The 2004 National Survey on Drug Use and Health (NSDUH)1 sample design is a continuation of a coordinated 5-year sample design providing estimates for all 50 States plus the District of Columbia for the years 1999 through 2003 and continuing through 2004. The respondent universe is the civilian, noninstitutionalized population aged 12 years old or older residing within the United States and the District of Columbia. Persons excluded from the universe include active-duty military personnel, persons with no fixed household address (e.g., homeless and/or transient persons not in shelters), and residents of institutional group quarters, such as jails and hospitals.
The coordinated design for 1999 through 2003 facilitated 50 percent overlap in first-stage units (area segments) between each 2 successive years. The 2004 NSDUH continued the 50 percent overlap by retaining approximately half of the first-stage sampling units from the 2003 survey. The remainder of the sample was drawn from the 1999 through 2003 reserve sample (i.e., area segments not used in previous years). Before selection, composite size measures2 were adjusted to the 2000 census data.3 The application of a special probability sampling procedure initially developed by Keyfitz (1951) ensured that most of the overlap segments from 2003 were included in the 2004 sample.
For the 50-State design, 8 States were designated as large sample States (California, Florida, Illinois, Michigan, New York, Ohio, Pennsylvania, and Texas) with samples large enough to support direct State estimates. In 2004, sample sizes in these States ranged from 3,575 to 3,725. For the remaining 42 States and the District of Columbia, smaller, but adequate, samples were selected to support State estimates using small area estimation (SAE) techniques.4 Sample sizes in these States ranged from 828 to 934 in 2004.
States were first stratified into a total of 900 field interviewer (FI) regions (48 regions in each large sample State and 12 regions in each small sample State). These regions were contiguous geographic areas designed to yield the same number of interviews on average. Within FI regions, adjacent census blocks were combined to form the first-stage sampling units, called area segments. A total of 96 segments per FI region were selected with probability proportional to population size to support the 5-year sample and any supplemental studies that the Substance Abuse and Mental Health Services Administration (SAMHSA) may choose to field.5 Of these segments, 24 were designated for the coordinated 5-year sample, while the other 72 were designated as "reserve" segments. It is from this reserve sample and the 2003 overlap sample that the 2004 NSDUH sample segments were selected. Eight sample segments per FI region were fielded during the 2004 survey year.
These sampled segments were allocated equally into four separate samples, one for each 3-month period (calendar quarter) during the year, so that the survey was essentially continuous in the field. In each of these area segments, a listing of all addresses was made, from which a sample of 169,514 addresses was selected. Of the selected addresses, 142,612 were determined to be eligible sample units. In these sample units (which can be either households or units within group quarters), sample persons were randomly selected using an automated screening procedure programmed in a handheld computer carried by the interviewers. The number of sample units completing the screening was 130,130. Youths aged 12 to 17 years and young adults aged 18 to 25 years were oversampled at this stage. Because of the large sample size, there was no need to oversample racial/ethnic groups, as was done on surveys prior to 1999. A total of 81,973 persons were selected nationwide. Consistent with previous surveys in this series, the final respondent sample of 67,760 persons was representative of the U.S. general population (since 1991, the civilian, noninstitutionalized population) aged 12 or older. In addition, State samples were representative of their respective State populations. More detailed information on the disposition of the national screening and interview sample can be found in Appendix B. Definitions of key terms are provided in Appendix C.
The survey covers residents of households (living in houses/townhouses, apartments, condominiums, etc.), persons in noninstitutional group quarters (e.g., shelters, rooming/boarding houses, college dormitories, migratory workers' camps, halfway houses), and civilians living on military bases. Although the survey covers these types of units (they are given a nonzero probability of selection), sample sizes of most specific groups are too small to provide separate estimates. Persons excluded from the survey include homeless people who do not use shelters, active military personnel, and residents of institutional group quarters, such as correctional facilities, nursing homes, mental institutions, and long-term hospitals. More information on the sample design can be found in a 2004 NSDUH report by Bowman, Chromy, Hunter, and Martin (2005a) on the OAS website (http://www.oas.samhsa.gov/nhsda/methods.cfm#2k3).
