Despite the strengths of this paper, including the use of a nationally representative sample, the availability of predicting and outcome variables by gender, and the inclusion of several substance use outcomes, the results of this study are to be interpreted with caution because of some limitations. First, the cross-sectional nature of this study limit sour ability to draw directional or causal conclusions. Second, missing data in communication and substance use variables and the subsequent exclusion of 251 participants (16.1%) from the present study might have slightly skewed the representativeness of the sample. Third, the present study relied on adolescents’ self-reports of their substance use behavior, which could introduce recall bias. Finally, our dataset did not have a variable on the frequency of contact with a parent. Therefore, we could not include this variable in our analyses. To address the first two limitations, future studies should examine the associations between parent-child communication and substance use outcomes longitudinally, and provide greater incentive to reduce missing data. In regards to the third limitation, our use of past 30 days substance use, rather than
