National Survey on Drug Use and Health, 2005: Statistical Inference Report

摘要

Statistical inference occurs whenever data obtained from sample observations belonging to and considered representative of a larger target population are used to make generalizations concerning the larger population. The target population for the 2005 National Survey on Drug Use and Health (NSDUH)1 was the U.S. civilian, noninstitutionalized population aged 12 or older (at the time of their interview) in 2005. Measurements for this target population were the responses to the survey questions provided by persons participating in the 2005 survey. Statistical inferences concerning characteristics of interest for this population and various subpopulations are presented in the form of estimates derived from the sample data collected. Examples of the inferences made from the 2005 NSDUH data include estimates of the number of persons who were substance users during the past month, past year, and their lifetime, as well as the associated percentages (prevalence rates) of substance use for these reference periods. Inferences also were made for such categories as substance initiation; risk and protective factors; substance dependence, dependence or abuse, and treatment; and measures related to mental health problems. Among some populations of interest, sample sizes were not adequate to support inferences; in these cases, estimates were produced from annual averages based on combined data.