National Survey on Drug Use and Health sample redesign issues and methodological studies
BACKGROUND: The Substance Abuse and Mental Health Services Administration (SAMHSA) conducted a study to determine how to modify the survey methodology to improve the quality of the estimates and the efficiency of data collection and processing for the National Survey on Drug Use and Health (NSDUH)....
| Main Author: | |
|---|---|
| Corporate Authors: | , , |
| Format: | eBook |
| Language: | English |
| Published: |
Rockville, Maryland
Substance Abuse and Mental Health Services Administration, Center for Behavioral Health Statistics and Quality
April 2012, 2012
|
| Series: | CBHSQ methodology report
|
| Online Access: | |
| Collection: | National Center for Biotechnology Information - Collection details see MPG.ReNa |
| Summary: | BACKGROUND: The Substance Abuse and Mental Health Services Administration (SAMHSA) conducted a study to determine how to modify the survey methodology to improve the quality of the estimates and the efficiency of data collection and processing for the National Survey on Drug Use and Health (NSDUH). With sample design (which includes data collection) driving much of the cost of NSDUH, it is prudent to implement a more efficient design that meets SAMHSA's data needs while maintaining the integrity of the results. METHOD: This report examines the feasibility and impacts of a variety of sample design changes on survey costs and data precision. These sample design changes included optimal cluster size, optimal sample distribution with respect to demographic and geographic groups, pros and cons of collecting data on a more continuous basis, feasibility of conducting the survey every other year, implications of expanding the target population to include children younger than age 12, and feasibility of a flexible design. RESULTS: SAMHSA found that a larger cluster size affords the ability to produce nearly as precise estimates at a greatly reduced cost; however, producing reliable state- and substate-level estimates may require a minimum number of clusters per state. The oversampling optimization investigation found that the optimal sample distribution was equal for the 12 to 17, 18 to 25, and 26 or older age groups. The continuous design assessment found pros and cons related to the training, field management activities, weighting, and analysis. CONCLUSION: In the event that SAMHSA is required to reduce or supplement the NSDUH sample to address a particular need, change from a 50-state design to national design, or make other significant changes, there are flexible design options from a sample design perspective. Depending on the objective, flexible design options for the NSDUH design could provide flexibility but at the cost of loss of trends, decreased precision, and cost implications The biennial design would be less costly; however, it would reduce response rates, affect data quality, and degrade the ability to produce state- and substate-level estimates. Expanding the target population to include children younger than age 12 would increase costs, reduce response rates, and be difficult to have cognitively comparable questions across age groups. Depending on the objective, flexible design options for the current and next NSDUH design include transitioning to a national or hybrid design (a design that preserves minimum state sample sizes at some level by supplementing a national sample), changing cluster sizes, using address-based sampling, switching to more continuous or biennial data collection, and expanding the target population. |
|---|---|
| Physical Description: | 1 PDF file (various pagings) illustrations |