Go to the Table Of Contents

2000 State Estimates of Substance Use & Mental Health

bulletNational data      bulletState level data       bulletMetropolitan and other subState area data

Appendix B: State Estimation Methodology

B.1 Measuring Prevalence Levels for 2000 and the Change Between 1999 and 2000

This appendix describes the methodology used to measure change in State estimates based on the 1999 and 2000 National Household Surveys on Drug Abuse (NHSDAs), validation of the methodology, alternative measures evaluated, the mean squared error of the measures, and options to be considered for the 2001 NHSDA. State estimates were produced by combining State-level NHSDA data with local-area county and Census block group/tract-level estimates from the States. This small area estimation (SAE) methodology is described in this appendix. Section B.6 provides an overview of SAE methodology, while earlier sections discuss State-level estimates of change between 1999 and 2000.

The original plan was to compare three alternative estimators of the 2000 prevalence levels for a few substance use measures and three associated 1999 to 2000 change measures. One set of 2000 prevalence estimates was based on an independent analysis of the 2000 NHSDA data using models that were nearly identical to those used in 1999. The exception was that fixed predictors, which were no longer significant at the 1 percent level in SUrvey DAta ANalysis (SUDAAN) logistic regression runs (RTI, 2001), were dropped from the 2000 models. This is referred to as Model 1. The other two estimators of the prevalence level were derived from a simultaneous modeling of the 1999 and 2000 survey data.

The simultaneous models had separate vectors of fixed effects for eight year-by-age group domains (four age groups by 2 years) and separate State and substate region random effects for each of these eight domains. The State-level vectors of eight year-by-age group random effects were allowed to have a general (8 × 8) covariance matrix. There were four substate regions in each of the 42 small sample States and 16 in each of the eight large sample States,1 which had eight element random effect vectors for the year-by-age group domains. These substate region random effect vectors were also allowed to have general 8 × 8 covariance matrices. One estimator of the 2000 prevalence (level) produced from these simultaneous model fits (Model 2) was the associated posterior mean for each State i: Symbolic representation of the posterior mean for the 2000 prevalence in State i given the data for 1999 and 2000.. The other level estimator for 2000 (Model 3) had the following ratio form:

The equation (B-1) gives a ratio type estimator of the 2000 prevalence for State i given the 1999 and 2000 data. The estimator is defined as the product of the posterior mean of the 1999 prevalence for the ith State given the 1999 data and the State-specific ratio of posterior means for 2000 prevalence estimator and the 1999 prevalence estimator given the 1999 and 2000 data for the ith state..      (B-1)

Change was measured as the 2000 prevalence estimate for a State divided by its previously published 1999 prevalence estimate, Symbolic representation of the posterior mean for the 1999 prevalence in State i given the data for 1999.(see Office of Applied Studies [OAS], 2000). Note that the ratio estimator, Notation symbolizing the ratio type estimator for the 2000 prevalence as defined in equation (B-1)., has the property that the associated change measure is the ratio of simultaneously modeled posterior means contained in the equation (B-1) square bracket. The other two change measures used Symbolic representation of the posterior mean for the 2000 prevalence in State i given the data for 2000. and Notation depicting the posterior mean for the 2000 prevalence in State i given the data for 1999 and 2000.as the numerators and Characters representing the posterior mean for the 1999 prevalence in State i given the data for 1999.as the denominator.

Estimates were produced for four outcome variables representative of a range of prevalence rates (see the following table). The age group and overall national design-based prevalence estimates based on the 1999 and 2000 data are given in the table. For the outcome variables, the modeling began with the fixed effect predictors from their 1999 SAE models. The predictors were updated where possible for the 2000 NHSDA data, and age-group-level SUDAAN fixed logistic regression models were fit to the 2000 data. The predictors still significant at the 1 percent level are referred to as "the reduced set of predictors." For an outcome variable and for a particular age group, the same set of reduced predictors was used in all three models mentioned above. This was done to minimize the effect on the accuracy of the change measure of using different sets of fixed effect predictors for different models.

Design-Based Prevalence Estimates for the 1999 and 2000 NHSDAs, by Age Group and Outcome Variable
Outcome Variable Name 1999 Design-Based Estimates 2000 Design-Based Estimates
12–17 18–25 26+ Total 12–17 18–25 26+ Total
Past Month Use of Marijuana MRJMON 7.22 14.22 2.79 4.73 7.18 13.63 2.97 4.80
Past Year Use of Cocaine COCYR 1.61 5.24 1.11 1.69 1.66 4.40 0.97 1.49
Past Month "Binge" Alcohol Use BNGALC 10.10 37.90 18.62 20.21 10.43 37.83 19.10 20.62
Past Month Use of Cigarettes CIGMON 14.90 39.67 24.91 25.76 13.44 38.28 24.24 24.93

Source: SAMHSA, Office of Applied Studies, National Household Surveys on Drug Abuse, 1999 and 2000.

To judge the quality of the three types of SAEs, the relative absolute biases (RABs) were compared with the corresponding design-based estimates based on very large sample sizes, at about 7,200 respondents per year. The latter served as the "true" benchmark values (described in more detail in Section B.2). To simplify the comparisons, the following simple "shorthand" is used to describe the different SAEs for the prevalence rates:

Later in the text (and in Table s B.2 to B.6), the three point estimates, P00, S00, and Adj00, are simply denoted by P1, P2, and P3, respectively.

B.2 Validation

The validation study was performed by pairing the eight large sample States to form design-based benchmark estimates based on samples of around 7,200 respondents per year. For each of these State pairs, eight sample replicates were formed that mimicked the overlapping of sample segments in the 1999 and 2000 surveys from 1 of the 43 small sample States. A key feature of this replicate formation strategy was mimicking the 50 percent overlap between the samples of 96 area segments surveyed in 1999 and 2000 in a small sample State. Because new samples of dwellings and persons were drawn from all sample segments for the 2000 NHSDA, the survey design-induced covariance between years is limited to this 50 percent overlap of sample block groups/segments.

B.2.1 Introduction

To validate the fit of the 1999 NHSDA SAE models, the eight large sample States had been used as internal benchmarks. For this purpose, 12 pseudo field interviewer (FI) regions had been created within each large sample State by pooling the 48 initial regions into groups of four. Each of these pseudo FI regions were then expected to have eight area segments per calendar quarter. For each of these pseudo FI region-by-quarter sets of eight area segments, any segments devoid of interviews were first randomly replaced by a selection from the non-empty segments in the set. The completed set of eight segments from each pseudo FI region-by-quarter combination was then randomly partitioned into four replicates of two segments each. Combined across the 12 pseudo FI regions and the four calendar quarters, each of the four substate replicates mimicked the size and design structure of a small State sample.

Having created four pseudo small State samples and associated universe-level files for each large State, SAEs were then produced for 75 States (43 + 32), which included the 42 small States and the District of Columbia and the 32 substate territories defined across the eight large sample States. To validate the SAE models, the 32 pseudo small State SAEs were compared with the corresponding large State design-based estimates.

Commenting on the 1999 validation study, William Bell, a member of the panel of SAE advisors (see Section B.6.1), noted that the design-based estimates from a single large State were probably not precise enough to serve as an internal benchmarks because a large State sample size is only 4 times larger than a small State sample size. Therefore, to provide more precise internal benchmarks for the 2000 NHSDA SAE model validation, the eight large States were grouped into four pairs. The SAE model validation methodology used for the 2000 NHSDA is described in the next section.

B.2.2 Subsampling Pairs of States

To validate the prevalence (level) estimates produced by the 2000 SAE models and the associated measures of change between the 1999 and 2000 SAE estimates, the eight large sample States were grouped into four pairs: (a) California and Texas, (b) New York and Florida, (c) Ohio and Michigan, and (d) Pennsylvania and Illinois. This represented an improvement over the 1999 validation design in that the design-based estimates for individual large States had been somewhat imprecise to serve as internal benchmarks (the estimated unbiased "true" value) for the small State SAEs. The large State sample sizes were only approximately 4 times larger than the small State sample sizes. By grouping the eight large sample States into four pairs, the individual design-based estimate for each pair was based on a sample that was now 8 times as large as that for a small State. These pairs were formed based on State population totals so as to minimize variance inflation in the design-based estimates for the State pairs that results from unequal weighting. The design-based estimates corresponding to the four pairs were used as internal benchmarks.

To select the eight pseudo small State samples from each large State pair, six pseudo FI regions were first created within both of the paired large sample States by pooling their 48 initial regions into groups of eight. Each of these pseudo FI regions was then expected to have 16 area segments per calendar quarter. For each of these pseudo FI region-by-quarter sets of 16 area segments, any segments that were devoid of interviews were first randomly replaced by a selection from the non-empty segments in the set. The segments for the 1999 and 2000 NHSDA data were filled in separately. Once complete sets of 16 non-empty segments for the 1999 and 2000 NHSDA data in each of the pseudo FI region by quarter sets were assembled, the 1999 and 2000 data were linked using State-by-pseudo FI region-by-quarter-by segment identification codes. Approximately half of the 16 segments represented cases where the 1999 segments were reused in 2000 (i.e., common segments in 1999 and 2000), and the remaining ones represented cases where a 1999 segment was linked with a new 2000 replacement segment. Next, the 16 linked 1999 and 2000 segment pairs were stratified into two strata—the common segment pairs and the uncommon 1999 and 2000 segment pairs. One segment pair was then drawn from each of these strata and combined to form eight replicate pairs such that one of the paired replicates would have common segments in the 1999 and 2000 surveys and the other replicate pair would have uncommon segments for 1999 and 2000.

This subsampling validation exercise was repeated for all four quarters in a pseudo FI region, for all six pseudo FI regions in both States of the pair, and for all four pairs of large States. This resulted in 32 (4 large State pairs × 8 subsamples from each large State pair) replicate subsamples from four large State pairs. These subsamples mimicked the design properties of small States. For example, let CA_TX1 represent the first subsample from the California and Texas pair. Then subsample State CA_TX1 will have 12 pseudo FI regions (six from California and six from Texas), each pseudo FI region will have four quarters, and each quarter will have two segments per survey year. Of these two 1999 segments, one segment would be retained in the 2000 survey and the other 1999 segment would be replaced. This was desirable because one of the goals was to estimate the change in drug use prevalence rates between 1999 and 2000. The planned 50 percent overlap in sample segments between the 1999 and 2000 surveys would be expected to yield positive between-year correlations and reduced posterior variances for the change measures. Survey weights for the eight subsamples formed from each large State pair were appropriately scaled so that their totals over the subsamples would reproduce the age-group-level large State pair totals.

Having created eight pseudo small State samples and associated universe-level files for each large State pair, SAEs for the models described in the following section were then produced for 75 States, which include the 42 small States and the District of Columbia and the 32 substate territories defined for each large sample State.

B.2.3 Results of SAE Models of Prevalence Levels

Table B.1, provided at the end of this appendix, compares Model 1 estimates of the prevalence levels for 2000 to the "true" design-based estimates for each of the four substances estimated in the validation. A "Model 1 estimate" is the mean over the eight substates of the prevalence estimates. As shorthand, Model 1 estimates of the prevalence level for 2000 have been denoted as P1. The P1 estimates are generally quite close to the design-based estimates for each of the large State sample pairs, for each of the age categories, and across all four substances measured.

Table s B.2 to B.5 present the means over the eight substates for the three types of SAEs mentioned above. The paired large State design-based estimates are also included for comparison purposes. The RABs are shown in these tables, including the average RABs over the four pairs of large States. Overall averages of these RABs across the four outcome variables are presented in Table B.6.

Focusing on just the 12 or older age group, the overall RABs are quite small, ranging from a low of about 1 percent (for past month cigarette use) to 6 percent (for past year cocaine and past month marijuana use). This is similar to the results for 1999.

The three point estimates (P1, P2, and P3) surprisingly gave similar results, with none consistently superior over the four substances. The overall RAB averages for the 26 or older age group and the total appear to be lowest for P1.

Table B.7 shows averages over eight substates of the 95 percent lower and upper prediction interval (PI) bounds of P1 for the four outcome variables. The corresponding paired large State design-based 95 percent confidence interval (CI) bounds are also included for comparison purposes to the model-based 95 percent PIs. It should be noted that the design-based intervals of the paired large States are typically smaller owing to the larger sample sizes of approximately 6,000 to 8,000 cases for the 12 or older age group. Table B.8 shows ratios of average widths of 95 percent PIs for the model-based estimates divided by the corresponding averaged design-based interval widths for the eight substate replicates (denoted W1, W2, and W3 to correspond to estimators P1, P2, and P3, respectively). Here, the model-based estimates mimic the design and hierarchical Bayes estimation of the 42 small States and the District of Columbia, with each substate replicate having a sample of approximately 900 total for the 12 or older age group. The overall average shows that the model-based PIs were 25 to 35 percent shorter than the design-based CIs. The largest reduction in width occurred for the P1 PIs.

B.2.4 Results of Small Area Estimates of Change Between 1999 and 2000

Three types of change estimators were considered. These estimators were defined in terms of ratios of three types of 2000 prevalence rates defined above with respect to the 1999 individual estimates:

The posterior variances of Adj00, R1, R2, and R3 and the posterior correlation between the S99 and S00 were all calculated on the logarithmic scale. Because there is no direct method of estimating posterior correlations between P99 and P00, S00, and Adj00, the correlations from the simultaneous model were used to approximate the correlations for the non-simultaneous solutions. Table s B.9 to B.12 (past month marijuana, past year cocaine, past month "binge" alcohol use, and past month cigarette use, respectively) present the model-based 32 substate change estimates and RABs relative to their corresponding large State pair design-based estimates.

Table B.13 shows the estimates of change for R1 compared with the corresponding design-based estimates of change. The design-based estimates of change were similar to those based on SAE. The paired large States' design-based estimates for change (D00/D99) were also included for comparison purposes where D99 and D00 are paired large State design-based estimates of prevalence rates for 1999 and 2000.

Table B.14 quantifies the estimated bias for all three measures of change by presenting averages over the eight substates for R1, R2, and R3. The RABs are shown in Table B.14, which also contains average RABs over the four pairs of large States and overall averages over the four outcome variables. The overall RAB averages for the 18 to 25 age group, the 26 or older age groups, and the total were lowest for R1 whereas for the 12 to 17 age group, the overall RAB was smallest for R3. In general, the estimated bias was quite small.

Table B.15 shows averages over eight substates of the 95 percent lower and upper PI bounds for the change measure R1. The corresponding paired large States' design-based 95 percent CI bounds for change are also included for comparison purposes. What is important to note from this table is that the model-based 95 percent PIs generally ranged from less than one to more than one, indicating that the interval sizes are such that the changes between 1999 and 2000 are not significant (i.e., not significantly different than "no change").

Table B.16 shows ratios of average widths of 95 percent PIs for model-based change estimates and corresponding design-based change estimates. The overall average shows that the model-based PIs were 30 to 40 percent smaller than the design-based CIs. Table B.16 basically confirms the fact that the model-based estimates of change for the 42 States and the District of Columbia were more precise than the corresponding design-based estimates based on samples of about 1,000 for persons aged 12 or older, although the estimates were still not precise enough to be very useful (Table B.15).

To summarize, the estimated RABs of all three level and change estimates were reasonably small. The past year cocaine biases for the 12 to 17 age group were larger (in the 20 percent range [see Table B.6]), but this was due to the small divisors (low cocaine prevalence rates) in the relative bias calculation. There were somewhat surprising results in Table s B.6, B.8, B.14, and B.16, including those showing the interval width ratios relative to the design-based intervals for the three types of estimators. These tables show very little if any difference between the three estimator types. It had been hypothesized that the change ratios based on the simultaneous solutions would be smaller due to simultaneous estimation of the 1999 and 2000 State-level random effects. This simultaneous estimation effect turned out to be minimal. For the relative widths of the intervals, the correlations between Symbols for the natural logarithm of the posterior prevalence estimator for 1999 given the data for 1999 and 2000. and Notation depicting the natural logarithm of the posterior prevalence estimator for 2000 given the data for 1999 and 2000. obtained from the simultaneous solution were used to approximate the unknown correlations between Symbols for the natural logarithm of the posterior prevalence estimator for 1999 given the data for 1999. and Symbols for the natural logarithm of the posterior prevalence estimator for 2000 given the data for 1999 and 2000.. It was also assumed that the independent and simultaneous 1999 prevalence estimates were perfectly correlated on the log scale. With these assumptions about the correlations, the only way the simultaneous solutions could yield narrower PIs for the 2000 level and 1999 versus the 2000 change ratios was if the posterior variances of the log prevalence were reduced by conditioning simultaneously on both years of data. For the four outcome variables examined, conditioning simultaneously on both years of data did not noticeably reduce the posterior variances of the log prevalence.

Referring again to Table B.15, which provides the average 95 percent PIs for each of the substances for each paired State, it is clear that very few of the changes were significant (i.e., had intervals that did not overlap 1.0), and those few that were significant carry with them such large PIs as to render them not very useful for analytic purposes. Table B.17 presents the 95 percent least significant lower and upper bounds of change estimates (R1) for 2000 and 1999. It shows how low or high the change estimate has to be for it to be a significant change.

B.3 Conclusions about Measuring Change

The information provided in Sections B.1 and B.2 was also provided to the expert panel on SAE for its review. The panel's consensus was that the NHSDA should not be used to measure change between 1999 and 2000. Instead, the panel indicated that the Substance Abuse and Mental Health Services Administration (SAMHSA) would be better served by providing improved estimates of the prevalence levels based on combining the 1999 and 2000 data while exploring other possibilities for measuring change for 2001.

Several options for measuring change were discussed:

  1. Calculate a 2-year moving average to increase the precision of estimated change. Change then would be measured as the difference between two consecutive 2-year moving averages (e.g., the combined 2000-2001 moving average minus the combined 1999-2000 moving average).
  2. Jointly model two related variables at one time to increase the precision of the individual yearly estimates, thereby increasing the precision of year-to-year change.
  3. Create a "100 percent overlap" by asking the respondent to provide retrospective information for the "year before last" as well as for the past year.
  4. Using the current 50 percent overlap of segments, reinterview respondents in those segments in 2 consecutive years.
  5. Encourage the Center for Substance Abuse Prevention (CSAP) and the Center for Substance Abuse Treatment (CSAT) to have States institute annual data collection of variables that are related to program participation and success and are based on consistent definitions across States.
For each option, the SAE expert panel discussed the advantages and disadvantages. The advantages were typically either a shorter time frame for implementation or a greater potential for significant improvement in precision. Results of the discussion are summarized below.

