Preventing the next sext: A behavioral economic approach to understanding non-consensual nude photo sharing decisions in a high school community

Sexting refers to the sending or receiving of sexually explicit or suggestive images or video via any cyber platform (Hinduja & Patchin, 2012). Adolescents sext at alarming rates: Between 12-27% report involvement as a recipient or participant (Madigan et al., 2018). Many students are faced with the decision to forward a sext or not. This phenomenon holds numerous irreparable consequences for students, as distributing child pornography is a criminal offense. However, prevention strategies are a-theoretical and measurement is insufficient. The current study sought to understand the behavioral mechanisms that underlie sexting and provide a behavioral measurement of sexting risk using a behavioral economic framework. Subjects: A sample of 213 high school students were recruited from a Chicago-area high school. Partnerships: Peagram Consulting, an organization that provides in-school bullying prevention curricula facilitated participant recruitment. Research Design and Methods: All aims were addressed using cross-sectional quantitative and qualitative data. Analysis: Aim 1. Examined current real-world sexting prevalence rates, types, and context characteristics by utilizing survey methods that yield in-depth quantitative and qualitative data. Sample descriptive statistics for each variable were inspected and visualized. Grounded theory was used to analyze qualitative data and create a ground-up conceptual model of non-consensual sexting. Results separated the ways photos originate from the ways in which they’re forwarded non-consensually, and showed that most of the sample had received a sext in the last year. Most students see nude images on multiple platforms. Aim 2. Using an adaptation of the established Social Discounting Task, this aim examined social discounting of preferential and popular peers as predictors of real-world self-reported sexting behavior. For each, a receiving operator curve (ROC) analysis was used to examine the accuracy of social discounting rate S (yielded by the social discounting task) in predicting whether or not a student has reported forwarding sexts in the past. Qualitative data provide context regarding gender and participant individual differences. Social discounting across the preference axis predicted sexting behavior significantly better than change, results were mixed regarding the popularity discounting. Aim 3. Test a novel Sext Discounting Task to capture hypothetical sexting choices. To examine systematicity of data yielded, all discounting data was assessed for orderliness and cleaned according to guidelines by Johnson and Bruner (2008), and was fairly systematic throughout the sample. Within-individual correlations of sext discount rates with the original discounting task and real-world self-reported sexting measures was examined, and showed higher correlations among within- modality measures than within construct measures. A latent class analyses showed four profiles of sexting behavior where the largest class does not discount to a high degree, signaling altruistic behavior trends. Differences between the other three indicate a domain-general quality in social discounting, though identify sexting as a unique behavior. Choice tasks and social discounting theory hold promise for future work with other interpersonal processes.