The comparison analyses paralleled those described in the prior report analyzing suicide rates in relation to medical marijuana policy (Anderson et al., 2014). Data from years 1990–2007 were analyzed; the most basic model included medical marijuana policy and categorical indicators for state and year. State covariates were added in the second model, and state-specific linear time trends were added in the third model. State-specific linear time-trends are modeled as state by year interactions, with year specified as a continuous, rather than a categorical variable.