Option 1 is the only short-term solution that could be used in 2001. The variance of a difference of two 2-year moving averages is one quarter of the variance of a difference of 2 consecutive years. This is a significant improvement; however, it still may not be possible to detect change for some measures in many States. There is a better opportunity among those States for which the trend between 1999 and 2001 continued in the same direction (either positive or negative).

Option 2 could be implemented in about 1 year and used to complement the other options in order to give added improvement to the precision of the change estimates. This option may require significant reprogramming implementation costs. The primary impact of this multivariate approach is the improved estimation of prevalence levels; however, this will also carry over to better estimation of yearly change and better correlation between years. The NHSDA data can be used to estimate the potential improvement resulting from its implementation. Some of the possible estimates that could be used together include modeling past month marijuana use and past month use of any illicit drug simultaneously, modeling past month use of a substance with past year use of the same substance, and modeling past month use of a substance with a scale based on risk factors.

Option 3 is believed to have potential; however, there can be problems because of memory recall decay. It may be best considered in conjunction with Option 2 in that retrospective data can be modeled jointly with the main outcome to improve precision. Various methods of adjusting for potential memory recall problems were considered. Data have been collected from earlier NHSDAs that may allow the estimation of potential gain under different alternatives relating to the size of the year-to-year correlation, depending on whether the recency measure is past month, past year, or some other reference period. This method would not be applicable to measures of perceived risk (asking a retrospective question about perceptions over a year ago) or to measures based on a series of questions, such as past year dependence.

A significant difficulty with Option 4 is that it represents a major departure from current data collection protocol with respect to confidentiality and follow-up. This would probably mean that other options would be pursued before this was given greater consideration. The NHSDA has typically been presented as an annual survey that does not reinterview individuals and only maintains the household address and selected respondent identification for quality verification follow-up. There is also the issue of nonresponse, especially in the reinterview. The $30 incentive would probably ameliorate the nonresponse problem to some extent.

Option 5 is more of a long-range goal. Its viability depends on adherence to consistent definitions and its ultimate predictive ability, which cannot be known in advance. Both CSAP and CSAT obtain State-level data, but the collection is typically not annual and often does not cover all 50 States and the District of Columbia.

The SAE experts suggested first conducting research based on available data and utilizing "off-the-shelf" software to better understand the potential gains that might be achievable.

B.4 Modeling Estimates Based on Combining Data from 1999 and 2000

B.4.1 Modeling Without the Region Main Effect

In addition to combining 2 years of data for the next set of State estimates, another concern was the potential impact on individual States within each region of including Census region in the model, as had been done for the 1999 State small area estimates. The concern was that the presence of a significant regional fixed effect would have an automatic result of raising or lowering the measure of substance use for each State in the region regardless of whether a given State followed the regional pattern or not. In addition to providing the best fit of the data for the States and the District of Columbia collectively, one of the primary goals was to distinguish real differences among States. Thus, including a fixed regional effect seemed counterintuitive. Although the regional effect could come collectively from a set of States in a region, it was also a concern that a single State having a relatively large population relative to the region (e.g., California in the West) could have a significant effect on the remaining States in the region, especially if its substance use level were significantly different.

To compare the impact of including versus excluding the region variable main effects, models were fit on a handful of substance use measures for the two alternatives using the SAE methodology. In addition to the region main effects, it was decided that any interactions of region with race/ethnicity or gender needed to be excluded as well because their effect on States within a region would be similar. The interactions of all other variables with region (e.g., the percentage of single mothers in the Census tract) were left in the model because differential State distributions of that data would still result in differential impacts among States within the same region.

Results are presented for four of the substances. The interest here was in comparing the average behavior across States of the two different models. Table B.18 displays the design-based estimates and SAE estimates for past month use of marijuana with region and without region. Table B.19 displays similar information for past year use of cocaine. Table s B.20 and B.21 display similar information for past month use of alcohol and past month use of cigarettes, respectively.

The estimates for past month use of marijuana for persons aged 12 or older showed that the average State differences between the SAE estimates and the design-based estimates were very similar—0.07 percent for the model without region and 0.11 percent for the model with region (Table B.18). The width of the 95 percent PIs also was similar—1.98 and 1.95 percent for the models without and with region, respectively. Also, the with-region estimate for Utah, a State that typically displays significantly lower prevalence rates than the remainder of the West region, was 0.58 percentage points higher than the design-based estimate, while the without-region estimate was only 0.14 percentage points higher.

A comparison of the estimates for alcohol use shows that the national SAE estimates were almost identical for the two models, and the difference from the national design-based estimate was only a quarter of a percent. The average difference across States between the model without region and the design-based estimate for those aged 12 or older was -0.44 percent, and the average absolute difference was 1.00 percent (Table B.20). For the model with region, the corresponding averages were -0.39 and 0.97 percent. The average PI widths were 6.26 percent for the model without region and 6.02 percent for the model with region. Thus, for example, a State such a Utah that displayed a much lower level of past month use of alcohol than other States in the West had a with-region SAE that was 3.20 percentage points higher than the corresponding design-based estimate and a without-region SAE that was only 1.92 percentage points higher.

In addition to these comparisons, the SAE estimates were validated for another substance, "the treatment gap rate" (not shown in tables) using the four large-State design-based estimates as the "true" value. The results showed that the SAE estimates for the without-region model were very close to the corresponding "true" values.

B.4.2 Validation of the SAE for the Combined 1999 and 2000 Data

Given that the decision had been made to produce SAE estimates based on combining 2 years of data, the next step was to produce those estimates for each of the substance measures for which there were 2 years of data and validate the estimates using the model that excluded any regional main effects and interactions, as described in Section B.4.1.

The 2-year estimates were validated on four variables: past month use of marijuana, past year use of cocaine, past month "binge" alcohol use, and past month use of cigarettes. For this validation exercise, large States were not combined. Each large State was divided into four substates. The results are presented in Table s B.22 to B.25. On average, the RABs were quite small. For the 12 or older age group, the RABs were as follows:

Also, compared with the design-based CIs, the 95 percent PIs were much shorter, about 75 percent as large for marijuana, "binge" alcohol, and cigarettes and 65 percent as large for cocaine (Table B.26).

In addition, the 2-year estimates were compared with the corresponding 1-year estimates to ascertain the extent of improvement in estimation for the 42 States and the District of Columbia, given that those sample sizes would now be approximately double their size in 1999. For example, comparing the PIs' widths across the 50 States and the District of Columbia, the average SAE PI width for past month use of marijuana among persons 12 or older was 2.40 percent in 1999, but only 1.98 percent for 1999 and 2000 combined. Just as importantly, because the States (and the District of Columbia) had smaller single-year sample sizes, the national model had a greater relative influence in the SAE estimates for 1999 than for 1999 and 2000 combined. Therefore, the 1999-2000 pooled State estimates would not be shrunk as much toward the national model-based estimate as would similar estimates based on a single year of data. One result is that the 2-year SAEs would tend to be closer to their corresponding design-based estimates than SAEs based on 1 year of data. The other implication is that States with design-based estimates that were relatively lower or higher than other States would retain that distinction, and the overall range and spread of the State estimates would tend to be larger, for example, than it was in 1999. This should make it easier to identify States that have significantly lower or higher substance use prevalence rates than other States.

B.5 Caveats

Most of the caveats regarding SAE, including differential nonresponse and response bias effects, are addressed in Chapter 6 of this report. Table s B.27 and B.28 show the screening and interview response rates for the 50 States and the District of Columbia in 1999 and 2000, respectively. One other possible contributor to bias in the State estimates, and the estimates in general, is the effect of editing and imputation of the summary data. In developing the editing and imputation process for 1999 and subsequent years, the desire was to minimize the amount of editing because of its somewhat subjective nature, and instead let the random imputation process supply any partially missing information. Overall, the percentage of imputed information is quite small for any given substance.

The method described in Appendix D is based on a multivariate imputation in which some demographic and other substance use information from the respondent is used to determine a donor who is similar in those characteristics but has supplied data for the drug in question. Often, information is also available from the partial respondent on the recency of drug use. For example, respondents may have indicated that they used the drug in their lifetime or in the past year, but left blank the question about use in the past month. For many of the records, this type of auxiliary information was available. In a small portion of the time, no auxiliary information was available, in which case a random donor with similar drug use patterns and demographic characteristics was used. For the different substances, the largest differences between the edited and the imputed estimates typically occurred when there was a lot of auxiliary information. For past month use of marijuana, based on the 1999 data, the State with the largest percentage change from edited to imputed data was Alabama, whose edited rate of use of marijuana was 2.1 percent and whose imputed rate of use was 3.1 percent—a relative increase of almost 50 percent.

B.6 SAE Methodology

B.6.1 Background

In response to the need for State-level information on substance abuse problems, SAMHSA began developing and testing SAE methods for the NHSDA in 1994 under a contract with RTI of Research Triangle Park, North Carolina. That developmental work used logistic regression models with data from the combined 1991 to 1993 NHSDAs and local area indicators, such as drug-related arrests, alcohol-related death rates, and block group/tract level characteristics from the 1990 Census that were found to be associated with substance abuse. In 1996, the results were published for 25 States for which there were sufficient sample data (OAS, 1996). A subsequent report described the methodology in detail and noted areas in which improvements were needed (Folsom & Judkins, 1997).

The increasing need for State-level estimates of substance use led to the decision to expand the NHSDA to provide estimates for all 50 States and the District of Columbia on an annual basis beginning in 1999. It was determined that, with the use of modeling similar to that used with the 1991 to 1993 NHSDA data in conjunction with a sample designed for State-level estimation, a sample of about 67,500 persons would be sufficient to make reasonably precise estimates.

The State-based NHSDA sample design implemented in 1999 and 2000 had the following characteristics:

In preparation for the modeling of the 1999 data, RTI used the data from the combined 1994-1996 NHSDAs to develop an improved methodology that utilized more local area data and produced better estimates of the accuracy of the State estimates (Folsom, Shah, & Vaish, 1999). That effort involved the development of procedures that would validate the results for geographic areas with large samples. This work was reviewed by a panel with SAE expertise.2 They approved of the methodology, but suggested further improvements for the modeling to be used to produce the 1999 State estimates. Those improvements were incorporated into the methodology finally used for the 1999 State estimates. Similar methodology (as described earlier) was used for this 2000 State report. The methodology, called Survey-Weighted Hierarchical Bayes Estimation (HB), is described below.

B.6.2 Goals of Modeling

There were several goals underlying the estimation process. The first was to model drug use at the lowest possible level and aggregate over the levels to form the State estimates. The chosen level of aggregation was the 32 age group (12 to 17, 18 to 25, 26 to 34, 35+) by race/ethnicity (white, non-Hispanic; black, non-Hispanic; Hispanic; Other) by gender cells at the block group level. Estimated population counts were obtained from a private vendor for each block group for each of the 32 cells. This level of aggregation was desired because the NHSDA first stage of sample selection was at the block group level, so that there would be data at this level to fit a model. In addition, there was a great deal of information from the Census at the block group level that could be used as predictors in the models. If prevalence rates could be estimated for each of the 32 cells at the block group level, it would only be necessary to multiply by the estimated population counts and aggregate to the State level.

Another goal of the estimation process was to include the sampling weight in the model in such a way that the small area estimates would converge to the design-based (sample-weighted) estimates when they were aggregated to a sufficient sample size. There was a desire for the estimates to have this characteristic so that there would be consistency with the survey-weighted national estimates based on the entire sample.

A third goal was to include as much local source data as possible, especially data related to each substance use measure. This would help provide a better fit beyond the strictly sociodemographic information. The desire was to use national sources of these data so that there would be consistency of collection and estimation methodology across States.

Recognizing that estimates based solely on these "fixed" effects would not reflect differences across States due to differences in laws, enforcement activities, advertising campaigns, outreach activities, and other such unique State contributions, a fourth goal was to include "random" effects to compensate for these differences. The types of random effects that could be supported by the NHSDA data were a function of the size of sample and the model fit to the sample data. Random effects were included at the State level and for substate regions comprised of three neighboring FI regions. Although this grouping of the three FI regions was principally motivated by the need to accumulate enough sample to support good model fitting for the low prevalence NHSDA outcomes, it was also reasoned that it would be possible to produce substate HB estimates for areas comprised of these FI region groups, once 2 or 3 years of NHSDA data were available, because that would yield substate region samples of at least 400 respondents. For substate areas that do not conform to the substate region boundaries (e.g., counties and large municipalities), HB estimates could be derived from their elemental block group level contributions, but the design-based data employed in the estimation of the associated substate region effects would not be restricted to the county or city of interest. This mismatch of FI region and county/large municipality boundaries weakens the theoretical appeal of the associated HB estimate. For this reason, substate HB estimates probably should be restricted to areas that can be matched reasonably well to FI region groups.

One of the difficulties of typical SAE has been obtaining good estimates of the accuracy of the estimates with PIs that give a good representation of the true probability of coverage of the intervals. Therefore, the final major goal was to provide accurate PIs—ones that would approach the usual sample-based intervals as the sample size increases.

B.6.3 Variables Modeled

A set of 18 measures covering a variety of aspects of substance use and abuse was designated for estimation from the 2000 NHSDA. The first 12 were based on the pooled 1999 and 2000 NHSDA data and the remaining 6 on 2000 data alone. Recall that the SAE measures of annual change evaluated in Section B.2 were not precise enough to declare significant the size of annual changes that were observed. After conferring with the SAE expert panel, the decision was made to produce averaged 1999 and 2000 State estimates for the 12 measures that were comparably defined. The 18 outcome variables are listed below:

  1. past month use of any illicit drug,
  2. past month use of marijuana,
  3. perceptions of great risk of smoking marijuana once a month,
  4. average annual rates of first use of marijuana,
  5. past month use of any illicit drug other than marijuana,
  6. past year use of cocaine,
  7. past month use of alcohol,
  8. past month "binge" alcohol use,
  9. perceptions of great risk of having five or more drinks of an alcoholic beverage once or twice a week,
  1. past month use of any tobacco product,
  2. past month use of cigarettes,
  3. perceptions of great risk of smoking one or more packs of cigarettes per day,
  4. past year alcohol dependence or abuse,
  5. past year alcohol dependence,
  6. past year any illicit drug dependence or abuse,
  7. past year any illicit drug dependence,
  8. past year dependence or abuse for any illicit drug or alcohol, and
  9. past year treatment gap.

Note that the "past year treatment gap" variable is not covered in this report (see, instead, OAS, 2002).

B.6.4 Predictors Used in Logistic Regression Models

Local area data used as potential predictor variables in the logistic regression models were obtained from several sources, including Claritas, the Census Bureau, the FBI (Uniform Crime Reports), Health Resources and Services Administration (Area Resource File), SAMHSA (Uniform Facility Data Set), and the National Center for Health Statistics (mortality data). The list of sources and potential data items used in the modeling are provided below.

The following lists provide the specific independent variables that were potential predictors in the models.
Claritas Data
Description Level
% Population aged 0–18 in block group Block group
% Population aged 19–24 in block group Block group
% Population aged 25–34 in block group Block group
% Population aged 35–44 in block group Block group
% Population aged 45–54 in block group Block group
% Population aged 55–64 in block group Block group
% Population aged 65+ in block group Block group
% Blacks in block group Block group
% Hispanics in block group Block group
% Other race in block group Block group
% Whites in block group Block group
% Males in block group Block group
% Females in block group Block group
% American Indian, Eskimo, Aleut in tract Tract
% Asian, Pacific Islander in tract Tract
% Population aged 0–18 in tract Tract
% Population aged 19–24 in tract Tract
% Population aged 25–34 in tract Tract
% Population aged 35–44 in tract Tract
% Population aged 45–54 in tract Tract
% Population aged 55–64 in tract Tract
% Population aged 65+ in tract Tract
% Blacks in tract Tract
% Hispanics in tract Tract
% Other race in tract Tract
% Whites in tract Tract
% Males in tract Tract
% Females in tract Tract
% Population aged 0–18 in county County
% Population aged 19–24 in county County
% Population aged 25–34 in county County
% Population aged 35–44 in county County
% Population aged 45–54 in county County
% Population aged 55–64 in county County
% Population aged 65+ in county County
% Blacks in county County
% Hispanics in county County
% Other race in county County
% Whites in county County
% Males in county County
% Females in county County

 

1990 Census Data
Description Level
% Population who dropped out of high school Tract
% Housing units built in 1940–1949 Tract
% Persons 16–64 with a work disability Tract
% Hispanics who are Cuban Tract
% Females 16 years or older in labor force Tract
% Females never married Tract
% Females separated/divorced/widowed/other Tract
% One–person households Tract
% Female head of household, no spouse, child 18 Tract
% Males 16 years or older in labor force Tract
% Males never married Tract
% Males separated/divorced/widowed/other Tract
% Housing units built in 1939 or earlier Tract
Average persons per room Tract
% Families below poverty level Tract
% Households with public assistance income Tract
% Housing units rented Tract
% Population 9–12 years of school, no high school diploma Tract
% Population 0–8 years of school Tract
% Population with associate's degree Tract
% Population some college and no degree Tract
% Population with bachelor's, graduate, professional degree Tract
Median rents for rental units Tract
Median value of owner–occupied housing units Tract
Median household income Tract

 

Uniform Crime Report Data
Description Level
Drug possession arrest rate County
Drug sale/manufacture arrest rate County
Drug violations' arrest rate County
Marijuana possession arrest rate County
Marijuana sale/manufacture arrest rate County
Opium cocaine possession arrest rate County
Opium cocaine sale/manufacture arrest rate County
Other drug possession arrest rate County
Other dangerous non-narcotics arrest rate County
Serious crime arrest rate County
Violent crime arrest rate County

 

Categorical Data
Description Source Level
=1 if Hispanic, =0 otherwise Sample Person
=1 if non-Hispanic Black, =0 otherwise Sample Person
=1 if non-Hispanic Other, =0 otherwise Sample Person
=1 if male, =0 if female Sample Person
=1 if Northeast region, =0 otherwise 1990 Census State
=1 if Midwest region, =0 otherwise 1990 Census State
=1 if South region, =0 otherwise 1990 Census State
=1 if MSA with 1 million +, =0 otherwise 1990 Census County
=1 if MSA with <1 million, =0 otherwise 1990 Census County
=1 if non-MSA urban, =0 otherwise 1990 Census Tract
=1 if underclass tract Urban Institute Tract
=1 if no Cubans in tract, =0 otherwise 1990 Census Tract
=1 if urban area, =0 if rural area 1990 Census Tract
=1 if no arrests for dangerous non-narcotics, =0 otherwise UCR County

 

Miscellaneous Data
Variable Description Source Level
Alcohol death rate, direct cause ICD-9 County
Alcohol death rate, indirect cause ICD-9 County
Cigarettes death rate, direct cause ICD-9 County
Cigarettes death rate, indirect cause ICD-9 County
Drug death rate, direct cause ICD-9 County
Drug death rate, indirect cause ICD-9 County
Alcohol treatment rate UFDS County
Alcohol and drug treatment rate UFDS County
Drug treatment rate UFDS County
% Families below poverty level ARF County
Unemployment rate ARF County
Per capita income (in thousands) ARF County
Food stamp participation rate Census Bureau County

B.6.5 Selection of Independent Variables for the Models

For outcomes modeled using 2000 data alone, independent variables for modeling each of the substance use measures were first identified by a CHAID (Chi-squared Automatic Interaction Detector) algorithm, which does not use sample weights. Prior to this process, all the continuous variables were categorized using deciles and were treated as ordinal in CHAID. Region was treated as nominal categorical variables in CHAID. Significant (at 3 percent level) independent variables from each model and final nodes in the tree-growing process were identified as predictor variables destined for inclusion at a later step.

