Measuring Attitudes Toward Cannabis: An Investigation of the Link Between Risk of Hazardous Use and Implicit and Explicit Cannabis-Harm Associations

Although existing research demonstrates that beliefs about the harmfulness of cannabis relate to cannabis-related behaviors, there has been very little work evaluating an implicit measure of cannabis-harm associations. In a large, preregistered study (N = 9,995 US adults), we investigated the internal reliability and predictive validity of an Implicit Association Test (IAT) designed to examine cognitive associations between cannabis and harm. We did an internal replication, dividing our sample into exploratory (35%) and confirmatory (65%) subgroups. Results demonstrate the internal reliability (r’s = 0.64, p < .001, CI95% = [0.63, 0.66]) and predictive validity of the IAT. Indeed, over and above explicit-cannabis harm-associations, implicit cannabis-harm associations predicted: (1) greater odds of reporting that one has not used cannabis in the prior six months (OR = 1.45, CI95% = [1.35, 1.55], p < .001) and (2) decreased risk of self-reported hazardous use (IRR = 0.95, CI95% = [0.92, 0.97], p = <.001). This research demonstrates the potential utility of measuring implicit and explicit cannabis-harm associations when predicting risk of hazardous use.