An additional stage of sampling occurred within the 2004 computer-assisted interviewing (CAI) questionnaire. Approximately 50 percent of adult respondents aged 18 or older were randomly assigned to receive the full module of serious psychological distress (SPD) questions. The remaining adults received a reduced number of SPD questions and a new set of questions on depression. These complementary samples are together referred to as the SPD "split sample," the full SPD module is referred to as "sample A," and the reduced SPD module is referred to as "sample B."
The split sample was originally set up so that 20 percent of the adult respondents received the full module and 80 percent received the reduced module. When a preliminary analysis indicated that there may be a difference between the two samples, the selection algorithm was modified such that 60 percent received the full module and 40 percent received the reduced module in Quarters 2, 3, and 4. As a result, the sample was split half and half for the year.
The data collection method used in NSDUH involves in-person interviews with sample persons, incorporating procedures that would be likely to increase respondents' cooperation and willingness to report honestly about their illicit drug use behavior. Confidentiality is stressed in all written and oral communications with potential respondents. Respondents' names are not collected with the data, and CAI methods, including audio computer-assisted self-interviewing (ACASI), are used to provide a private and confidential setting to complete the interview.
Introductory letters are sent to sampled addresses, followed by an interviewer visit. A 5-minute screening procedure using a handheld computer involves listing all household members along with their basic demographic data. The computer uses the demographic data in a preprogrammed selection algorithm to select zero to two sample person(s), depending on the composition of the household. This selection process is designed to provide the necessary sample sizes for the specified population age groupings.
Interviewers immediately attempt to conduct the NSDUH interview with each selected person in the household. The interviewer requests the selected respondent to identify a private area in the home to conduct the interview away from other household members. The interview averages about an hour and includes a combination of CAPI (computer-assisted personal interviewing) and ACASI. The interview begins in CAPI mode with the FI reading the questions from the computer screen and entering the respondent's replies into the computer. The interview then transitions to the ACASI mode for the sensitive questions. In this mode, the respondent can read the questions silently on the computer screen and/or listen to the questions read through headphones and enter his or her responses directly into the computer. At the conclusion of the ACASI section, the interview returns to the CAPI mode with the interviewer completing the questionnaire. All respondents who complete a full interview are given a $30.00 cash payment as a token of appreciation for their time.
No personal identifying information is captured in the CAI record for the respondent. At the end of the day when an interviewer has completed one or more interviews, he or she transmits the data to RTI in Research Triangle Park, North Carolina, via home telephone lines.
Interviewers initiate nightly data transmissions of interview data and call records on days when they work. Computers at RTI direct the information to a raw data file that consists of one record for each completed interview. Even though editing and consistency checks are done by the CAI program during the interview, additional, more complex, edits and consistency checks are completed at RTI. Cases are retained only if respondents provided data on lifetime use of cigarettes and at least nine other substances. An important aspect of subsequent editing routines involves assignment of codes when respondents legitimately were skipped out of questions that definitely did not apply to them (e.g., if respondents never used a drug of interest). For key drug use measures, the editing procedures identify inconsistencies between related variables. Inconsistencies in variables pertaining to the most recent period that respondents used a drug are edited by assigning an "indefinite" period of use (e.g., use at some point in the lifetime, which could mean use in the past 30 days or past 12 months). Inconsistencies in other key drug use variables are edited by assigning missing data codes. These inconsistencies then are resolved through statistical imputation procedures, as discussed below.
For some key variables that still have missing or ambiguous values after editing, statistical imputation is used to replace these values with appropriate response codes. For example, the response is ambiguous if the editing procedures assigned a respondent's most recent use of a drug to "use at some point in the lifetime," with no definite period within the lifetime. In this case, the imputation procedures assign a definite value for when the respondent last used the drug (e.g., in the past 30 days, more than 30 days ago but within the past 12 months, more than 12 months ago). Similarly, if the response is completely missing, the imputation procedures replace missing values with nonmissing ones.