Independently, a SAS stepwise logistic regression model was fit for each dependent variable by age group. The SAS stepwise was used because it was able to quickly run all of the variables for all of the models, although it was recognized that the software would not take into account the complex sample design. The independent variables included all the first-order or linear polynomial trend contrasts across the 10 levels of the categorized variables plus the gender, region, and race variables. Significant variables (at the 3 percent level) were identified from this process. Based on the combined list from CHAID and SAS, a list of variables was created that included the corresponding second- and third-order polynomials and the interaction of the first-order polynomials with the gender, race, and region variables.

Next, the variables were entered into a SAS stepwise logistic model at the 1 percent significance level. Because of past concerns about overfitting of the data in earlier estimation using the 1991 to 1993 NHSDA data, the significance levels were made quite stringent. These variables were then entered into a SUDAAN logistic regression model because the SUDAAN software would adjust for the effects of the weights and other aspects of the complex sample design. All variables that were still significant at the 1 percent significance level were entered into the survey-weighted HB process.

For outcome variables modeled using pooled 1999 and 2000 data, the starting predictor set was the final predictor set used in the 1999 analyses. This set was further reduced by modeling the combined data using SUDAAN selection at the 1 percent level of significance.

B.6.6 General Model Description

The model can be characterized as a complex mixed model (including both fixed and random effects) of the form:

This notation depicts a complex mixed logistic model. Lambda equals X times beta plus Z times U. X times beta is the usual (fixed) regression contribution, and Z times U represents random effects for the States and FI region groups. Lambda is a vector of the log odds of the propensity for a particular person in a particular FI composite region in a given State to engage in the behavior of interest..

Each of the symbols represents a matrix or vector. The leading term This notation depicts X times beta which is the usual (fixed) regression contribution. is the usual (fixed) regression contribution, and This notation depicts Z times U which represents random effects for the States and FI region groups. represents random effects for the States and FI region groups that the data will support and for which estimates are desired. Not obvious from the notation is that the form of the model is a logistic model used to estimate dichotomous data. The lambda vector has elements This notation depicts Lambda which is a vector of the log odds of the propensity for a particular person-k in a particular FI composite region-j in a given State i to engage in the behavior of interest., where the This notation depicts Pie sub i, j, k which is the propensity for the kth person in the jth FI composite region in the ith State to engage in the behavior of interest. is the propensity for the kth person in the jth FI composite region in the ith State to engage in the behavior of interest (e.g., to use marijuana in the past month). Also not obvious from the notation is that the model fitting utilizes the final "sample" weights as discussed above. The "sample" weights have been adjusted for nonresponse and poststratified to known Census counts.

The estimate for each State behaves like a "weighted" average of the design-based estimate in that State and the predicted value based on the national regression model. The "weights" in this case are functions of the relative precision of the sample-based estimate for the State and the predicted estimate based on the national model. The eight large States have large samples, and thus more "weight" is given to the sample estimate relative to the model-based regression estimate. The 42 small States and the District of Columbia put relatively more "weight" on the regression estimate because of their smaller samples. The national regression estimate actually uses national parameters that are based on the pooled 1999 and 2000 sample of approximately 138,000 persons; however, the regression estimate for a specific State is based on applying the national regression parameters to that State's "local" county, block group, and tract-level predictor variables and summing to the State level. Therefore, even the national regression component of the estimate for a State includes "local" State data. For the five outcome variables presented in this report, whose State estimates were based on the 2000 data alone, the national or fixed regression coefficients were based on a sample of roughly 72,000 persons.

The goal then was to come up with the best estimates of beta and U. This would lead to the best estimates of lambda, which would in turn lead to the best estimate of pi. Once the best estimate of pi for each block group and each age/race/gender cell within a block group has been estimated, the results could be weighted by the projected Census population counts at that level to make estimates for any geographic area larger than a block group.

In the model fitting for the pooled 1999 and 2000 data, the small number of predictor variables updated in 2000 were used in both their 1999 and 2000 versions when they appeared in a model. At the time the decision was made to form pooled data estimates, the schedule did not permit merging the updated versions of the predictors onto the 1999 analysis file. To produce the pooled data estimates, the updated versions of the predictors were always used (i.e., the versions on the 2000 universe file). For the block group-level population counts by age, race, and gender used to aggregate the block group-level prevalence estimates, the average of the 1999 and 2000 population counts were used in each demographic domain.

B.6.7 Implementation of Modeling

The solution to the equation for pi in Section B.6.6 is not straightforward but involves a series of iterative steps to generate values of the desired fixed and random effects from the underlying joint distribution. The basic process can be described as follows.

Let beta denote the matrix of fixed effects, eta be the matrix of State random effects i = 1–51, and nu denote the matrix of FI composite region effects j within State i. Because the goal is to estimate separate models for four age groups, it is assumed that the random effect vectors are four variate Normal with null mean vectors and 4×4 covariance matrices This notation depicts D sub eta, which is the 44 variance-covariance matrix of the State random effects.and This notation depicts D sub nu, which is the 44 variance-covariance matrix of the FI composite region level random effects., respectively. To estimate the individual effects, a Bayesian approach is used to represent the joint density function given the data by Notation depicting joint probability density function of fixed effects (beta), State random effects (eta), composite field interviewer region effects (nu) within the State, and associated 4 by 4 variance-covariance matrices (D sub nu) and (D sub eta) assuming that the data (y) are known.. According to the Bayes process, this can be estimated once the conditional distributions are known:

Notation depicting conditional probability distribution of fixed effects (beta) assuming that the data (y) and the following parameters are known: State random effects (eta), composite field interviewer region effects (nu) within the State, and associated 4 by 4 variance-covariance matrices (D sub nu) and (D sub eta). Notation depicting conditional probability distribution of the 4 by 4 variance-covariance matrices (D sub nu) and (D sub eta) assuming that the data (y) and the following parameters are known: fixed effects (beta), State random effects (eta), composite field interviewer region effects (nu) within the State. and Notation depicting conditional probability distribution of State random effects (eta) and composite field interviewer region effects (nu) within the State, assuming that the data (y) and the following parameters are known: fixed effects (beta)  and associated 4 by 4 variance-covariance matrices (D sub nu) and (D sub eta)..

To generate random draws from these distributions, Markov Chain Monte Carlo (MCMC) processes need to be used. There are a body of methods for generating pseudo-random draws from probability distributions via Markov chains. A Markov chain is fully specified by its starting distribution Notation depicting probability of X sub zero, where X sub zero is the starting point.and the transition kernel Notation depicting probability of X sub t, given X sub (t-1), where t represents the current time or step..

Each MCMC step that involves the vector of binary outcome variables y in the conditioning set needs first to be modified by defining a pseudo-likelihood using survey weights. In defining pseudo-likelihood, weights are introduced after scaling them to the effective sample size based on a suitable design effect. Note that with the pseudo-likelihood, the covariance matrix of the pseudo-score functions is no longer equal to the pseudo-information matrix; therefore, a sandwich-type of covariance matrix was used to compute the design effect. In this process, weights are largely assumed to be noninformative (i.e., unrelated to the outcome variable y). The assumption of noninformative weights is useful in finding tractable expressions for the appropriate information matrix of the pseudo score functions. The pseudo log-likelihood remains an unbiased estimate of the finite-population log-likelihood regardless of this assumption.

Step I Notation depicting the conditional probability of fixed effects (beta) assuming that the data (y) and the following parameters are known: State random effects (eta), composite field interviewer region effects (nu) within the State. (this does not depend on Notation depicting D sub eta., Notation depicting D sub nu.).

With flat prior for Notation depicting Fixed effect, Beta sub alpha, where alpha denotes a specific age-group., the conditional posterior is proportional to the pseudo-likelihood function. For large samples, this posterior can be approximated by the multivariate normal distribution with mean vector equal to the pseudo-maximum likelihood estimate and with asymptotic covariance matrix having the associated sandwich form. Assuming that the survey weights are noninformative makes the age group specific Notation depicting Fixed effect, Beta sub alpha, where alpha denotes a specific age-group. vectors conditionally independent of each other. Therefore, the Notation depicting Fixed effect, Beta sub alpha, where alpha denotes a specific age-group. can be updated separately at each MCMC cycle.

Step II Notation depicting the conditional probability of State random effects (eta) for State i, assuming that the data (y) and the following parameters are known:  fixed effects (beta), composite field interviewer region effects (nu) and the associated 4 by 4 variance-covariance matrix (D sub eta). (this does not depend on Notation depicting D sub nu.)

Here, the conditional posterior is proportional to the product of the prior Notation depicting the prior distribution of the State i random effects, eta sub i..), the pseudo-likelihood function Notation depicting the psuedo-likelihood function of the data given the parameters. as well as the prior Notation depicting the prior distribution of the fixed effects (beta) and variance-covariance matrix (D sub eta).; this last prior can be omitted as it does not involve This notation depicts the State random effect for State i. To calculate the denominator (or the normalization constant) of the posterior distribution for This notation depicts the State random effect for State i requires multidimensional integration and is numerically intractable. To get around this problem, the Metropolis-Hastings (M-H) algorithm is used that requires a dominating density convenient for Monte Carlo sampling. For this purpose, the mode and curvature of the conditional posterior distribution are used; these can be simply obtained from its numerator. Then a Gaussian distribution is used with matching mode and curvature to define the dominating density for M-H. As with the age group specific Notation depicting the Fixed effect, Beta sub alpha, where alpha denotes a specific age-group. parameters, the State-specific random effect vectors This notation depicts the State random effect for State -i are conditionally independent of each other and can be updated separately at each MCMC cycle.

Step III Notation depicting the conditional probability of composite field interviewer region effects (nu) within the State, assuming that the data (y) and the following parameters are known: fixed effects (beta), State random effects (eta) and the associated 4 by 4 variance-covariance matrix (D sub nu). (this does not depend on Notation depicting D sub eta.)

Similar to step II.

Step IVNotation depicting the conditional probability of D sub eta, given State random effects eta., Conditional probability of D sub nu, given composite FI region random effects nu. (here,etaand nu include all the information from y)

Here, the pseudo-likelihood involving design weights comes in implicitly through the conditioning parameters eta and nu evaluated at the current cycle. An exact conditional posterior distribution is obtained because the inverse Wishart priors for Notation depictin D sub eta. and Notation depictin D sub nu are conjugate.

B.6.8 Remarks

B.7 References

Folsom, R. E., & Judkins, D. R. (1997). Substance abuse in states and metropolitan areas: Model based estimates from the 1991-1993 National Household Surveys on Drug Abuse: Methodology report (DHHS Publication No. SMA 97-3140, Methodology Series M-1; September 1996 summary report available as a WordPerfect file at /methods.htm#methods). Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies.

Folsom , R. E., Shah, B., & Vaish, A. (1999). Substance abuse in states: A methodological report on model based estimates from the 1994-1996 National Household Surveys on Drug Abuse. In Proceedings of the Section on Survey Research Methods of the American Statistical Association (pp. 371-375). Washington, DC: American Statistical Association.

Office of Applied Studies. (1996). Substance abuse in states and metropolitan areas: Model based estimates from the 1991-1993 National Household Surveys on Drug Abuse—Summary report (WordPerfect 6.1 file available at /analytic.htm). Rockville, MD: Substance Abuse and Mental Health Services Administration.

Office of Applied Studies. (2000). Summary of findings from the 1999 National Household Survey on Drug Abuse (DHHS Publication No. SMA 00-3466, NHSDA Series H-12; available at /p0000016.htm#special). Rockville, MD: Substance Abuse and Mental Health Services Administration.

Office of Applied Studies. (2002). National and state estimates of the drug abuse treatment gap: 2000 National Household Survey on Drug Abuse (NHSDA Series H-14, DHHS Publication No. SMA 02-3640, /nsduh.htm). Rockville, MD: Substance Abuse and Mental Health Services Administration.

RTI. (2001). SUDAAN user's manual: Release 8.0. Research Triangle Park, NC: RTI.

Table B.1 Level Estimates (P1) for 2000
State Age in Years Total
12-17 18-25 26+
Past Month Use of Marijuana
CA_TX (design-based) 6.93 11.98 3.19 4.85
CA_TX (average over substates) 7.02 11.49 3.60 5.10
NY_FL (design-based) 6.95 14.61 2.33 4.20
NY_FL (average over substates) 6.84 13.97 2.76 4.46
OH_MI (design-based) 6.95 15.96 3.09 5.17
OH_MI (average over substates) 7.52 15.46 2.83 4.96
PA_IL (design-based) 6.69 14.38 2.90 4.71
PA_IL (average over substates) 6.92 14.23 2.83 4.67
Past Year Use of Cocaine
CA_TX (design-based) 2.12 4.57 0.99 1.62
CA_TX (average over substates) 2.39 4.53 1.19 1.79
NY_FL (design-based) 1.20 4.16 1.06 1.44
NY_FL (average over substates) 1.45 4.14 1.11 1.50
OH_MI (design-based) 0.96 4.20 0.78 1.24
OH_MI (average over substates) 1.30 4.12 0.83 1.31
PA_IL (design-based) 0.97 4.14 1.16 1.51
PA_IL (average over substates) 1.22 4.10 1.18 1.55
Past Month "Binge" Alcohol Use
CA_TX (design-based) 9.64 32.54 19.60 20.29
CA_TX (average over substates) 9.79 32.10 19.39 20.09
NY_FL (design-based) 9.89 37.53 17.60 19.17
NY_FL (average over substates) 10.07 36.75 18.15 19.53
OH_MI (design-based) 11.31 43.60 18.89 21.33
OH_MI (average over substates) 11.15 43.13 19.74 21.90
PA_IL (design-based) 10.35 43.41 21.81 23.37
PA_IL (average over substates) 11.00 42.93 20.88 22.66
Past Month Use of Cigarettes
CA_TX (design-based) 10.08 29.83 23.59 22.94
CA_TX (average over substates) 10.41 30.51 23.44 22.96
NY_FL (design-based) 11.67 34.94 23.38 23.59
NY_FL (average over substates) 12.36 35.14 23.12 23.48
OH_MI (design-based) 14.74 43.48 25.48 26.71
OH_MI (average over substates) 15.14 44.19 25.85 27.13
PA_IL (design-based) 13.93 44.47 24.68 26.08
PA_IL (average over substates) 14.30 43.68 24.45 25.84

Note: The average over substates is calculated as the average of the individual 2000 SAEs (P1) over the 8 substates.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 2000.

