Modeling the Probability of Fraud in Social Media in a National Cannabis Survey

Abstract

Cannabis legalization has spread rapidly in the United States. Although national surveys provide robust information on the prevalence of cannabis use, cannabis disorders, and related outcomes, information on knowledge, attitudes, and beliefs (KABs) about cannabis is lacking. To inform the relationship between cannabis legalization and cannabis-related KABs, RTI International launched the National Cannabis Climate Survey (NCCS) in 2016. The survey sampled US residents 18 years or older via mail (n = 2,102), mail-to-web (n = 1,046), and two social media data collections (n = 11,957). This report outlines two techniques that we used to problem-solve several challenges with the resulting data: (1) developing a model for detecting fraudulent cases in social media completes after standard fraud detection measures were insufficient and (2) designing a weighting scheme to pool multiple probability and nonprobability samples. We also describe our approach for validating the pooled dataset. The fraud prevention and detection processes, predictive model of fraud, and the methods used to weight the probability and nonprobability samples can be applied to current and future complex data collections and analysis of existing datasets.

Perceived Safety, Not Perceived Legality, Mediates the Relationship Between Cannabis Legalization and Drugged Driving

“…All demographic covariates were maintained in models (regardless of significance) to create comparability with previous analyses (Dutra et al, 2019). As described previously (Dever, 2018;Dutra et al, 2021), we used jackknife weights to calibrate groups of participants recruited by different methods to each other to ensure realistic standard errors (Dever, 2018;Dutra et al, 2021;Kott, 2019;Ridenhour & Kott, 2018). Then, we used demographic variables to calibrate the mail sample to the U.S. population (Dever, 2018;Dutra et al, 2021).…”

“…The methods we used to recruit the NCCS sample have been described previously (Dever, 2018;Dutra et al, 2021). Briefly, between August 2016 and June 2017, we recruited participants 18 years and older living in the United States via two address-based samples (ABS) and two social media samples.…”

“…We obtained 11,957 responses. After implementing standard fraud detection methods and using a regression model to detect additional fraudulent responses (Dutra et al, 2021), the social media sample totaled 4,779 participants. All procedures were approved by RTI International's Institutional Review Board.…”

“…All analyses were adjusted by state cannabis legal status to account for the effect of cannabis legal status on perceptions about the legality of DUIC. As described in a previous analyses (Dutra et al, 2019(Dutra et al, , 2021, in 2018, we categorized states as having legal recreational and medical cannabis, medical cannabis only, or neither at the time of survey administration based on their effective date. Laws had to be in effect 1 month or more by the time that the survey was completed online or received by mail (between 2016 and 2017).…”