|An Analysis of Low-Response Characteristics in Rural Community Drug and Alcohol Survey Data
| Rebecca Orsi
Master's Candidate, Department of Statistics, Colorado State University
Wednesday, June 21, 2007
The Tri-Ethnic Center for Prevention Research (TEC) at Colorado State University is a part of the Department of Psychology in the College of Natural Sciences . For 30 years TEC has conducted multidisciplinary research aimed at understanding community dynamics and the social, psychological and cultural factors that contribute to substance abuse. Between 1996 and 2000 TEC conducted a major, one-of-a kind survey of drug and alcohol use among junior high and high-school students in rural communities in the United States . To study drug use in non-minority communities, TEC has used data from 193 of the communities in which a very high survey participation percentage (80% or greater) was achieved. This study looks at an additional 65 non-minority communities in which the survey participation percentage was lower than 80%. Hierarchical linear models (Raudenbush and Bryk, 2002) were estimated using the specialized statistical software package HLM6 (Raudenbush, et al , 2005) to determine whether incidence of drug use, or the relationship between incidence of drug use and predictor variables, differs between high- and low-participation communities. As a result of this study, we cannot conclude that drug use behavior differs substantially between the two types of communities. This conclusion opens the door for further research into the question of whether the two groups can be combined. If TEC can expand the size of its database by including communities with lower than 80% survey participation, this could be of substantial benefit to future research in the area of how a community's characteristics influence the success of its drug and alcohol prevention programs. Understanding the similarities between high- and low-participation groups also has implications for how TEC and other organizations conducting survey research might coach low-responding communities or clients regarding appropriate use of the community's own data for substance abuse prevention programs.
Support for the Tri-Ethnic Center project which assembled the data used in this study was provided, in part, by grants R01 DA09349, Ruth W. Edwards, principal investigator, and P50 DA07074, Eugene R. Oetting, principal investigator, from the National Institute on Drug Abuse of the Department of Health, Education, and Welfare.