 

Table B.2 Relative Absolute Bias of Three Types of Level Estimates for 2000: Past Month Use of Marijuana
State P1 P2 P3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 6.93 11.98 3.19 4.85 6.93 11.98 3.19 4.85 6.93 11.98 3.19 4.85
CA_TX1 9.63 11.08 4.12 5.72 10.00 11.11 4.14 5.78 9.87 10.95 4.12 5.73
CA_TX2 6.16 12.17 3.67 5.15 5.96 12.30 3.51 5.03 5.86 12.00 3.26 4.79
CA_TX3 7.02 10.17 3.21 4.62 7.13 10.20 3.21 4.64 7.19 10.16 3.17 4.61
CA_TX4 6.21 9.50 3.15 4.39 6.43 9.65 3.42 4.64 6.51 9.96 3.57 4.80
CA_TX5 6.50 14.11 3.66 5.45 6.66 14.65 3.76 5.62 6.68 14.72 3.80 5.66
CA_TX6 6.93 11.88 4.00 5.44 6.82 11.72 3.81 5.27 6.71 11.39 3.49 4.99
CA_TX7 7.25 12.64 3.47 5.19 7.00 12.57 3.23 4.97 6.89 12.02 3.10 4.78
CA_TX8 6.49 10.35 3.53 4.82 6.52 10.33 3.61 4.88 6.62 10.46 3.75 5.01
Average over substates 7.02 11.49 3.60 5.10 7.07 11.57 3.59 5.10 7.04 11.46 3.53 5.05
Relative Absolute Bias 1.28 4.13 12.96 5.13 1.89 3.47 12.57 5.26 1.57 4.37 10.84 4.07
NY_FL (design-based) 6.95 14.61 2.33 4.20 6.95 14.61 2.33 4.20 6.95 14.61 2.33 4.20
NY_FL1 6.63 15.27 2.80 4.62 6.57 15.32 2.81 4.63 6.56 15.04 2.71 4.52
NY_FL2 7.28 15.13 2.54 4.46 7.34 15.27 2.55 4.49 7.34 15.23 2.59 4.52
NY_FL3 7.77 13.42 2.63 4.38 8.05 13.57 2.77 4.53 8.04 13.71 2.86 4.62
NY_FL4 6.78 12.33 2.83 4.32 6.81 12.18 2.94 4.38 7.00 12.54 3.09 4.57
NY_FL5 6.81 11.63 2.45 3.94 6.95 11.59 2.60 4.06 7.21 11.79 2.73 4.21
NY_FL6 6.84 15.55 3.30 5.06 6.74 15.65 3.27 5.05 6.71 15.73 3.31 5.08
NY_FL7 5.08 12.70 2.32 3.79 4.92 13.00 2.47 3.93 5.15 13.52 2.66 4.17
NY_FL8 7.56 15.72 3.23 5.10 7.47 15.62 3.16 5.02 7.42 15.22 3.07 4.90
Average over substates 6.84 13.97 2.76 4.46 6.86 14.03 2.82 4.51 6.93 14.10 2.88 4.57
Relative Absolute Bias 1.49 4.39 18.68 6.13 1.30 4.00 21.16 7.41 0.24 3.51 23.51 8.81
OH_MI (design-based) 6.95 15.96 3.09 5.17 6.95 15.96 3.09 5.17 6.95 15.96 3.09 5.17
OH_MI1 6.73 14.45 2.81 4.73 6.59 14.47 2.79 4.71 6.59 14.41 2.74 4.67
OH_MI2 7.13 14.56 2.67 4.68 7.07 14.50 2.62 4.63 7.11 14.66 2.69 4.71
OH_MI3 8.10 15.38 2.69 4.90 7.91 15.23 2.58 4.78 7.72 14.87 2.59 4.73
OH_MI4 8.04 16.97 3.43 5.67 7.56 16.40 3.07 5.27 7.18 15.60 2.76 4.89
OH_MI5 7.49 15.35 2.95 5.03 7.37 15.26 2.85 4.94 7.25 15.13 2.82 4.89
OH_MI6 7.01 18.04 3.21 5.54 6.60 17.98 2.98 5.31 6.40 17.64 2.83 5.13
OH_MI7 7.24 13.27 2.52 4.41 7.12 13.09 2.53 4.38 7.14 13.00 2.53 4.37
OH_MI8 8.41 15.63 2.39 4.74 8.49 15.81 2.43 4.80 8.42 15.43 2.47 4.78
Average over substates 7.52 15.46 2.83 4.96 7.34 15.34 2.73 4.85 7.22 15.09 2.68 4.77
Relative Absolute Bias 8.24 3.14 8.36 3.95 5.64 3.86 11.61 6.09 3.99 5.43 13.33 7.72
PA_IL (design-based) 6.69 14.38 2.90 4.71 6.69 14.38 2.90 4.71 6.69 14.38 2.90 4.71
PA_IL1 7.39 15.74 2.90 4.96 7.19 15.37 2.72 4.75 7.05 14.98 2.57 4.57
PA_IL2 8.50 15.02 3.61 5.53 8.68 14.85 3.62 5.53 8.50 14.69 3.57 5.46
PA_IL3 7.88 15.44 3.08 5.11 7.98 15.45 3.08 5.12 7.84 15.25 2.99 5.01
PA_IL4 5.92 12.18 2.47 4.03 5.79 12.00 2.48 4.00 5.73 11.90 2.33 3.88
PA_IL5 7.14 14.80 3.60 5.36 6.99 14.36 3.40 5.13 6.86 14.02 3.07 4.85
PA_IL6 6.95 15.16 2.32 4.39 6.88 15.46 2.37 4.46 6.87 15.68 2.39 4.50
PA_IL7 6.09 13.06 2.06 3.84 6.28 13.86 2.31 4.15 6.55 14.32 2.47 4.36
PA_IL8 5.45 12.42 2.59 4.10 5.37 12.46 2.77 4.25 5.44 12.69 2.75 4.27
Average over substates 6.92 14.23 2.83 4.67 6.89 14.23 2.84 4.67 6.85 14.19 2.77 4.61
Relative Absolute Bias 3.44 1.02 2.32 1.01 3.11 1.05 1.85 0.84 2.49 1.29 4.42 2.13
Overall 3.61 3.17 10.58 4.05 2.99 3.10 11.80 4.90 2.07 3.65 13.03 5.68

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.3 Relative Absolute Bias of Three Types of Level Estimates for 2000: Past Year Use of Cocaine
State P1 P2 P3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 2.12 4.57 0.99 1.62 2.12 4.57 0.99 1.62 2.12 4.57 0.99 1.62
CA_TX1 2.23 4.65 1.06 1.70 2.21 4.77 1.14 1.77 2.24 4.73 1.17 1.79
CA_TX2 2.83 3.76 1.31 1.83 2.83 3.65 1.27 1.78 2.89 3.62 1.20 1.73
CA_TX3 2.29 4.63 1.10 1.73 2.25 4.58 1.10 1.72 2.28 4.53 1.05 1.68
CA_TX4 2.26 3.72 1.08 1.59 2.25 3.68 1.20 1.67 2.24 3.75 1.24 1.71
CA_TX5 2.42 5.66 1.18 1.95 2.41 5.66 1.19 1.95 2.37 5.69 1.22 1.98
CA_TX6 2.18 4.70 1.00 1.66 2.17 4.82 1.06 1.71 2.13 4.76 1.08 1.71
CA_TX7 2.66 5.24 1.22 1.95 2.70 5.30 1.19 1.94 2.68 5.27 1.23 1.97
CA_TX8 2.25 3.89 1.52 1.93 2.20 3.85 1.50 1.91 2.20 3.90 1.48 1.90
Average over substates 2.39 4.53 1.19 1.79 2.38 4.54 1.21 1.81 2.38 4.53 1.21 1.81
Relative Absolute Bias 12.64 0.74 19.61 10.51 12.14 0.59 21.75 11.48 12.15 0.84 22.06 11.60
NY_FL (design-based) 1.20 4.16 1.06 1.44 1.20 4.16 1.06 1.44 1.20 4.16 1.06 1.44
NY_FL1 1.41 4.18 1.12 1.50 1.37 4.17 1.13 1.51 1.38 4.14 1.11 1.48
NY_FL2 1.44 3.63 0.95 1.31 1.45 3.67 0.98 1.34 1.47 3.69 0.99 1.35
NY_FL3 1.33 3.98 0.99 1.37 1.29 4.09 1.03 1.41 1.28 4.08 1.06 1.44
NY_FL4 1.26 3.28 0.97 1.27 1.23 3.30 1.05 1.33 1.25 3.31 1.06 1.34
NY_FL5 1.43 4.10 1.01 1.41 1.43 4.15 1.05 1.44 1.44 4.13 1.06 1.46
NY_FL6 1.94 5.79 1.46 2.01 1.90 5.64 1.34 1.89 1.91 5.45 1.27 1.81
NY_FL7 1.19 3.30 0.75 1.09 1.14 3.38 0.85 1.17 1.10 3.47 0.84 1.17
NY_FL8 1.63 4.88 1.63 2.01 1.57 4.79 1.48 1.87 1.57 4.74 1.45 1.84
Average over substates 1.45 4.14 1.11 1.50 1.42 4.15 1.11 1.50 1.42 4.13 1.10 1.48
Relative Absolute Bias 20.87 0.32 4.47 4.19 18.23 0.19 4.82 4.22 18.42 0.72 3.75 3.36
OH_MI (design-based) 0.96 4.20 0.78 1.24 0.96 4.20 0.78 1.24 0.96 4.20 0.78 1.24
OH_MI1 1.17 4.59 0.77 1.31 1.17 4.67 0.81 1.35 1.15 4.73 0.87 1.41
OH_MI2 1.07 3.21 0.70 1.06 1.05 3.30 0.79 1.14 1.09 3.29 0.83 1.17
OH_MI3 1.60 3.99 0.85 1.34 1.63 3.92 0.82 1.31 1.66 4.01 0.83 1.33
OH_MI4 1.23 4.94 0.80 1.39 1.19 5.10 0.78 1.39 1.17 5.03 0.76 1.36
OH_MI5 1.38 4.11 0.91 1.37 1.40 4.11 0.90 1.37 1.43 4.10 0.97 1.43
OH_MI6 1.10 4.58 0.72 1.26 1.05 4.68 0.74 1.29 1.04 4.63 0.77 1.30
OH_MI7 1.44 4.17 1.15 1.57 1.36 4.05 1.04 1.47 1.35 4.05 0.98 1.42
OH_MI8 1.39 3.40 0.76 1.17 1.41 3.39 0.79 1.19 1.44 3.38 0.78 1.19
Average over substates 1.30 4.12 0.83 1.31 1.28 4.15 0.83 1.31 1.29 4.15 0.85 1.33
Relative Absolute Bias 35.79 1.78 7.23 5.54 34.19 1.08 7.33 5.77 34.93 1.13 9.24 6.77
PA_IL (design-based) 0.97 4.14 1.16 1.51 0.97 4.14 1.16 1.51 0.97 4.14 1.16 1.51
PA_IL1 1.15 5.16 1.17 1.67 1.09 5.22 1.17 1.67 1.08 5.17 1.18 1.67
PA_IL2 1.09 3.98 1.40 1.69 1.01 3.75 1.38 1.64 1.00 3.83 1.31 1.61
PA_IL3 1.51 4.39 1.36 1.76 1.54 4.43 1.32 1.73 1.50 4.36 1.30 1.71
PA_IL4 1.01 3.55 0.90 1.25 0.97 3.60 1.02 1.34 0.95 3.65 1.02 1.34
PA_IL5 1.08 3.61 1.07 1.39 1.04 3.50 1.13 1.42 1.04 3.52 1.12 1.42
PA_IL6 1.27 4.87 1.15 1.63 1.29 5.04 1.17 1.66 1.27 4.95 1.17 1.65
PA_IL7 1.31 3.76 0.92 1.32 1.39 3.87 0.99 1.39 1.39 3.90 1.06 1.45
PA_IL8 1.33 3.51 1.48 1.72 1.30 3.41 1.47 1.70 1.31 3.44 1.44 1.68
Average over substates 1.22 4.10 1.18 1.55 1.20 4.10 1.21 1.57 1.19 4.10 1.20 1.57
Relative Absolute Bias 25.35 0.83 2.27 2.69 23.90 0.84 4.23 3.76 22.93 0.79 3.95 3.61
Overall 23.66 0.92 8.40 5.73 22.12 0.67 9.54 6.31 22.11 0.87 9.75 6.34

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.4 Relative Absolute Bias of Three Types of Level Estimates for 2000: Past Month "Binge" Alcohol Use
State P1 P2 P3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 9.64 32.54 19.60 20.29 9.64 32.54 19.60 20.29 9.64 32.54 19.60 20.29
CA_TX1 11.21 31.64 19.38 20.18 11.42 31.68 19.36 20.20 11.56 31.88 19.54 20.37
CA_TX2 9.54 28.90 19.71 19.85 9.75 29.02 20.41 20.42 9.77 29.06 20.97 20.81
CA_TX3 9.29 30.16 17.42 18.29 9.39 29.87 17.25 18.13 9.42 29.69 17.13 18.03
CA_TX4 8.97 33.49 19.79 20.50 8.82 34.02 20.37 20.99 8.89 34.47 20.54 21.18
CA_TX5 9.28 32.71 19.01 19.84 9.37 33.42 19.43 20.26 9.62 33.69 20.10 20.81
CA_TX6 9.71 32.28 18.07 19.13 9.81 32.79 18.17 19.28 9.78 33.16 18.22 19.37
CA_TX7 9.77 33.15 21.59 21.88 9.51 33.01 21.64 21.87 9.53 33.03 21.49 21.76
CA_TX8 10.54 34.44 20.16 21.08 10.31 33.73 19.44 20.42 10.24 32.82 18.79 19.79
Average over substates 9.79 32.10 19.39 20.09 9.80 32.19 19.51 20.20 9.85 32.22 19.60 20.26
Relative Absolute Bias 1.56 1.35 1.06 0.98 1.63 1.06 0.45 0.47 2.20 0.96 0.00 0.13
NY_FL (design-based) 9.89 37.53 17.60 19.17 9.89 37.53 17.60 19.17 9.89 37.53 17.60 19.17
NY_FL1 9.94 37.78 17.57 19.18 9.90 38.19 17.51 19.18 10.17 38.50 17.48 19.20
NY_FL2 9.99 37.17 19.81 20.88 9.80 36.91 19.99 20.98 9.71 36.94 19.84 20.85
NY_FL3 10.41 37.40 17.90 19.45 10.49 37.39 17.93 19.48 10.62 37.66 18.01 19.57
NY_FL4 9.86 38.10 18.02 19.56 9.69 38.07 17.87 19.43 9.62 37.76 17.40 19.00
NY_FL5 11.77 36.70 18.80 20.20 12.24 36.80 18.99 20.41 12.64 37.30 19.40 20.82
NY_FL6 9.55 36.69 17.27 18.78 9.51 36.82 17.26 18.79 9.79 36.92 17.47 18.99
NY_FL7 7.92 34.76 15.48 16.99 7.94 35.57 15.56 17.15 8.45 35.68 15.80 17.40
NY_FL8 11.14 35.45 20.33 21.20 11.41 34.84 20.59 21.36 11.50 35.04 20.92 21.64
Average over substates 10.07 36.75 18.15 19.53 10.12 36.82 18.22 19.60 10.31 36.97 18.29 19.69
Relative Absolute Bias 1.84 2.06 3.11 1.87 2.36 1.88 3.49 2.21 4.25 1.47 3.91 2.68
OH_MI (design-based) 11.31 43.60 18.89 21.33 11.31 43.60 18.89 21.33 11.31 43.60 18.89 21.33
OH_MI1 9.90 42.49 16.77 19.41 9.94 42.72 16.48 19.23 10.14 42.42 16.42 19.16
OH_MI2 10.80 40.97 18.80 20.86 10.80 40.57 18.47 20.56 10.66 40.41 18.39 20.46
OH_MI3 11.14 41.19 19.55 21.50 11.36 41.32 19.57 21.56 11.60 41.35 20.04 21.94
OH_MI4 11.74 43.98 21.12 23.13 11.74 43.61 20.86 22.88 11.44 43.46 20.51 22.56
OH_MI5 11.77 44.37 19.21 21.72 11.63 44.24 18.56 21.19 11.47 43.25 18.06 20.65
OH_MI6 9.84 44.45 19.44 21.70 9.56 44.66 19.26 21.56 9.41 44.26 19.14 21.41
OH_MI7 11.59 43.67 21.65 23.48 11.53 43.38 21.46 23.29 11.24 43.19 21.04 22.90
OH_MI8 12.39 43.95 21.38 23.39 12.43 43.62 21.12 23.15 12.12 43.71 20.84 22.91
Average over substates 11.15 43.13 19.74 21.90 11.12 43.02 19.47 21.68 11.01 42.76 19.30 21.50
Relative Absolute Bias 1.41 1.07 4.48 2.68 1.63 1.34 3.07 1.64 2.62 1.94 2.18 0.80
PA_IL (design-based) 10.35 43.41 21.81 23.37 10.35 43.41 21.81 23.37 10.35 43.41 21.81 23.37
PA_IL1 10.47 43.48 20.59 22.45 10.13 43.74 20.76 22.58 10.23 43.71 20.87 22.67
PA_IL2 10.73 42.29 21.37 22.93 10.59 42.26 21.55 23.05 10.48 42.23 21.67 23.13
PA_IL3 12.59 44.10 23.07 24.66 12.36 43.38 22.89 24.41 11.90 43.18 22.45 23.99
PA_IL4 9.61 42.33 20.13 21.86 9.27 42.37 20.13 21.83 9.37 42.05 19.75 21.50
PA_IL5 11.52 44.27 20.80 22.81 11.35 44.06 20.71 22.70 11.26 43.75 20.76 22.69
PA_IL6 12.20 47.99 21.00 23.50 12.03 48.47 20.78 23.37 12.40 48.07 20.88 23.44
PA_IL7 10.27 40.46 18.79 20.65 10.37 40.84 19.17 21.01 10.50 41.38 19.62 21.44
PA_IL8 10.59 38.50 21.30 22.39 10.47 37.79 21.58 22.50 10.28 37.76 21.25 22.23
Average over substates 11.00 42.93 20.88 22.66 10.82 42.86 20.95 22.68 10.80 42.77 20.91 22.64
Relative Absolute Bias 6.30 1.12 4.24 3.05 4.60 1.26 3.95 2.94 4.42 1.49 4.13 3.13
Overall 2.78 1.40 3.22 2.14 2.56 1.38 2.74 1.82 3.37 1.46 2.56 1.69

