamic Descriptive Norms Through Automated and Crowdsourced Content Analysis

Abstract

Media content can shape people’s descriptive norm perceptions by presenting either population-level prevalence information or descriptions of individuals’ behaviors. Supervised machine learning and crowdsourcing can be combined to answer new, theoretical questions about the ways in which normative perceptions form and evolve through repeated, incidental exposure to normative mentions emanating from the media environment. Applying these methods, this study describes tobacco and e-cigarette norm prevalence and trends over 37 months through an examination of a census of 135,764 long-form media texts, 12,262 popular YouTube videos, and 75,322,911 tweets. Long-form texts mentioned tobacco population norms (4–5%) proportionately less often than e-cigarette population norms (20%). Individual use norms were common across sources, particularly YouTube (tobacco long-form: 34%; Twitter: 33%; YouTube: 88%; e-cigarette long form: 17%; Twitter: 16%; YouTube: 96%). The capacity to capture aggregated prevalence and temporal dynamics of normative media content permits asking population-level media effects questions that would otherwise be infeasible to address.

Marijuana Content on Digital Media and Marijuana Use among Young People in the United States

“…Using an exogenous measure of exposure to topical messaging may be complicated by the fact that consumers' digital media patterns are recorded in cookies and search histories and used to fine-tune targeted marketing strategies. Nonetheless, our team has demonstrated in prior work that such measures are correlated with tobacco product sales and youth attitudes and beliefs about tobacco use (Berg et al, 2019;Liu et al, 2019). Third, using the VADER to calculate sentiment score may not accurately capture pro-versus anti-marijuana opinions; it reflects emotional valence in tweets, and some pro-marijuana tweets may convey negative sentiment, for example.…”

An Experimental Test of the Effects of Digital Content Permanency on Perceived Anonymity and Indirect Effects on Cyber Bullying Intentions

“…Although normative beliefs develop over time, often through social learning processes (Bandura, 1986), they are not static phenomena and can be situation specific (Huesmann & Guerra, 1997; Rimal & Lapinski, 2015). Furthermore, media play a significant role in impacting the formation of norms, and even subtle exposure to changing normative messages in the media can impact audience norms (Liu et al, 2019). Similarly, based on priming theory (Berkowitz, 1984), recent stimuli can change how norms are used (Huesmann & Guerra, 1997).…”

The Effects of Tobacco Coverage in the Public Communication Environment on Young People’s Decisions to Smoke Combustible Cigarettes

“…While the behavioral beliefs pathway of effects is examined here, the test of the pathway through descriptive norm beliefs requires a different dataset. Analysis of the norms path requires coverage estimating the volume of norm information embodied in the coverage; that is, how often smoking is portrayed in the messages ( Gibson et al, 2019 ; Liu et al, 2019 ) rather than the volume of messages with positive or negative valence. Siegel et al (2022) present evidence for these effects.…”

“…Analysis of the norms path requires coverage estimating the volume of norm information embodied in the coverage; that is, how often smoking is portrayed in the messages ( Gibson et al, 2019 ; Liu et al, 2019 ) rather than the volume of messages with positive or negative valence. Siegel et al (2022) present evidence for these effects. The third path, through self-efficacy beliefs, would require yet another re-coding of the coverage data focused on, for example, content which addressed how hard or easy it was to quit.…”