In most cases, missing or ambiguous values are imputed using a methodology called predictive mean neighborhoods (PMN), which was developed specifically for the 1999 survey and used in all subsequent survey years. PMN is a combination of a model-assisted imputation methodology and a random nearest neighbor hot-deck procedure. The hot-deck procedure is set up in such a way that imputed values are made consistent with preexisting nonmissing values for other variables. Whenever feasible, the imputation of variables using PMN is multivariate, in which imputation is accomplished on several response variables at once. Variables requiring imputation using PMN were the core demographic variables, core drug use variables (recency of use, frequency of use, and age at first use), income, health insurance, and noncore demographic variables for work status, immigrant status, and the household roster. A weighted regression imputation was used to impute some of the missing values in the nicotine dependence variables.
In the modeling stage of PMN, the model chosen depends on the nature of the response variable Y. In the 2004 NSDUH, the models included binomial logistic regression, multinomial logistic regression, Poisson regression, and ordinary linear regression, where the models incorporated the design weights.
In general, hot-deck imputation replaces a missing or ambiguous value taken from a "similar" respondent who has complete data. For random nearest neighbor hot-deck imputation, the missing or ambiguous value is replaced by a responding value from a donor randomly selected from a set of potential donors. Potential donors are those defined to be "close" to the unit with the missing or ambiguous value, according to a predefined function, called a distance metric. In the hot-deck stage of PMN, the set of candidate donors (the "neighborhood") consists of respondents with complete data who have a predicted mean close to that of the item nonrespondent. In particular, the neighborhood consists of either the set of the closest 30 respondents or the set of respondents with a predicted mean (or means) within 5 percent of the predicted mean(s) of the item nonrespondent, whichever set is smaller. If no respondents are available who have a predicted mean (or means) within 5 percent of the item nonrespondent, the respondent with the predicted mean(s) closest to that of the item nonrespondent is selected as the donor.
In the univariate case, the neighborhood of potential donors is determined by calculating the relative distance between the predicted mean for an item nonrespondent and the predicted mean for each potential donor, then choosing those means defined by the distance metric. The pool of donors is further restricted to satisfy logical constraints whenever necessary (e.g., age at first crack use must not be younger than age at first cocaine use).
Whenever possible, missing or ambiguous values for more than one response variable are considered at a time. In this (multivariate) case, the distance metric is a Mahalanobis distance (Manly, 1986) rather than a relative Euclidean distance. Whether the imputation is univariate or multivariate, only missing or ambiguous values are replaced, and donors are restricted to be logically consistent with the response variables that are not missing. Furthermore, donors are restricted to satisfy "likeness constraints" whenever possible. That is, donors are required to have the same values for variables highly correlated with the response. If no donors are available who meet these conditions, these likeness constraints can be loosened. For example, donors for the age at first use variable are required to be of the same age as recipients, if at all possible. Further details on the PMN methodology are provided in RTI International (2005b) and Singh, Grau, and Folsom (2001, 2002).
Although statistical imputation could not proceed separately within each State due to insufficient pools of donors, information about each respondent's State of residence was incorporated in the modeling and hot-deck steps. For most drugs, respondents were separated into three "State usage" categories as follows: respondents from States with high usage of a given drug were placed in one category, respondents from States with medium usage into another, and the remainder into a third category. This categorical "State rank" variable was used as one set of covariates in the imputation models. In addition, eligible donors for each item nonrespondent were restricted to be of the same State usage category (i.e., the same "State rank") as the nonrespondent.
The general approach to developing and calibrating analysis weights involved developing design-based weights, dk, as the inverse of the selection probabilities of the households and persons. Adjustment factors, ak(), then were applied to the design-based weights to adjust for nonresponse, to poststratify to known population control totals, and to control for extreme weights when necessary. In view of the importance of State-level estimates with the 50-State design, it was necessary to control for a much larger number of known population totals. Several other modifications to the general weight adjustment strategy that had been used in past surveys also were implemented for the first time beginning with the 1999 CAI sample.
Weight adjustments were based on a generalization of Deville and Särndal's (1992) logit model. This generalized exponential model (GEM) (Folsom & Singh, 2000b) incorporates unit-specific bounds (k, uk), ks, for the adjustment factor ak() as follows:
where ck are prespecified centering constants, such that k < ck < uk and Ak = (uk - k) / (uk - ck)(ck - k). The variables k, ck, and uk are user-specified bounds, and is the column vector of p model parameters corresponding to the p covariates x. The -parameters are estimated by solving
where denotes control totals that could be either nonrandom, as is generally the case with poststratification, or random, as is generally the case for nonresponse adjustment.