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.5 Relative Absolute Bias of Three Types of Level Estimates for 2000: Past Month Use of Cigarettes
State P1 P2 P3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 10.08 29.83 23.59 22.94 10.08 29.83 23.59 22.94 10.08 29.83 23.59 22.94
CA_TX1 11.78 30.28 23.44 23.08 11.95 29.86 23.34 22.97 12.04 29.52 23.31 22.90
CA_TX2 10.51 26.51 22.34 21.59 10.84 26.15 22.50 21.69 10.98 26.18 22.57 21.76
CA_TX3 10.53 32.31 22.40 22.45 10.42 32.47 21.98 22.15 10.38 32.66 21.79 22.03
CA_TX4 10.32 30.64 22.48 22.25 10.36 30.67 22.40 22.20 10.34 30.61 22.34 22.15
CA_TX5 10.60 31.40 24.99 24.27 10.58 31.46 25.27 24.48 10.58 31.65 25.35 24.56
CA_TX6 9.13 27.76 22.55 21.77 9.07 28.13 22.44 21.73 9.07 28.53 22.47 21.81
CA_TX7 10.63 34.24 24.94 24.63 10.40 34.52 25.08 24.75 10.50 34.21 24.89 24.57
CA_TX8 9.79 30.91 24.41 23.67 9.53 30.80 24.56 23.73 9.49 30.71 24.37 23.58
Average over substates 10.41 30.51 23.44 22.96 10.39 30.51 23.45 22.96 10.42 30.51 23.39 22.92
Relative Absolute Bias 3.32 2.27 0.63 0.10 3.15 2.26 0.63 0.09 3.44 2.27 0.88 0.09
NY_FL (design-based) 11.67 34.94 23.38 23.59 11.67 34.94 23.38 23.59 11.67 34.94 23.38 23.59
NY_FL1 11.89 38.29 24.44 24.84 11.45 38.49 24.61 24.95 11.60 38.10 24.56 24.88
NY_FL2 11.94 32.77 21.92 22.22 11.99 32.65 22.00 22.27 11.99 33.22 22.08 22.39
NY_FL3 13.05 34.88 23.61 23.90 13.04 34.47 23.64 23.88 13.12 34.50 23.48 23.76
NY_FL4 11.19 32.95 21.91 22.16 11.12 33.23 21.75 22.05 11.40 33.87 21.49 21.93
NY_FL5 13.63 37.49 24.08 24.63 13.71 37.38 24.15 24.68 13.79 37.59 24.00 24.59
NY_FL6 13.03 35.51 24.82 24.92 12.95 35.25 25.13 25.13 13.15 34.94 24.97 24.99
NY_FL7 10.77 32.25 22.21 22.27 10.74 32.62 22.23 22.33 10.91 32.98 22.49 22.59
NY_FL8 13.41 37.00 21.97 22.89 13.35 36.90 21.62 22.59 13.64 36.92 21.39 22.43
Average over substates 12.36 35.14 23.12 23.48 12.29 35.12 23.14 23.49 12.45 35.26 23.06 23.45
Relative Absolute Bias 5.97 0.57 1.11 0.48 5.37 0.52 1.02 0.45 6.72 0.92 1.38 0.62
OH_MI (design-based) 14.74 43.48 25.48 26.71 14.74 43.48 25.48 26.71 14.74 43.48 25.48 26.71
OH_MI1 14.49 43.65 25.30 26.57 14.73 43.77 25.20 26.53 14.92 43.88 25.44 26.76
OH_MI2 16.30 42.63 26.60 27.62 16.69 41.96 26.70 27.65 16.56 42.18 26.51 27.51
OH_MI3 15.11 42.62 22.57 24.41 15.39 42.55 22.24 24.18 15.70 42.35 22.35 24.27
OH_MI4 14.95 47.16 27.41 28.70 14.97 47.37 27.56 28.84 15.04 47.28 27.69 28.94
OH_MI5 15.98 46.61 25.05 26.92 15.89 46.48 24.70 26.63 16.00 45.82 24.49 26.39
OH_MI6 12.68 46.71 27.11 28.17 12.21 47.39 27.30 28.35 12.30 47.16 27.58 28.55
OH_MI7 15.15 41.87 26.73 27.50 15.42 41.66 27.06 27.76 15.51 41.92 27.34 28.02
OH_MI8 16.42 42.27 26.00 27.13 16.77 41.77 26.06 27.15 16.89 41.98 25.90 27.06
Average over substates 15.14 44.19 25.85 27.13 15.26 44.12 25.85 27.14 15.37 44.07 25.91 27.19
Relative Absolute Bias 2.65 1.64 1.44 1.55 3.48 1.48 1.47 1.58 4.22 1.37 1.71 1.78
PA_IL (design-based) 13.93 44.47 24.68 26.08 13.93 44.47 24.68 26.08 13.93 44.47 24.68 26.08
PA_IL1 13.84 42.31 24.66 25.79 13.71 42.12 24.75 25.82 13.79 42.33 24.45 25.61
PA_IL2 14.99 40.62 24.98 25.94 15.07 40.23 25.19 26.06 15.21 40.21 25.19 26.07
PA_IL3 14.14 45.30 23.91 25.62 13.97 45.04 23.61 25.33 13.94 44.78 23.53 25.24
PA_IL4 12.83 42.19 23.17 24.52 12.55 42.26 23.08 24.44 12.28 42.81 22.97 24.36
PA_IL5 13.52 48.49 25.96 27.54 13.05 48.84 26.00 27.57 12.86 48.63 26.59 27.96
PA_IL6 15.38 44.65 24.44 26.06 15.43 44.76 24.24 25.94 15.22 44.61 24.23 25.88
PA_IL7 16.07 45.51 24.38 26.20 16.25 45.68 24.19 26.09 16.08 45.40 24.37 26.18
PA_IL8 13.65 40.35 24.06 25.06 13.77 40.25 24.11 25.10 13.76 40.52 23.93 24.98
Average over substates 14.30 43.68 24.45 25.84 14.23 43.65 24.40 25.79 14.14 43.66 24.41 25.79
Relative Absolute Bias 2.64 1.79 0.93 0.93 2.10 1.85 1.13 1.11 1.51 1.82 1.09 1.14
Overall 3.65 1.57 1.03 0.76 3.53 1.53 1.06 0.81 3.97 1.59 1.27 0.91

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 2000.

Table B.6 Relative Absolute Bias of Three Types of Level Estimates for 2000
State P1 P2 P3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
Past Month Use of Marijuana
CA_TX 1.28 4.13 12.96 5.13 1.89 3.47 12.57 5.26 1.57 4.37 10.84 4.07
NY_FL 1.49 4.39 18.68 6.13 1.30 4.00 21.16 7.41 0.24 3.51 23.51 8.81
OH_MI 8.24 3.14 8.36 3.95 5.64 3.86 11.61 6.09 3.99 5.43 13.33 7.72
PA_IL 3.44 1.02 2.32 1.01 3.11 1.05 1.85 0.84 2.49 1.29 4.42 2.13
Average 3.61 3.17 10.58 4.05 2.99 3.10 11.80 4.90 2.07 3.65 13.03 5.68
Past Year Use of Cocaine
CA_TX 12.64 0.74 19.61 10.51 12.14 0.59 21.75 11.48 12.15 0.84 22.06 11.60
NY_FL 20.87 0.32 4.47 4.19 18.23 0.19 4.82 4.22 18.42 0.72 3.75 3.36
OH_MI 35.79 1.78 7.23 5.54 34.19 1.08 7.33 5.77 34.93 1.13 9.24 6.77
PA_IL 25.35 0.83 2.27 2.69 23.90 0.84 4.23 3.76 22.93 0.79 3.95 3.61
Average 23.66 0.92 8.40 5.73 22.12 0.67 9.54 6.31 22.11 0.87 9.75 6.34
Past Month "Binge" Alcohol Use
CA_TX 1.56 1.35 1.06 0.98 1.63 1.06 0.45 0.47 2.20 0.96 0.00 0.13
NY_FL 1.84 2.06 3.11 1.87 2.36 1.88 3.49 2.21 4.25 1.47 3.91 2.68
OH_MI 1.41 1.07 4.48 2.68 1.63 1.34 3.07 1.64 2.62 1.94 2.18 0.80
PA_IL 6.30 1.12 4.24 3.05 4.60 1.26 3.95 2.94 4.42 1.49 4.13 3.13
Average 2.78 1.40 3.22 2.14 2.56 1.38 2.74 1.82 3.37 1.46 2.56 1.69
Past Month Use of Cigarettes
CA_TX 3.32 2.27 0.63 0.10 3.15 2.26 0.63 0.09 3.44 2.27 0.88 0.09
NY_FL 5.97 0.57 1.11 0.48 5.37 0.52 1.02 0.45 6.72 0.92 1.38 0.62
OH_MI 2.65 1.64 1.44 1.55 3.48 1.48 1.47 1.58 4.22 1.37 1.71 1.78
PA_IL 2.64 1.79 0.93 0.93 2.10 1.85 1.13 1.11 1.51 1.82 1.09 1.14
Average 3.65 1.57 1.03 0.76 3.53 1.53 1.06 0.81 3.97 1.59 1.27 0.91
Overall 8.43 1.76 5.81 3.17 7.80 1.67 6.28 3.46 7.88 1.89 6.65 3.65

Note: Relative absolute bias (Pi) = 100 × abs(mean of 8 substate Pi - P) / P, i = 1,2, and 3.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.7 Average 95 Percent Lower and Upper Bounds for Level Estimates (P1) for 2000
State Age in Years Total
12-17 18-25 26+
Lower Upper Lower Upper Lower Upper Lower Upper
Past Month Use of Marijuana
CA_TX (design-based) 5.96 8.06 10.32 13.87 2.50 4.06 4.24 5.54
CA_TX (avg. over substates) 5.17 9.28 8.69 14.82 2.23 5.48 3.80 6.68
NY_FL (design-based) 5.96 8.08 12.85 16.56 1.76 3.08 3.66 4.82
NY_FL (avg. over substates) 4.86 9.33 10.73 17.75 1.71 4.22 3.37 5.78
OH_MI (design-based) 6.05 7.97 14.12 17.98 2.42 3.95 4.53 5.90
OH_MI (avg. over substates) 5.44 10.09 12.07 19.36 1.78 4.26 3.85 6.29
PA_IL (design-based) 5.72 7.80 12.27 16.77 2.20 3.81 4.02 5.53
PA_IL (avg. over substates) 5.02 9.26 10.95 18.06 1.78 4.28 3.56 5.99
Past Year Use of Cocaine
CA_TX (design-based) 1.67 2.69 3.69 5.65 0.58 1.68 1.25 2.10
CA_TX (avg. over substates) 1.52 3.57 2.97 6.61 0.55 2.23 1.18 2.62
NY_FL (design-based) 0.76 1.90 3.37 5.11 0.72 1.57 1.11 1.85
NY_FL (avg. over substates) 0.83 2.38 2.64 6.16 0.53 2.05 0.94 2.26
OH_MI (design-based) 0.64 1.43 3.43 5.13 0.50 1.21 0.97 1.58
OH_MI (avg. over substates) 0.72 2.16 2.66 6.10 0.40 1.53 0.86 1.91
PA_IL (design-based) 0.68 1.39 3.21 5.31 0.75 1.78 1.15 1.99
PA_IL (avg. over substates) 0.68 2.01 2.61 6.12 0.57 2.19 0.97 2.35
Past Month "Binge" Alcohol Use
CA_TX (design-based) 8.60 10.79 30.28 34.88 17.81 21.51 18.88 21.78
CA_TX (avg. over substates) 7.70 12.22 27.72 36.72 15.84 23.35 17.21 23.23
NY_FL (design-based) 8.72 11.20 35.04 40.08 15.88 19.46 17.73 20.70
NY_FL (avg. over substates) 7.73 12.84 31.92 41.79 14.76 21.94 16.65 22.68
OH_MI (design-based) 10.02 12.73 40.64 46.61 17.13 20.79 19.86 22.87
OH_MI (avg. over substates) 8.69 14.02 38.19 48.18 16.39 23.44 19.10 24.91
PA_IL (design-based) 9.14 11.69 40.46 46.41 19.99 23.74 21.87 24.94
PA_IL (avg. over substates) 8.61 13.78 38.01 47.96 17.39 24.72 19.71 25.82
Past Month Use of Cigarettes
CA_TX (design-based) 8.96 11.31 27.34 32.45 21.26 26.09 21.14 24.85
CA_TX (avg. over substates) 8.25 12.91 26.17 35.12 19.43 27.85 19.82 26.35
NY_FL (design-based) 10.15 13.38 32.68 37.28 21.64 25.21 22.19 25.05
NY_FL (avg. over substates) 9.73 15.41 30.38 40.13 19.35 27.23 20.38 26.80
OH_MI (design-based) 13.39 16.21 40.89 46.10 23.64 27.41 25.25 28.23
OH_MI (avg. over substates) 12.17 18.50 39.17 49.29 22.00 29.99 24.01 30.42
PA_IL (design-based) 12.54 15.45 42.28 46.68 22.63 26.85 24.36 27.88
PA_IL (avg. over substates) 11.53 17.45 38.63 48.82 20.65 28.56 22.73 29.14

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 2000.

Table B.8 Ratio of Widths of 95 Percent Confidence Intervals of Level Estimates for 2000
State W1 W2 W3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
Past Month Use of Marijuana
CA_TX 0.76 0.69 0.62 0.70 0.78 0.72 0.60 0.65 0.79 0.72 0.61 0.67
NY_FL 0.67 0.74 0.66 0.72 0.72 0.76 0.66 0.69 0.74 0.77 0.71 0.73
OH_MI 0.77 0.68 0.57 0.66 0.79 0.70 0.54 0.61 0.79 0.69 0.55 0.62
PA_IL 0.71 0.64 0.57 0.61 0.76 0.67 0.57 0.58 0.76 0.67 0.58 0.60
Average 0.72 0.69 0.60 0.67 0.76 0.71 0.59 0.63 0.77 0.72 0.61 0.65
Past Year Use of Cocaine
CA_TX 0.65 0.71 0.51 0.62 0.69 0.73 0.47 0.57 0.71 0.75 0.51 0.59
NY_FL 0.50 0.70 0.59 0.72 0.54 0.75 0.55 0.67 0.57 0.77 0.57 0.69
OH_MI 0.52 0.65 0.49 0.59 0.58 0.70 0.47 0.56 0.62 0.73 0.51 0.58
PA_IL 0.53 0.62 0.62 0.67 0.59 0.66 0.59 0.63 0.62 0.68 0.63 0.64
Average 0.55 0.67 0.55 0.65 0.60 0.71 0.52 0.61 0.63 0.73 0.55 0.62
Past Month "Binge" Alcohol Use
CA_TX 0.68 0.73 0.65 0.67 0.74 0.78 0.71 0.71 0.75 0.78 0.72 0.71
NY_FL 0.64 0.69 0.73 0.73 0.71 0.71 0.76 0.74 0.73 0.71 0.77 0.75
OH_MI 0.66 0.66 0.77 0.74 0.73 0.68 0.81 0.77 0.73 0.68 0.81 0.76
PA_IL 0.72 0.63 0.70 0.69 0.77 0.67 0.75 0.71 0.78 0.67 0.75 0.72
Average 0.68 0.68 0.71 0.71 0.74 0.71 0.76 0.73 0.75 0.71 0.76 0.74
Past Month Use of Cigarettes
CA_TX 0.69 0.74 0.61 0.61 0.75 0.79 0.67 0.66 0.76 0.79 0.68 0.67
NY_FL 0.65 0.69 0.74 0.75 0.71 0.72 0.81 0.81 0.73 0.73 0.82 0.81
OH_MI 0.77 0.77 0.72 0.71 0.84 0.80 0.80 0.77 0.86 0.81 0.81 0.78
PA_IL 0.68 0.75 0.68 0.66 0.73 0.80 0.74 0.71 0.74 0.81 0.75 0.72
Average 0.70 0.74 0.69 0.68 0.76 0.78 0.76 0.74 0.77 0.79 0.76 0.74
Overall 0.66 0.69 0.64 0.68 0.71 0.73 0.66 0.68 0.73 0.74 0.67 0.69

W1 = WP1 / WP, W2 = WP2 / WP, and W3 = WP3 / WP

where

WP1 = mean of widths of PIs of P1 over substates,
WP2 = mean of widths of PIs of P2 over substates,
WP3 = mean of widths of PIs of P3 over substates, and
WP = mean of widths of CIs of 2000 design-based estimate over substates.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.9 Relative Absolute Bias of Three Types of Change Estimates for 2000/1999: Past Month Use of Marijuana
State R1 R2 R3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 0.99 0.92 1.06 1.00 0.99 0.92 1.06 1.00 0.99 0.92 1.06 1.00
CA_TX1 1.13 0.76 1.20 1.03 1.17 0.77 1.20 1.04 1.16 0.75 1.20 1.03
CA_TX2 0.96 1.16 1.24 1.17 0.93 1.17 1.18 1.14 0.91 1.14 1.10 1.09
CA_TX3 0.97 0.86 1.07 0.98 0.99 0.87 1.08 0.99 1.00 0.86 1.06 0.98
CA_TX4 0.85 0.72 0.91 0.84 0.88 0.74 0.99 0.89 0.89 0.76 1.04 0.92
CA_TX5 0.71 0.98 1.00 0.95 0.73 1.02 1.03 0.98 0.73 1.02 1.04 0.98
CA_TX6 0.93 1.00 1.48 1.21 0.92 0.99 1.41 1.17 0.90 0.96 1.29 1.11
CA_TX7 1.39 1.13 1.15 1.18 1.34 1.12 1.07 1.13 1.32 1.07 1.03 1.08
CA_TX8 1.10 0.75 1.03 0.94 1.10 0.75 1.05 0.95 1.12 0.76 1.09 0.97
Average over substates 1.01 0.92 1.14 1.04 1.01 0.93 1.13 1.04 1.00 0.92 1.11 1.02
Relative Absolute Bias 1.27 0.47 7.01 3.72 1.50 1.07 6.27 3.58 1.16 0.03 4.23 2.18
NY_FL (design-based) 1.22 0.91 0.78 0.88 1.22 0.91 0.78 0.88 1.22 0.91 0.78 0.88
NY_FL1 0.94 1.03 1.13 1.06 0.93 1.03 1.14 1.07 0.93 1.02 1.09 1.04
NY_FL2 1.17 1.02 0.87 0.97 1.18 1.03 0.88 0.98 1.18 1.03 0.89 0.98
NY_FL3 1.10 0.84 0.85 0.88 1.14 0.85 0.90 0.91 1.14 0.85 0.93 0.93
NY_FL4 1.15 0.77 0.87 0.87 1.16 0.76 0.90 0.88 1.19 0.79 0.95 0.92
NY_FL5 1.05 0.72 0.90 0.85 1.07 0.72 0.96 0.88 1.11 0.73 1.01 0.91
NY_FL6 1.02 0.99 1.07 1.04 1.00 1.00 1.07 1.03 1.00 1.00 1.08 1.04
NY_FL7 0.88 0.90 0.75 0.82 0.86 0.92 0.80 0.85 0.90 0.96 0.86 0.90
NY_FL8 1.06 1.03 1.20 1.12 1.05 1.02 1.17 1.10 1.05 1.00 1.14 1.07
Average over substates 1.05 0.91 0.96 0.95 1.05 0.92 0.98 0.96 1.06 0.92 0.99 0.97
Relative Absolute Bias 14.24 0.19 23.35 7.73 14.11 0.25 25.80 8.99 13.10 0.76 27.95 10.34
OH_MI (design-based) 1.02 1.08 1.09 1.08 1.02 1.08 1.09 1.08 1.02 1.08 1.09 1.08
OH_MI1 0.84 0.99 1.11 1.02 0.82 0.99 1.11 1.01 0.82 0.99 1.09 1.00
OH_MI2 1.09 0.91 0.91 0.94 1.08 0.91 0.90 0.93 1.08 0.92 0.92 0.94
OH_MI3 1.27 1.00 0.95 1.01 1.24 0.99 0.91 0.99 1.21 0.97 0.91 0.98
OH_MI4 1.13 1.29 1.50 1.35 1.06 1.25 1.34 1.25 1.01 1.19 1.20 1.17
OH_MI5 0.96 1.07 1.02 1.03 0.94 1.06 0.99 1.01 0.93 1.06 0.98 1.00
OH_MI6 0.93 1.29 1.21 1.19 0.88 1.28 1.12 1.14 0.85 1.26 1.06 1.11
OH_MI7 1.03 0.94 1.02 0.99 1.01 0.92 1.02 0.98 1.02 0.92 1.02 0.98
OH_MI8 1.06 0.98 0.87 0.95 1.07 0.99 0.89 0.96 1.06 0.97 0.90 0.96
Average over substates 1.04 1.06 1.07 1.06 1.01 1.05 1.03 1.04 1.00 1.03 1.01 1.02
Relative Absolute Bias 1.24 1.83 1.11 1.43 1.22 2.61 4.78 3.74 2.78 4.26 6.87 5.52
PA_IL (design-based) 0.87 1.02 1.20 1.07 0.87 1.02 1.20 1.07 0.87 1.02 1.20 1.07
PA_IL1 1.23 1.17 1.14 1.17 1.20 1.14 1.07 1.12 1.17 1.11 1.01 1.08
PA_IL2 1.06 0.90 1.19 1.05 1.08 0.89 1.19 1.05 1.06 0.88 1.17 1.04
PA_IL3 0.97 0.94 1.17 1.04 0.98 0.94 1.17 1.05 0.96 0.93 1.13 1.02
PA_IL4 1.01 0.94 1.13 1.03 0.99 0.92 1.13 1.03 0.98 0.92 1.06 0.99
PA_IL5 0.86 1.09 1.55 1.24 0.84 1.05 1.46 1.19 0.82 1.03 1.32 1.12
PA_IL6 0.99 1.09 0.82 0.95 0.98 1.11 0.84 0.96 0.98 1.13 0.84 0.97
PA_IL7 0.78 0.81 0.72 0.77 0.80 0.86 0.80 0.83 0.84 0.89 0.86 0.87
PA_IL8 0.75 0.89 1.08 0.95 0.74 0.90 1.16 0.98 0.75 0.91 1.15 0.99
Average over substates 0.96 0.98 1.10 1.03 0.95 0.98 1.10 1.03 0.94 0.97 1.07 1.01
Relative Absolute Bias 10.45 4.08 8.70 4.45 9.91 4.22 8.43 4.43 9.22 4.49 11.13 5.80
Overall 6.80 1.64 10.04 4.33 6.69 2.04 11.32 5.18 6.57 2.39 12.55 5.96

