Sociology, Department of


Date of this Version



Published in The American Journal of Drug and Alcohol Abuse 2020, VOL. 46, NO. 4, 401–407



Copyright © 2019 Taylor & Francis Group, LLC. Used by permission.


Background: There are several benefits to using short message service surveying (SMS) to gather data on substance use from homeless youth, including capturing data “in the moment” and verifying the timing of one behavior relative to another. Though SMS is a valuable data collection tool with highly mobile populations that otherwise are difficult to longitudinally sample, the reliability of SMS compared with surveys is largely unknown with homeless youth. Examining the reliability of SMS is important because these data can provide a more nuanced understanding of the relationships between various risk behaviors, which may lead to better intervention strategies with these youth.

Objectives: We compared past 30-day survey and SMS data for youth’s alcohol and marijuana use.

Methods: Interviewed 150 homeless youth (51% female) using surveys and SMS. Results: Past 30-day survey and SMS data revealed moderately strong correlations for alcohol (rs = .563) and marijuana (rs = .564). Regression analysis revealed that independent variables were similarly associated with alcohol and marijuana use when comparing survey and SMS data with two exceptions: heterosexual youth reported less alcohol use in SMS data compared to survey data (β = −.212; p < .05 vs. β = −.006; p > .05, respectively) and youth whose parents had alcohol problems reported less marijuana use in survey data compared to SMS data (β = −.277; p < .01 vs. β = −.150; p > .05, respectively).

Conclusion: Findings indicate SMS and surveys are both reliable methods of gathering data from homeless youth on substance use.