Methods
Sample
Data from the 2006 German Epidemiological Survey of Substance Abuse (ESA) were used (Kraus & Baumeister, 2008). A two-stage probability sampling design was used to select a random sample of non-institutionalized adults aged 18 to 64 years. In a first step, communities were selected proportional to population size. In a second step, individuals were randomly drawn from population registers, with younger age groups being oversampled and older age groups being undersampled. This disproportionate sampling strategy was used because the percentage of younger adults in the general population is considerably lower than the percentage of older individuals. A mixed-mode design was applied in the survey. All participants received a questionnaire by post. Those participants who did not wish to respond by post could answer the questionnaire by telephone. Mode-effect analyses generally revealed a high degree of comparability between the administration modes. Out of all eligible subjects, 7,912 individuals participated in the survey. Considering sample-irrelevant drop-outs (death, address unknown), the response rate was 45%. The ESA sample comprised 51.4% males; the prevalence of alcohol abuse and dependence was 2.6% and 3.9%, respectively. The sample did not perfectly represent the general population; in particular, people with lower educational levels and foreign nationalities were underrepresented.