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.10 Relative Absolute Bias of Three Types of Change Estimates for 2000/1999: Past Year Use of Cocaine
State R1 R2 R3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 0.86 0.75 0.80 0.79 0.86 0.75 0.80 0.79 0.86 0.75 0.80 0.79
CA_TX1 0.96 0.68 0.70 0.72 0.95 0.69 0.75 0.75 0.97 0.69 0.77 0.76
CA_TX2 1.14 0.66 1.03 0.90 1.14 0.64 0.99 0.88 1.16 0.63 0.94 0.85
CA_TX3 1.09 0.96 0.89 0.94 1.07 0.95 0.89 0.94 1.09 0.94 0.85 0.92
CA_TX4 0.89 0.59 0.65 0.66 0.89 0.59 0.72 0.69 0.89 0.60 0.75 0.71
CA_TX5 1.01 0.99 0.73 0.86 1.01 0.99 0.73 0.86 0.99 1.00 0.75 0.87
CA_TX6 0.91 0.70 0.72 0.74 0.91 0.72 0.76 0.76 0.89 0.71 0.77 0.76
CA_TX7 1.26 0.84 0.76 0.84 1.28 0.85 0.74 0.84 1.27 0.85 0.76 0.85
CA_TX8 0.87 0.62 1.05 0.87 0.86 0.62 1.04 0.85 0.85 0.63 1.03 0.85
Average over substates 1.02 0.76 0.81 0.82 1.01 0.76 0.83 0.82 1.01 0.75 0.83 0.82
Relative Absolute Bias 18.13 0.70 2.23 3.33 17.63 0.74 3.86 4.10 17.66 0.47 3.78 4.08
NY_FL (design-based) 0.82 0.77 0.96 0.88 0.82 0.77 0.96 0.88 0.82 0.77 0.96 0.88
NY_FL1 0.87 0.85 0.71 0.76 0.85 0.85 0.71 0.76 0.86 0.84 0.70 0.75
NY_FL2 0.89 0.75 0.59 0.66 0.90 0.76 0.61 0.68 0.91 0.76 0.61 0.68
NY_FL3 0.82 0.71 0.58 0.64 0.80 0.73 0.61 0.66 0.79 0.73 0.63 0.67
NY_FL4 0.81 0.64 0.60 0.62 0.79 0.64 0.65 0.66 0.81 0.64 0.65 0.66
NY_FL5 1.03 0.78 0.59 0.67 1.03 0.79 0.61 0.69 1.03 0.79 0.62 0.69
NY_FL6 1.25 1.14 0.90 1.00 1.23 1.11 0.83 0.94 1.23 1.07 0.78 0.90
NY_FL7 0.60 0.77 0.51 0.59 0.58 0.79 0.57 0.63 0.56 0.81 0.57 0.63
NY_FL8 1.05 0.98 0.90 0.94 1.01 0.97 0.82 0.88 1.01 0.96 0.80 0.86
Average over substates 0.92 0.83 0.67 0.74 0.90 0.83 0.68 0.74 0.90 0.83 0.67 0.73
Relative Absolute Bias 11.60 7.11 29.72 16.12 9.25 7.26 29.32 16.02 9.50 6.75 30.05 16.72
OH_MI (design-based) 1.15 0.72 0.77 0.77 1.15 0.72 0.77 0.77 1.15 0.72 0.77 0.77
OH_MI1 1.07 0.82 0.63 0.74 1.07 0.83 0.65 0.75 1.04 0.84 0.71 0.79
OH_MI2 1.14 0.51 0.60 0.59 1.12 0.53 0.67 0.63 1.15 0.52 0.71 0.65
OH_MI3 1.80 0.86 0.78 0.87 1.83 0.84 0.75 0.85 1.86 0.86 0.76 0.87
OH_MI4 1.20 1.08 0.97 1.04 1.15 1.11 0.94 1.03 1.14 1.09 0.91 1.01
OH_MI5 1.53 0.76 0.72 0.78 1.56 0.76 0.72 0.78 1.58 0.76 0.77 0.81
OH_MI6 1.15 1.00 0.71 0.86 1.10 1.02 0.73 0.88 1.09 1.01 0.76 0.89
OH_MI7 1.48 0.95 1.26 1.15 1.40 0.93 1.14 1.07 1.39 0.93 1.07 1.04
OH_MI8 1.51 0.69 0.81 0.80 1.53 0.69 0.84 0.82 1.56 0.68 0.83 0.82
Average over substates 1.36 0.83 0.81 0.85 1.34 0.84 0.81 0.85 1.35 0.84 0.81 0.86
Relative Absolute Bias 17.73 15.76 5.15 11.09 16.45 16.49 4.79 11.04 17.19 16.40 5.97 11.77
PA_IL (design-based) 0.84 0.99 1.22 1.10 0.84 0.99 1.22 1.10 0.84 0.99 1.22 1.10
PA_IL1 1.14 1.11 1.32 1.22 1.09 1.12 1.32 1.23 1.08 1.11 1.34 1.23
PA_IL2 0.90 1.17 1.60 1.38 0.83 1.10 1.58 1.34 0.83 1.12 1.51 1.32
PA_IL3 1.32 0.96 1.53 1.28 1.34 0.97 1.47 1.26 1.31 0.95 1.46 1.24
PA_IL4 0.76 0.86 1.07 0.96 0.73 0.88 1.21 1.03 0.71 0.89 1.21 1.03
PA_IL5 0.91 0.75 1.27 1.01 0.87 0.73 1.34 1.03 0.87 0.73 1.32 1.03
PA_IL6 1.12 1.05 1.27 1.17 1.14 1.09 1.29 1.20 1.12 1.07 1.29 1.19
PA_IL7 1.17 0.74 0.92 0.87 1.24 0.76 0.99 0.91 1.24 0.77 1.06 0.95
PA_IL8 1.26 0.85 1.58 1.28 1.23 0.83 1.56 1.26 1.25 0.83 1.53 1.25
Average over substates 1.07 0.94 1.32 1.15 1.06 0.93 1.35 1.16 1.05 0.93 1.34 1.15
Relative Absolute Bias 28.03 5.70 8.09 3.77 26.59 5.97 10.21 4.76 25.65 5.82 9.80 4.54
Overall 18.87 7.32 11.30 8.58 17.48 7.61 12.04 8.98 17.50 7.36 12.40 9.28

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.11 Relative Absolute Bias of Three Types of Change Estimates for 2000/1999: Past Month "Binge" Alcohol Use
State R1 R2 R3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 0.95 0.93 1.05 1.02 0.95 0.93 1.05 1.02 0.95 0.93 1.05 1.02
CA_TX1 1.10 0.90 1.00 0.98 1.12 0.90 0.99 0.98 1.14 0.91 1.00 0.99
CA_TX2 0.90 0.81 1.08 1.00 0.92 0.81 1.12 1.03 0.92 0.81 1.15 1.05
CA_TX3 1.01 0.96 1.12 1.07 1.02 0.95 1.11 1.06 1.02 0.94 1.10 1.06
CA_TX4 0.82 0.97 1.00 0.98 0.81 0.99 1.03 1.01 0.82 1.00 1.04 1.02
CA_TX5 0.77 0.85 1.00 0.95 0.77 0.87 1.02 0.97 0.79 0.88 1.06 1.00
CA_TX6 0.85 0.96 0.97 0.96 0.86 0.98 0.98 0.97 0.85 0.99 0.98 0.98
CA_TX7 1.14 1.02 1.02 1.03 1.11 1.02 1.03 1.03 1.11 1.02 1.02 1.02
CA_TX8 1.38 1.01 1.21 1.17 1.35 0.99 1.17 1.13 1.34 0.97 1.13 1.10
Average over substates 1.00 0.94 1.05 1.02 0.99 0.94 1.06 1.02 1.00 0.94 1.06 1.03
Relative Absolute Bias 4.87 0.55 0.32 0.41 4.77 0.80 0.83 0.84 5.26 0.89 1.21 1.11
NY_FL (design-based) 1.17 1.00 0.94 0.96 1.17 1.00 0.94 0.96 1.17 1.00 0.94 0.96
NY_FL1 1.20 1.05 0.87 0.92 1.19 1.06 0.86 0.92 1.23 1.07 0.86 0.92
NY_FL2 1.08 0.99 1.16 1.12 1.06 0.99 1.17 1.12 1.05 0.99 1.16 1.11
NY_FL3 1.07 1.03 0.97 0.99 1.08 1.03 0.97 0.99 1.10 1.04 0.98 1.00
NY_FL4 1.21 1.14 1.04 1.07 1.19 1.14 1.03 1.06 1.18 1.13 1.00 1.03
NY_FL5 1.22 0.91 0.91 0.93 1.27 0.92 0.92 0.93 1.31 0.93 0.94 0.95
NY_FL6 1.14 0.99 0.89 0.92 1.14 0.99 0.89 0.92 1.17 1.00 0.90 0.93
NY_FL7 0.88 0.90 0.90 0.90 0.88 0.92 0.91 0.91 0.94 0.92 0.92 0.92
NY_FL8 1.01 0.87 1.15 1.08 1.03 0.85 1.17 1.09 1.04 0.86 1.19 1.10
Average over substates 1.10 0.99 0.99 0.99 1.11 0.99 0.99 0.99 1.13 0.99 0.99 1.00
Relative Absolute Bias 6.17 1.21 5.30 3.05 5.81 1.02 5.70 3.38 4.06 0.63 6.09 3.81
OH_MI (design-based) 1.19 1.08 0.92 0.97 1.19 1.08 0.92 0.97 1.19 1.08 0.92 0.97
OH_MI1 0.93 1.04 0.85 0.90 0.94 1.04 0.84 0.90 0.96 1.04 0.84 0.89
OH_MI2 1.14 1.07 0.94 0.98 1.14 1.06 0.92 0.96 1.13 1.05 0.92 0.96
OH_MI3 1.04 0.92 0.94 0.94 1.06 0.92 0.94 0.94 1.08 0.92 0.97 0.96
OH_MI4 1.13 1.12 1.04 1.06 1.13 1.11 1.03 1.05 1.10 1.10 1.01 1.04
OH_MI5 1.38 1.13 1.01 1.06 1.37 1.13 0.98 1.03 1.35 1.10 0.95 1.01
OH_MI6 1.07 1.15 0.99 1.03 1.04 1.16 0.98 1.03 1.02 1.15 0.98 1.02
OH_MI7 1.19 1.14 1.06 1.08 1.18 1.13 1.05 1.07 1.15 1.12 1.03 1.05
OH_MI8 1.15 1.12 0.93 0.98 1.16 1.11 0.92 0.97 1.13 1.11 0.90 0.96
Average over substates 1.13 1.08 0.97 1.01 1.13 1.08 0.96 0.99 1.11 1.07 0.95 0.99
Relative Absolute Bias 5.25 0.20 5.84 3.71 5.52 0.08 4.39 2.64 6.52 0.69 3.47 1.77
PA_IL (design-based) 0.99 1.07 1.08 1.08 0.99 1.07 1.08 1.08 0.99 1.07 1.08 1.08
PA_IL1 1.07 1.06 0.99 1.01 1.04 1.06 1.00 1.02 1.05 1.06 1.01 1.02
PA_IL2 1.00 1.05 1.07 1.06 0.98 1.05 1.08 1.07 0.97 1.05 1.09 1.07
PA_IL3 1.07 1.10 1.23 1.19 1.05 1.09 1.22 1.18 1.01 1.08 1.19 1.16
PA_IL4 1.03 1.05 1.04 1.05 1.00 1.05 1.04 1.04 1.01 1.04 1.02 1.03
PA_IL5 1.06 1.05 1.06 1.06 1.05 1.04 1.06 1.06 1.04 1.03 1.06 1.06
PA_IL6 1.20 1.06 0.94 0.98 1.18 1.07 0.93 0.97 1.22 1.06 0.93 0.98
PA_IL7 0.87 0.97 0.91 0.93 0.88 0.98 0.93 0.94 0.89 1.00 0.96 0.96
PA_IL8 1.11 1.01 1.09 1.07 1.10 0.99 1.10 1.08 1.08 0.99 1.08 1.06
Average over substates 1.05 1.04 1.04 1.04 1.03 1.04 1.05 1.05 1.03 1.04 1.04 1.04
Relative Absolute Bias 6.66 2.60 3.71 3.04 4.89 2.78 3.41 2.94 4.78 3.00 3.63 3.16
Overall 5.74 1.14 3.79 2.55 5.25 1.17 3.58 2.45 5.15 1.30 3.60 2.46

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.12 Relative Absolute Bias of Three Types of Change Estimates for 2000/1999: Past Month Use of Cigarettes
State R1 R2 R3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
CA_TX (design-based) 0.95 0.89 1.11 1.05 0.95 0.89 1.11 1.05 0.95 0.89 1.11 1.05
CA_TX1 1.05 0.85 1.08 1.03 1.07 0.84 1.08 1.02 1.08 0.83 1.07 1.02
CA_TX2 0.94 0.75 1.04 0.97 0.97 0.74 1.04 0.97 0.98 0.74 1.05 0.98
CA_TX3 0.89 1.00 1.00 1.00 0.88 1.01 0.98 0.98 0.87 1.01 0.98 0.98
CA_TX4 0.91 0.89 1.01 0.98 0.92 0.89 1.00 0.98 0.92 0.89 1.00 0.97
CA_TX5 0.84 0.92 1.13 1.07 0.84 0.92 1.15 1.08 0.84 0.93 1.15 1.08
CA_TX6 0.78 0.88 1.03 0.98 0.78 0.89 1.02 0.98 0.78 0.90 1.02 0.98
CA_TX7 1.01 0.97 1.15 1.11 0.99 0.98 1.16 1.11 1.00 0.97 1.15 1.10
CA_TX8 0.88 0.91 1.17 1.10 0.86 0.91 1.18 1.10 0.85 0.91 1.17 1.09
Average over substates 0.91 0.90 1.08 1.03 0.91 0.90 1.08 1.03 0.91 0.90 1.07 1.03
Relative Absolute Bias 4.09 0.37 2.75 2.34 4.25 0.39 2.73 2.35 3.97 0.42 2.98 2.53
NY_FL (design-based) 1.00 0.97 0.91 0.92 1.00 0.97 0.91 0.92 1.00 0.97 0.91 0.92
NY_FL1 0.88 0.99 1.01 1.00 0.85 0.99 1.01 1.00 0.86 0.98 1.01 1.00
NY_FL2 0.82 0.89 0.88 0.88 0.83 0.89 0.88 0.88 0.83 0.90 0.89 0.89
NY_FL3 0.95 0.93 0.97 0.96 0.95 0.92 0.97 0.96 0.96 0.92 0.96 0.96
NY_FL4 0.85 0.92 0.81 0.83 0.85 0.93 0.81 0.83 0.87 0.95 0.80 0.82
NY_FL5 0.92 0.98 0.90 0.92 0.93 0.97 0.91 0.92 0.93 0.98 0.90 0.92
NY_FL6 1.01 0.89 1.03 1.00 1.01 0.89 1.04 1.01 1.02 0.88 1.03 1.01
NY_FL7 0.82 0.87 0.94 0.92 0.82 0.88 0.94 0.93 0.83 0.89 0.95 0.94
NY_FL8 1.06 1.03 0.83 0.87 1.05 1.03 0.82 0.86 1.08 1.03 0.81 0.86
Average over substates 0.92 0.94 0.92 0.92 0.91 0.94 0.92 0.92 0.92 0.94 0.92 0.92
Relative Absolute Bias 8.29 2.82 1.18 0.07 8.83 2.87 1.30 0.02 7.63 2.45 0.95 0.18
OH_MI (design-based) 0.88 0.94 0.91 0.91 0.88 0.94 0.91 0.91 0.88 0.94 0.91 0.91
OH_MI1 0.77 0.93 0.85 0.86 0.78 0.93 0.85 0.86 0.79 0.94 0.85 0.87
OH_MI2 0.95 1.02 0.88 0.91 0.97 1.00 0.88 0.91 0.96 1.01 0.88 0.91
OH_MI3 0.90 0.92 0.81 0.84 0.92 0.92 0.80 0.83 0.94 0.91 0.80 0.83
OH_MI4 0.81 1.03 0.95 0.96 0.81 1.03 0.96 0.96 0.82 1.03 0.96 0.97
OH_MI5 1.05 1.04 0.93 0.96 1.04 1.04 0.92 0.95 1.05 1.02 0.91 0.94
OH_MI6 0.76 0.99 0.99 0.98 0.73 1.00 1.00 0.98 0.74 0.99 1.01 0.99
OH_MI7 0.82 0.88 0.91 0.90 0.84 0.87 0.92 0.91 0.84 0.88 0.93 0.91
OH_MI8 1.00 1.00 0.90 0.93 1.03 0.99 0.90 0.93 1.03 0.99 0.90 0.92
Average over substates 0.88 0.97 0.90 0.92 0.89 0.97 0.90 0.92 0.90 0.97 0.91 0.92
Relative Absolute Bias 0.50 3.39 0.22 0.48 1.28 3.19 0.21 0.50 2.00 3.09 0.02 0.68
PA_IL (design-based) 0.78 1.08 0.95 0.97 0.78 1.08 0.95 0.97 0.78 1.08 0.95 0.97
PA_IL1 0.90 1.06 0.91 0.93 0.89 1.05 0.91 0.94 0.90 1.06 0.90 0.93
PA_IL2 0.92 0.93 0.95 0.95 0.92 0.92 0.96 0.95 0.93 0.92 0.96 0.95
PA_IL3 0.82 1.07 0.91 0.94 0.81 1.07 0.90 0.93 0.81 1.06 0.90 0.93
PA_IL4 0.77 1.14 0.90 0.94 0.75 1.14 0.90 0.94 0.74 1.16 0.90 0.93
PA_IL5 0.68 1.06 1.06 1.03 0.65 1.07 1.06 1.03 0.64 1.06 1.08 1.05
PA_IL6 0.95 1.06 0.92 0.95 0.96 1.06 0.91 0.94 0.94 1.05 0.91 0.94
PA_IL7 0.99 1.06 0.97 0.99 1.00 1.06 0.97 0.99 0.99 1.06 0.97 0.99
PA_IL8 0.80 1.01 0.82 0.85 0.81 1.01 0.82 0.85 0.81 1.01 0.82 0.85
Average over substates 0.85 1.05 0.93 0.95 0.85 1.05 0.93 0.95 0.84 1.05 0.93 0.95
Relative Absolute Bias 9.48 2.58 2.51 2.02 8.98 2.66 2.71 2.21 8.37 2.59 2.64 2.23
Overall 5.59 2.29 1.66 1.23 5.83 2.28 1.74 1.27 5.49 2.14 1.65 1.40

