Development and Initial Validation of Marijuana Identity Implicit Associations Tests among Late Adolescents in Washington State

Abstract

The current climate surrounding adolescent marijuana use in the U.S. is facing unprecedented circumstances. Rates of daily use are at or near all-time highs and perceptions of risk are at an all-time low in the history of the Monitoring the Future study among 8th, 10th, and 12th graders. These rates are occurring despite research demonstrating worse long-term health outcomes associated with earlier age of marijuana use onset and increasing THC levels among marijuana products. As a result, there is an urgent need to identify risk factors that may represent screening markers of risk or targets for prevention and intervention among adolescents. One important risk factor for alcohol and tobacco is the extent to which one identifies with each substance. This aspect of identity can be measured with adaptations of the Implicit Association Test (IAT), a reaction time measure that aims to assess associations held in memory between constructs (e.g., marijuana and one’s self-concept). The aim of the current study was to develop and test two Marijuana Identity IATs among late adolescents in Washington State, one using images and another using words to represent marijuana and its control category. The current study included 169 adolescents between the ages of 15-18 (Mean age = 16.9, SD age = 0.9, 50% female, 66% high school student) with recruitment stratified by marijuana use (to include participants that range from non-users to heavy users) and gender. Data described here come from the online baseline assessment that included the Marijuana Identity IATs and self-report measures of marijuana use, consequences, and explicit (i.e., self-reported) marijuana identification. Results from the IATs reveal two normal distributions of IAT scores that were both positive on average indicating faster reaction times when marijuana was categorized with the self (and a neutral category categorized with other people). Split-half reliabilities of the IATs revealed internal consistencies in the range of previous substance-related IATs (word-based IAT, r = 0.52; imaged-based IAT, r = 0.40). In negative binomial regression models that controlled for age and sex, both IATs were significantly associated with use and consequences such that faster reaction times categorizing marijuana with the self were associated with more marijuana use and consequences (ps< .01). When controlling for self-reported identification marijuana, only the image-based IAT was significantly associated with marijuana use and consequences (ps< .05). The findings demonstrate relationships between IAT performance and marijuana use outcomes that compare favorably to past marijuana-related IATs lending support to implicit associations between the marijuana and the self as an important marker of marijuana use behaviors. Despite this promise, the relative inferiority of the internal consistency of these IATs to self-report measures may limit their utility as tools for screening. Future experimental and longitudinal research is warranted however, to examine identification with marijuana as a causal candidate for marijuana misuse to examine its potential as a prevention and intervention target.