The final weights wk = dkak() minimize the distance function Δ(w,d) defined as
This general approach was used at several stages of the weight adjustment process, including (1) adjustment of household weights for nonresponse at the screener level, (2) poststratification of household weights to meet population controls for various demographic groups by State, (3) adjustment of household weights for extremes, (4) poststratification of selected person weights, (5) adjustment of responding person weights for nonresponse at the questionnaire level, (6) poststratification of responding person weights, and (7) adjustment of responding person weights for extremes.
Every effort was made to include as many relevant State-specific covariates (typically defined by demographic domains within States) as possible in the multivariate models used to calibrate the weights (nonresponse adjustment and poststratification steps). Because further subdivision of State samples by demographic covariates often produced small cell sample sizes, it was not possible to retain all State-specific covariates (even after meaningful collapsing of covariate categories) and still estimate the necessary model parameters with reasonable precision. Therefore, a hierarchical structure was used in grouping States with covariates defined at the national level, at the census division level within the Nation, at the State group within the census division, and, whenever possible, at the State level. In every case, the controls for total population within State and the five age groups (12 to 17, 18 to 25, 26 to 34, 35 to 49, 50 or older) within State were maintained except that, in the last step of poststratification of person weights, six age groups (12 to 17, 18 to 25, 26 to 34, 35 to 49, 50 to 64, 65 or older) were used. Census control totals by age, race, gender, and Hispanicity were required for the civilian, noninstitutionalized population of each State. Beginning with the 2002 NSDUH, the Population Estimates Branch of the U.S. Bureau of the Census produced the necessary population estimates in response to a special request based on the 2000 census.
Consistent with the surveys from 1999 onward, control of extreme weights through separate bounds for adjustment factors was incorporated into the GEM calibration processes for both nonresponse and poststratification. This is unlike the traditional method of winsorization in which extreme weights are truncated at prespecified levels and the trimmed portions of weights are distributed to the nontruncated cases. In GEM, it is possible to set bounds around the prespecified levels for extreme weights, and then the calibration process provides an objective way of deciding the extent of adjustment (or truncation) within the specified bounds. A step was added to poststratify the household-level weights to obtain census-consistent estimates based on the household rosters from all screened households; these household roster-based estimates then provided the control totals needed to calibrate the respondent pair weights for subsequent planned analyses. An additional step poststratified the selected person sample to conform to the adjusted roster estimates. This additional step takes advantage of the inherent two-phase nature of the NSDUH design. The final step poststratified the respondent person sample to external census data (defined within the State whenever possible, as discussed above). For more detailed information, see the 2003 NSDUH Methodological Resource Book (RTI International, 2005b).
For certain populations of interest, 2 years of NSDUH data were combined to obtain annual averages. The person-level weights for estimates based on the annual averages were obtained by dividing the analysis weights for the 2 specific years by a factor of two.
For the sections on SPD and adult depression in the 2004 questionnaire, the adult (aged 18 or older) sample was divided between two complementary modules: the full SPD module (sample A) and the reduced SPD module plus depression module (sample B). Therefore, two additional sets of analysis weights were required (i.e., one for sample A and one for sample B). The weights for sample A were used as the analysis weights for producing the SPD estimates, and the weights for sample B were used as the analysis weights for producing the adult depression module estimates. These two weights were created by incorporating the inverse quarterly sampling fractions associated with the random sample splits for the two modules into the weights after the person-level nonresponse adjustment. Each subsample then was poststratified separately to the census estimates of the civilian noninstitutionalized population aged 18 or older for various domains defined by age group, race/ethnicity, gender, and State. Note there are six respondents aged 18 or older who had a missing value for the SPD sample indicator variable. It appears that these six respondents broke off the interview before they could be assigned to the full or reduced SPD module. Those six respondents were excluded from either sample A or sample B; thus, they had zero weight of sample A or sample B.
This page was last updated on June 16, 2008.