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.13 Change Estimates (R1) for 2000/1999
State Age in Years Total
12-17 18-25 26+
Past Month Use of Marijuana
CA_TX (design-based) 0.99 0.92 1.06 1.00
CA_TX (average over substates) 1.01 0.92 1.14 1.04
NY_FL (design-based) 1.22 0.91 0.78 0.88
NY_FL (average over substates) 1.05 0.91 0.96 0.95
OH_MI (design-based) 1.02 1.08 1.09 1.08
OH_MI (average over substates) 1.04 1.06 1.07 1.06
PA_IL (design-based) 0.87 1.02 1.20 1.07
PA_IL (average over substates) 0.96 0.98 1.10 1.03
Past Year Use of Cocaine
CA_TX (design-based) 0.86 0.75 0.80 0.79
CA_TX (average over substates) 1.02 0.76 0.81 0.82
NY_FL (design-based) 0.82 0.77 0.96 0.88
NY_FL (average over substates) 0.92 0.83 0.67 0.74
OH_MI (design-based) 1.15 0.72 0.77 0.77
OH_MI (average over substates) 1.36 0.83 0.81 0.85
PA_IL (design-based) 0.84 0.99 1.22 1.10
PA_IL (average over substates) 1.07 0.94 1.32 1.15
Past Month "Binge" Alcohol Use
CA_TX (design-based) 0.95 0.93 1.05 1.02
CA_TX (average over substates) 1.00 0.94 1.05 1.02
NY_FL (design-based) 1.17 1.00 0.94 0.96
NY_FL (average over substates) 1.10 0.99 0.99 0.99
OH_MI (design-based) 1.19 1.08 0.92 0.97
OH_MI (average over substates) 1.13 1.08 0.97 1.01
PA_IL (design-based) 0.99 1.07 1.08 1.08
PA_IL (average over substates) 1.05 1.04 1.04 1.04
Past Month Use of Cigarettes
CA_TX (design-based) 0.95 0.89 1.11 1.05
CA_TX (average over substates) 0.91 0.90 1.08 1.03
NY_FL (design-based) 1.00 0.97 0.91 0.92
NY_FL (average over substates) 0.92 0.94 0.92 0.92
OH_MI (design-based) 0.88 0.94 0.91 0.91
OH_MI (average over substates) 0.88 0.97 0.90 0.92
PA_IL (design-based) 0.78 1.08 0.95 0.97
PA_IL (average over substates) 0.85 1.05 0.93 0.95

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.14 Relative Absolute Bias of Three Types of Change Estimates for 2000/1999
State R1 R2 R3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
Past Month Use of Marijuana
CA_TX 1.27 0.47 7.01 3.72 1.50 1.07 6.27 3.58 1.16 0.03 4.23 2.18
NY_FL 14.24 0.19 23.35 7.73 14.11 0.25 25.80 8.99 13.10 0.76 27.95 10.34
OH_MI 1.24 1.83 1.11 1.43 1.22 2.61 4.78 3.74 2.78 4.26 6.87 5.52
PA_IL 10.45 4.08 8.70 4.45 9.91 4.22 8.43 4.43 9.22 4.49 11.13 5.80
Average 6.80 1.64 10.04 4.33 6.68 2.04 11.32 5.18 6.57 2.39 12.55 5.96
Past Year Use of Cocaine
CA_TX 18.13 0.70 2.23 3.33 17.63 0.74 3.86 4.10 17.66 0.47 3.78 4.08
NY_FL 11.60 7.11 29.72 16.12 9.25 7.26 29.32 16.02 9.50 6.75 30.05 16.72
OH_MI 17.73 15.76 5.15 11.09 16.45 16.49 4.79 11.04 17.19 16.40 5.97 11.77
PA_IL 28.03 5.70 8.09 3.77 26.59 5.97 10.21 4.76 25.65 5.82 9.80 4.54
Average 18.87 7.32 11.30 8.58 17.48 7.61 12.04 8.98 17.50 7.36 12.40 9.28
Past Month "Binge" Alcohol Use
CA_TX 4.87 0.55 0.32 0.41 4.77 0.80 0.83 0.84 5.26 0.89 1.21 1.11
NY_FL 6.17 1.21 5.30 3.05 5.81 1.02 5.70 3.38 4.06 0.63 6.09 3.81
OH_MI 5.25 0.20 5.84 3.71 5.52 0.08 4.39 2.64 6.52 0.69 3.47 1.77
PA_IL 6.66 2.60 3.71 3.03 4.89 2.78 3.41 2.94 4.77 3.00 3.63 3.16
Average 5.74 1.14 3.79 2.55 5.25 1.17 3.58 2.45 5.15 1.30 3.60 2.46
Past Month Use of Cigarettes
CA_TX 4.09 0.37 2.75 2.34 4.25 0.39 2.73 2.35 3.97 0.42 2.98 2.53
NY_FL 8.29 2.82 1.18 0.07 8.83 2.87 1.30 0.02 7.63 2.45 0.95 0.18
OH_MI 0.50 3.39 0.22 0.48 1.28 3.19 0.21 0.50 2.00 3.08 0.02 0.68
PA_IL 9.48 2.58 2.51 2.02 8.98 2.66 2.71 2.21 8.37 2.59 2.64 2.23
Average 5.59 2.29 1.66 1.23 5.83 2.28 1.74 1.27 5.49 2.14 1.65 1.40
Overall 9.25 3.10 6.70 4.17 8.81 3.27 7.17 4.47 8.68 3.30 7.55 4.78

Note: Relative absolute bias (Ri) = 100 × abs(mean of 8 substate Ri - R) / R, i = 1, 2, and 3, where R = ratio of 2000 design-based estimates to 1999 design-based estimates.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.15 Average 95 Percent Lower and Upper Bounds for Change Estimates (R1) for 2000/1999
State Age in Years Total
12-17 18-25 26+
Lower Upper Lower Upper Lower Upper Lower Upper
Past Month Use of Marijuana
CA_TX (design-based) 0.83 1.19 0.75 1.12 0.73 1.54 0.82 1.22
CA_TX (avg. over substates) 0.68 1.49 0.64 1.32 0.62 2.08 0.73 1.47
NY_FL (design-based) 0.95 1.57 0.77 1.09 0.52 1.15 0.74 1.06
NY_FL (avg. over substates) 0.66 1.66 0.64 1.29 0.53 1.73 0.68 1.33
OH_MI (design-based) 0.83 1.27 0.89 1.30 0.74 1.60 0.87 1.32
OH_MI (avg. over substates) 0.67 1.61 0.76 1.47 0.61 1.90 0.78 1.44
PA_IL (design-based) 0.69 1.09 0.83 1.25 0.81 1.79 0.88 1.31
PA_IL (avg. over substates) 0.62 1.48 0.70 1.37 0.62 1.95 0.75 1.40
Past Year Use of Cocaine
CA_TX (design-based) 0.61 1.21 0.57 0.99 0.38 1.67 0.56 1.11
CA_TX (avg. over substates) 0.56 1.84 0.46 1.26 0.34 1.93 0.50 1.33
NY_FL (design-based) 0.41 1.63 0.58 1.03 0.53 1.71 0.61 1.26
NY_FL (avg. over substates) 0.44 1.93 0.48 1.44 0.29 1.54 0.42 1.28
OH_MI (design-based) 0.56 2.37 0.50 1.03 0.39 1.52 0.52 1.13
OH_MI (avg. over substates) 0.62 2.97 0.48 1.44 0.34 1.92 0.51 1.43
PA_IL (design-based) 0.48 1.46 0.70 1.42 0.65 2.31 0.75 1.63
PA_IL (avg. over substates) 0.50 2.28 0.53 1.64 0.58 3.01 0.68 1.95
Past Month "Binge" Alcohol Use
CA_TX (design-based) 0.80 1.12 0.85 1.02 0.90 1.23 0.90 1.14
CA_TX (avg. over substates) 0.72 1.37 0.77 1.13 0.79 1.39 0.83 1.26
NY_FL (design-based) 0.94 1.47 0.91 1.10 0.78 1.12 0.84 1.10
NY_FL (avg. over substates) 0.76 1.60 0.81 1.19 0.74 1.31 0.79 1.23
OH_MI (design-based) 1.01 1.41 0.99 1.18 0.81 1.04 0.88 1.06
OH_MI (avg. over substates) 0.80 1.60 0.91 1.29 0.75 1.26 0.83 1.22
PA_IL (design-based) 0.80 1.22 0.98 1.17 0.95 1.23 0.97 1.19
PA_IL (avg. over substates) 0.75 1.48 0.89 1.23 0.81 1.35 0.86 1.27
Past Month Use of Cigarettes
CA_TX (design-based) 0.82 1.10 0.79 1.01 0.96 1.28 0.94 1.18
CA_TX (avg. over substates) 0.68 1.23 0.74 1.09 0.83 1.39 0.84 1.26
NY_FL (design-based) 0.81 1.23 0.88 1.06 0.81 1.03 0.84 1.02
NY_FL (avg. over substates) 0.66 1.27 0.77 1.14 0.72 1.18 0.76 1.13
OH_MI (design-based) 0.76 1.02 0.86 1.03 0.80 1.02 0.83 1.00
OH_MI (avg. over substates) 0.66 1.18 0.83 1.14 0.72 1.13 0.77 1.09
PA_IL (design-based) 0.67 0.90 1.00 1.16 0.83 1.09 0.87 1.07
PA_IL (avg. over substates) 0.64 1.14 0.89 1.24 0.73 1.18 0.79 1.14

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.16 Ratio of Widths of 95 Percent Confidence Intervals of Change Estimates for 2000/1999
State W1 W2 W3
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
Past Month Use of Marijuana
CA_TX 0.70 0.65 0.61 0.58 0.71 0.66 0.60 0.56 0.71 0.66 0.56 0.52
NY_FL 0.45 0.75 0.43 0.58 0.46 0.77 0.43 0.58 0.47 0.78 0.42 0.55
OH_MI 0.61 0.63 0.46 0.56 0.61 0.64 0.44 0.53 0.60 0.63 0.41 0.49
PA_IL 0.62 0.62 0.51 0.52 0.64 0.64 0.51 0.50 0.64 0.64 0.47 0.47
Average 0.59 0.66 0.50 0.56 0.60 0.68 0.50 0.54 0.61 0.68 0.47 0.51
Past Year Use of Cocaine
CA_TX 0.64 0.63 0.88 0.47 0.66 0.64 0.84 0.45 0.68 0.64 0.78 0.43
NY_FL 0.36 0.63 0.31 0.51 0.38 0.66 0.30 0.48 0.39 0.68 0.29 0.47
OH_MI 0.74 0.66 0.81 0.45 0.78 0.69 0.78 0.44 0.82 0.70 0.74 0.42
PA_IL 0.57 0.38 0.40 0.50 0.61 0.39 0.39 0.48 0.64 0.41 0.37 0.47
Average 0.58 0.57 0.60 0.48 0.61 0.60 0.57 0.46 0.63 0.61 0.54 0.45
Past Month "Binge" Alcohol Use
CA_TX 0.70 0.75 0.63 0.65 0.73 0.77 0.65 0.67 0.74 0.79 0.66 0.66
NY_FL 0.54 0.68 0.63 0.64 0.57 0.69 0.64 0.65 0.59 0.70 0.65 0.65
OH_MI 0.57 0.69 0.74 0.73 0.59 0.70 0.76 0.74 0.60 0.70 0.75 0.72
PA_IL 0.63 0.65 0.64 0.64 0.64 0.66 0.66 0.65 0.66 0.67 0.66 0.64
Average 0.61 0.69 0.66 0.67 0.63 0.71 0.68 0.68 0.65 0.72 0.68 0.67
Past Month Use of Cigarettes
CA_TX 0.64 0.75 0.60 0.61 0.66 0.77 0.63 0.64 0.68 0.79 0.66 0.66
NY_FL 0.53 0.66 0.68 0.67 0.56 0.68 0.71 0.70 0.59 0.71 0.74 0.72
OH_MI 0.70 0.74 0.68 0.68 0.73 0.76 0.71 0.71 0.77 0.79 0.74 0.73
PA_IL 0.72 0.70 0.64 0.65 0.75 0.72 0.66 0.67 0.76 0.74 0.69 0.68
Average 0.65 0.71 0.65 0.65 0.68 0.73 0.68 0.68 0.70 0.76 0.71 0.70
Overall 0.61 0.66 0.60 0.59 0.63 0.68 0.61 0.59 0.65 0.69 0.60 0.58

W1 = WR1 / WR, W2 = WR2 / WR, and W3 = WR3 / WR

where

WR1 = mean of widths of PIs of R1 over substates,
WR2 = mean of widths of PIs of R2 over substates,
WR3 = mean of widths of PIs of R3 over substates, and
WR = mean of widths of CIs of design-based estimate over substates.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.17 95 Percent Least Significant Lower and Upper Bounds of Change Estimates (R1) for 2000/1999
State 12-17 18-25 26+ Total
Lower Upper Lower Upper Lower Upper Lower Upper
Past Month Use of Marijuana
CA_TX 0.67 1.49 0.70 1.43 0.55 1.83 0.71 1.42
NY_FL 0.63 1.59 0.71 1.42 0.55 1.81 0.71 1.40
OH_MI 0.64 1.56 0.72 1.39 0.56 1.78 0.74 1.36
PA_IL 0.65 1.55 0.72 1.40 0.56 1.78 0.73 1.37
Average 0.65 1.54 0.71 1.41 0.56 1.80 0.72 1.39
Past Year Use of Cocaine
CA_TX 0.55 1.81 0.60 1.66 0.42 2.36 0.61 1.63
NY_FL 0.47 2.11 0.57 1.74 0.43 2.31 0.57 1.74
OH_MI 0.46 2.19 0.58 1.73 0.42 2.37 0.59 1.68
PA_IL 0.47 2.14 0.57 1.76 0.44 2.29 0.59 1.70
Average 0.49 2.06 0.58 1.72 0.43 2.33 0.59 1.69
Past Month "Binge" Alcohol Use
CA_TX 0.73 1.38 0.83 1.21 0.76 1.32 0.81 1.23
NY_FL 0.69 1.45 0.82 1.21 0.75 1.33 0.80 1.25
OH_MI 0.71 1.42 0.84 1.19 0.77 1.30 0.82 1.21
PA_IL 0.71 1.40 0.85 1.18 0.77 1.29 0.82 1.21
Average 0.71 1.41 0.84 1.20 0.76 1.31 0.82 1.23
Past Month Use of Cigarettes
CA_TX 0.74 1.35 0.82 1.22 0.77 1.30 0.82 1.22
NY_FL 0.72 1.39 0.82 1.22 0.78 1.28 0.82 1.22
OH_MI 0.75 1.34 0.85 1.17 0.80 1.26 0.84 1.19
PA_IL 0.75 1.34 0.85 1.18 0.79 1.27 0.83 1.20
Average 0.74 1.36 0.84 1.20 0.78 1.28 0.83 1.21

