NSDUH 2021 Statistical Inference Report
Learn how to produce basic estimates with the 2021 National Survey on Drug Use and Health (NSDUH). The report describes the techniques that were used to make the 2021 NSDUH Detailed Tables and the 2021 NSDUH Annual National Report, but users may also find these techniques useful for their own research with NSDUH. The report describes the calculation of estimates and sampling errors, degrees of freedom, and the procedures for determining when low-precision estimates should be suppressed. It also includes sample code in several statistical languages that data users can modify to use in their own research.
- Introduction to the report.
- Background on the survey design, including redesign and questionnaire changes.
- Prevalence estimates and how they were calculated, including specifics on various topics presented in the detailed tables.
- Discussion of how missing item responses of variables that are not imputed may lead to biased estimates.
- Discussion of sampling errors and how they were calculated.
- Description of degrees of freedom and how they were used to compare estimates.
- Discussion of how the statistical significance of differences between estimates was determined.
- Discussion of confidence interval estimation.
- Discussion of when estimates with low precision were suppressed.
Appendix A contains code samples for various statistical procedures documented within the report.
- (pdf | 1.55 MB)