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.18 Past Month Use of Marijuana Based on Pooled 1999 and 2000 Data
State Design-Based Estimates SAE with Region SAE Without Region
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
National 7.20 13.92 2.88 4.76 7.25 13.69 2.98 4.81 7.24 13.69 2.97 4.80
Northeast 7.57 17.74 3.05 5.24 7.69 17.50 3.11 5.26 7.87 16.77 3.23 5.29
Midwest 7.40 14.11 2.46 4.52 7.44 13.95 2.50 4.53 7.38 14.06 2.56 4.59
South 6.32 11.76 2.31 3.95 6.40 11.72 2.43 4.05 6.38 11.78 2.56 4.15
West 8.08 14.13 4.13 5.92 8.03 13.52 4.30 5.96 7.95 13.88 3.85 5.65
Alabama 7.02 11.12 1.43 3.26 6.49 11.24 2.19 3.80 6.40 11.18 2.27 3.84
Alaska 8.73 18.84 4.02 6.77 8.74 16.82 4.37 6.73 8.65 17.35 3.76 6.35
Arizona 8.00 10.09 3.46 4.85 7.47 11.05 3.48 4.93 7.48 11.40 2.89 4.53
Arkansas 7.96 11.06 1.88 3.71 7.10 9.93 2.30 3.78 7.09 10.00 2.43 3.89
California 7.60 13.94 4.16 5.86 7.56 13.16 4.35 5.90 7.50 13.49 4.02 5.69
Colorado 11.75 22.47 5.30 8.24 10.97 19.91 5.73 8.14 10.80 20.49 5.21 7.80
Connecticut 10.85 17.10 3.16 5.47 9.79 18.38 3.35 5.66 9.99 17.05 3.57 5.70
Delaware 13.50 21.05 3.51 6.70 12.00 20.83 4.07 6.93 11.89 20.81 4.48 7.24
District of Columbia 7.21 11.83 2.89 4.53 7.24 12.40 3.74 5.25 7.25 12.48 3.69 5.23
Florida 6.33 13.31 3.39 4.73 6.68 13.05 3.29 4.65 6.63 13.11 3.40 4.74
Georgia 5.44 11.22 3.79 4.97 5.85 11.85 2.68 4.25 5.89 12.01 2.91 4.45
Hawaii 10.07 14.67 5.78 7.26 8.86 13.65 4.86 6.30 8.72 14.35 4.48 6.07
Idaho 5.01 10.33 2.85 4.25 6.00 10.83 2.96 4.52 5.99 11.12 2.36 4.12
Illinois 8.57 14.31 2.51 4.70 8.18 14.18 2.62 4.73 8.13 14.25 2.67 4.77
Indiana 7.70 11.87 3.11 4.76 7.50 12.09 2.54 4.33 7.48 12.26 2.65 4.43
Iowa 3.71 9.27 1.24 2.57 4.88 9.29 1.40 2.81 4.94 9.43 1.34 2.79
Kansas 6.93 12.18 1.38 3.46 6.85 12.03 1.77 3.71 6.82 12.00 1.74 3.68
Kentucky 6.50 13.30 1.74 3.76 6.80 12.55 2.26 4.07 6.79 12.71 2.41 4.21
Louisiana 6.53 11.00 1.63 3.57 6.10 11.17 1.76 3.64 6.07 11.19 1.89 3.74
Maine 8.61 21.37 3.25 5.88 9.01 21.40 3.27 5.93 9.25 20.31 3.42 5.95
Maryland 9.23 12.61 1.59 3.64 8.39 13.13 2.54 4.37 8.25 13.23 2.66 4.46
Massachusetts 13.30 26.58 6.96 9.90 12.20 26.01 5.79 8.77 12.35 24.81 6.28 9.03
Michigan 7.77 16.64 3.53 5.68 8.01 16.54 3.34 5.55 7.94 16.61 3.49 5.66
Minnesota 10.72 14.62 1.71 4.44 9.24 14.97 2.25 4.71 9.13 15.08 2.27 4.73
Mississippi 5.17 11.63 0.49 2.62 5.28 10.63 1.33 3.11 5.28 10.67 1.40 3.16
Missouri 5.84 14.28 1.92 3.95 6.62 13.40 2.38 4.26 6.55 13.62 2.45 4.33
Montana 10.51 15.71 2.41 5.01 9.42 14.80 3.10 5.30 9.26 15.05 2.55 4.89
Nebraska 5.64 11.42 1.19 3.09 6.16 11.55 1.65 3.50 6.17 11.54 1.57 3.45
Nevada 10.05 14.58 3.65 5.55 9.72 13.21 4.15 5.75 9.54 13.73 3.54 5.31
New Hampshire 11.20 18.56 2.87 5.59 10.54 19.95 3.26 5.98 10.73 18.74 3.39 5.96
New Jersey 6.01 16.79 2.37 4.39 6.63 15.86 2.81 4.69 6.81 14.88 2.86 4.63
New Mexico 10.52 15.95 3.55 6.12 9.74 14.86 4.13 6.30 9.52 15.15 3.60 5.93
New York 6.32 16.77 2.02 4.26 6.64 16.26 2.38 4.51 6.83 15.79 2.42 4.50
North Carolina 5.76 15.34 4.16 5.69 6.53 14.29 3.71 5.29 6.51 14.35 3.98 5.50
North Dakota 7.41 10.20 0.57 2.75 6.80 10.27 1.38 3.27 6.80 10.16 1.27 3.17
Ohio 6.07 14.31 2.49 4.40 6.46 13.66 2.38 4.26 6.44 13.77 2.41 4.30
Oklahoma 5.13 7.69 1.36 2.63 5.58 8.39 1.64 2.98 5.66 8.39 1.68 3.02
Oregon 8.83 17.37 4.17 6.28 9.48 17.16 4.95 6.90 9.39 17.68 4.40 6.53
Pennsylvania 5.83 14.16 2.79 4.42 6.28 14.43 2.74 4.45 6.46 13.96 2.81 4.47
Rhode Island 9.17 23.20 4.77 7.32 10.20 22.71 4.33 7.00 10.35 21.79 4.70 7.20
South Carolina 6.41 12.03 1.98 3.68 6.57 12.46 2.21 3.93 6.54 12.49 2.33 4.02
South Dakota 7.04 10.14 1.91 3.69 6.49 11.37 1.84 3.72 6.50 11.37 1.85 3.73
Tennessee 6.49 11.13 2.41 3.94 6.61 11.14 2.72 4.17 6.57 11.23 2.88 4.31
Texas 6.00 10.41 1.34 3.22 5.91 10.32 1.51 3.32 5.92 10.34 1.56 3.36
Utah 4.02 5.77 2.01 3.01 4.85 7.34 2.27 3.59 4.85 7.44 1.58 3.15
Vermont 11.23 26.28 3.99 7.48 10.50 25.65 3.79 7.18 10.62 24.60 4.05 7.26
Virginia 5.06 11.54 2.25 3.67 5.83 12.31 2.43 3.98 5.82 12.50 2.60 4.13
Washington 8.40 14.47 5.05 6.58 8.83 13.96 4.50 6.14 8.70 14.28 3.80 5.62
West Virginia 8.32 10.86 0.72 2.70 7.39 10.57 1.89 3.48 7.32 10.67 1.90 3.49
Wisconsin 8.05 16.30 2.69 5.10 8.20 16.31 2.70 5.09 8.08 16.60 2.86 5.24
Wyoming 7.28 13.08 2.50 4.58 7.51 12.69 2.91 4.85 7.42 12.99 2.26 4.40

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.19 Past Year Use of Cocaine Based on Pooled 1999 and 2000 Data
State Design-Based Estimates SAE with Region SAE Without Region
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
National 1.63 4.82 1.04 1.59 1.67 4.73 1.13 1.65 1.67 4.73 1.12 1.64
Northeast 1.16 4.56 1.16 1.56 1.18 4.55 1.27 1.65 1.43 4.63 1.09 1.54
Midwest 1.16 4.65 0.80 1.35 1.18 4.48 0.86 1.37 1.42 4.50 1.00 1.50
South 1.75 4.75 1.04 1.60 1.80 4.72 1.13 1.66 1.66 4.74 1.19 1.70
West 2.32 5.28 1.18 1.86 2.35 5.14 1.28 1.92 2.13 5.03 1.18 1.80
Alabama 2.36 3.88 1.12 1.60 1.71 4.13 1.28 1.69 1.54 4.17 1.38 1.76
Alaska 0.92 5.39 2.38 2.63 1.93 5.36 1.86 2.37 1.62 5.17 1.71 2.20
Arizona 4.05 6.70 1.36 2.37 3.41 6.12 1.43 2.27 3.15 5.98 1.27 2.11
Arkansas 1.26 3.63 1.07 1.42 1.62 3.70 1.12 1.50 1.44 3.72 1.15 1.51
California 2.05 4.79 1.29 1.85 2.14 4.73 1.26 1.83 2.00 4.67 1.16 1.73
Colorado 2.93 10.88 1.31 2.74 2.74 8.98 1.54 2.64 2.42 8.73 1.38 2.45
Connecticut 1.32 2.94 1.19 1.40 1.22 4.27 1.31 1.64 1.55 4.43 1.08 1.51
Delaware 1.79 5.75 3.98 3.98 1.75 6.05 2.08 2.53 1.62 6.10 2.02 2.48
District of Columbia 1.50 4.30 2.10 2.34 1.01 4.14 1.87 2.09 0.89 4.12 1.88 2.08
Florida 1.52 5.96 1.18 1.73 1.70 5.32 1.14 1.64 1.58 5.31 1.22 1.69
Georgia 1.13 5.29 1.40 1.90 1.33 4.82 1.25 1.74 1.20 4.80 1.33 1.79
Hawaii 1.20 4.93 1.69 2.03 1.86 3.93 1.72 1.99 1.61 3.72 1.54 1.81
Idaho 2.06 2.54 0.36 0.90 2.20 3.56 0.84 1.42 1.86 3.51 0.79 1.33
Illinois 0.96 3.93 1.11 1.47 1.08 4.12 0.97 1.40 1.28 4.13 1.08 1.51
Indiana 1.66 4.88 0.23 1.00 1.34 4.55 0.67 1.26 1.64 4.61 0.85 1.43
Iowa 1.00 3.79 0.59 1.06 1.19 4.05 0.72 1.21 1.44 4.07 0.83 1.33
Kansas 0.16 4.26 0.63 1.07 1.01 4.33 0.82 1.31 1.22 4.41 0.99 1.48
Kentucky 0.99 5.27 0.80 1.41 1.63 5.02 1.11 1.68 1.45 4.99 1.12 1.66
Louisiana 1.21 3.93 1.16 1.57 1.31 4.02 1.29 1.69 1.17 4.05 1.34 1.71
Maine 2.21 3.41 0.69 1.16 1.55 3.90 0.97 1.37 1.92 4.06 0.83 1.31
Maryland 1.79 3.09 0.61 1.02 1.51 3.82 0.88 1.29 1.35 3.80 0.90 1.29
Massachusetts 1.10 6.89 2.52 2.90 1.28 6.32 1.77 2.25 1.63 6.44 1.46 2.06
Michigan 1.02 5.04 0.86 1.42 1.09 4.69 0.91 1.42 1.27 4.68 1.04 1.54
Minnesota 2.39 4.82 1.03 1.69 1.63 4.90 0.84 1.47 2.02 4.88 1.01 1.64
Mississippi 1.07 3.82 0.44 1.00 1.21 3.84 1.02 1.44 1.06 3.84 1.11 1.49
Missouri 0.75 3.57 0.45 0.89 1.03 3.79 0.75 1.17 1.27 3.84 0.87 1.30
Montana 3.03 5.32 0.40 1.32 2.64 5.13 0.90 1.63 2.28 4.95 0.86 1.54
Nebraska 0.86 4.32 0.52 1.08 1.18 4.42 0.74 1.29 1.46 4.42 0.88 1.43
Nevada 2.61 6.72 1.14 1.94 2.47 5.81 1.72 2.27 2.10 5.53 1.64 2.14
New Hampshire 1.92 5.39 0.39 1.14 1.46 4.83 0.90 1.42 1.82 4.99 0.80 1.40
New Jersey 0.76 4.77 0.91 1.35 1.03 4.58 1.21 1.59 1.27 4.71 1.01 1.47
New Mexico 3.22 8.49 2.01 3.07 3.49 7.52 1.95 2.92 3.19 7.33 1.84 2.77
New York 1.18 3.87 1.01 1.37 1.16 4.04 1.27 1.60 1.38 4.10 1.10 1.49
North Carolina 1.65 3.77 1.12 1.50 1.57 4.15 1.19 1.59 1.41 4.19 1.20 1.59
North Dakota 1.42 3.89 0.20 0.87 1.34 3.88 0.64 1.18 1.62 3.91 0.77 1.31
Ohio 0.78 4.98 0.92 1.43 0.99 4.54 0.93 1.40 1.16 4.53 1.05 1.51
Oklahoma 1.05 3.21 0.70 1.07 1.60 3.79 1.03 1.46 1.39 3.77 1.08 1.47
Oregon 1.98 4.70 0.69 1.32 2.14 4.82 1.08 1.65 1.83 4.65 0.97 1.51
Pennsylvania 1.18 4.39 1.00 1.41 1.16 4.31 1.13 1.50 1.37 4.37 1.00 1.43
Rhode Island 0.72 7.25 0.84 1.56 1.16 5.83 1.29 1.79 1.47 5.96 1.11 1.69
South Carolina 1.52 3.86 0.96 1.38 1.45 4.17 1.15 1.55 1.31 4.20 1.26 1.63
South Dakota 2.35 5.73 0.26 1.28 1.79 4.62 0.72 1.39 2.20 4.63 0.87 1.56
Tennessee 2.08 2.64 2.18 2.23 1.84 3.84 1.50 1.83 1.67 3.89 1.49 1.81
Texas 2.66 6.10 0.83 1.82 2.64 5.69 1.01 1.89 2.56 5.72 1.10 1.95
Utah 1.62 2.31 0.44 0.97 2.01 3.54 0.91 1.56 1.68 3.42 0.90 1.48
Vermont 1.49 7.62 1.16 1.98 1.40 6.21 1.22 1.84 1.77 6.36 1.05 1.77
Virginia 1.29 4.70 0.41 1.02 1.46 4.76 0.96 1.47 1.30 4.77 1.00 1.49
Washington 2.46 4.04 0.82 1.39 2.34 4.49 1.05 1.62 2.02 4.35 0.95 1.49
West Virginia 1.60 3.47 0.52 0.99 1.84 3.78 0.84 1.30 1.62 3.81 0.86 1.30
Wisconsin 1.78 6.19 0.99 1.77 1.49 5.42 0.92 1.58 1.82 5.50 1.06 1.73
Wyoming 1.58 4.78 0.61 1.32 2.01 4.89 0.96 1.65 1.68 4.68 0.89 1.53

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.20 Past Month Use of Alcohol Based on Pooled 1999 and 2000 Data
State Design-Based Estimates SAE with Region SAE Without Region
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
National 16.45 57.04 48.88 46.53 16.40 56.82 48.55 46.26 16.40 56.81 48.55 46.25
Northeast 18.08 62.76 54.04 51.56 18.12 63.08 54.14 51.69 17.90 62.40 54.28 51.70
Midwest 17.25 62.68 51.70 49.51 17.18 62.39 50.54 48.59 17.34 62.27 50.33 48.44
South 15.47 53.06 42.45 40.99 15.37 52.84 42.16 40.73 15.25 52.91 42.82 41.23
West 15.88 52.91 51.73 47.95 15.85 52.38 51.84 47.96 16.05 52.86 50.82 47.28
Alabama 14.94 51.25 37.42 36.93 14.60 50.18 36.23 35.84 14.46 50.18 37.59 36.87
Alaska 16.41 56.82 59.59 53.66 16.35 57.28 58.91 53.23 16.57 58.13 58.22 52.87
Arizona 17.09 53.66 54.13 49.90 16.77 54.42 49.92 46.78 17.06 54.95 49.22 46.36
Arkansas 19.11 46.94 36.37 35.93 17.69 46.38 35.12 34.75 17.34 46.23 35.92 35.32
California 15.50 51.77 50.36 46.71 15.50 51.06 50.86 46.99 15.56 51.28 50.15 46.49
Colorado 19.90 70.64 63.21 59.61 19.76 66.83 63.76 59.53 20.08 68.16 62.15 58.51
Connecticut 22.03 67.20 57.74 55.43 21.07 68.19 58.98 56.47 20.58 67.17 58.61 56.01
Delaware 19.33 59.55 55.60 52.37 18.35 60.80 54.58 51.65 17.82 60.83 55.69 52.46
District of Columbia 12.79 56.79 47.05 44.86 12.30 54.46 47.40 44.75 12.19 54.78 47.55 44.90
Florida 14.29 54.97 49.41 46.74 14.49 54.54 48.18 45.73 14.37 54.52 48.63 46.08
Georgia 15.25 51.36 44.32 42.17 15.21 51.99 44.52 42.40 15.08 52.12 45.60 43.23
Hawaii 18.06 53.19 47.56 45.34 16.69 53.32 45.70 43.76 16.83 54.25 44.93 43.28
Idaho 13.26 45.80 44.71 41.07 13.87 46.71 45.86 42.14 14.26 47.38 44.83 41.54
Illinois 18.57 60.45 52.50 50.08 17.93 60.94 53.43 50.79 18.05 60.88 53.25 50.66
Indiana 12.36 57.54 43.45 42.11 13.20 54.87 41.21 40.13 13.54 54.63 40.62 39.69
Iowa 18.14 72.11 54.93 53.35 19.08 70.09 52.58 51.41 19.31 70.01 52.82 51.60
Kansas 16.21 59.50 51.88 48.90 15.92 60.52 50.39 47.90 16.15 59.84 49.92 47.48
Kentucky 14.37 53.65 31.89 33.01 14.72 52.77 33.36 34.05 14.65 52.98 33.81 34.41
Louisiana 20.91 57.09 44.38 43.45 19.81 57.60 43.99 43.13 19.44 57.94 44.99 43.88
Maine 19.57 66.12 50.37 49.15 19.59 65.57 51.67 50.12 19.42 64.98 51.49 49.90
Maryland 16.73 58.43 50.50 48.12 15.83 57.64 52.02 49.13 15.70 58.01 52.11 49.23
Massachusetts 22.64 72.54 64.18 61.11 22.45 72.50 62.94 60.14 21.94 71.74 63.35 60.33
Michigan 16.76 62.59 49.83 48.01 16.59 61.81 48.81 47.11 16.70 61.71 48.65 46.99
Minnesota 18.87 70.14 57.10 54.66 18.73 70.34 56.15 53.97 18.88 70.30 55.86 53.76
Mississippi 13.74 44.82 29.16 29.62 13.52 43.60 28.99 29.28 13.50 43.48 29.75 29.83
Missouri 15.74 58.28 46.88 45.09 15.77 59.29 46.91 45.24 16.06 58.94 46.37 44.81
Montana 22.06 64.60 60.06 56.45 21.80 63.29 59.72 56.01 22.00 64.31 58.61 55.32
Nebraska 18.27 71.64 55.17 53.26 18.71 71.40 54.32 52.69 19.03 71.32 53.95 52.44
Nevada 18.56 56.27 57.20 53.40 17.51 56.68 57.14 53.30 17.89 57.53 56.10 52.62
New Hampshire 21.83 65.03 62.06 58.22 21.72 68.70 59.95 57.01 21.15 67.38 60.50 57.23
New Jersey 18.44 60.22 54.05 51.45 18.00 60.28 54.39 51.68 17.69 58.95 54.36 51.47
New Mexico 17.79 58.97 57.04 52.70 18.41 56.98 55.34 51.22 18.60 58.00 54.71 50.92
New York 17.25 60.82 51.58 49.31 17.30 60.87 51.74 49.44 17.18 60.39 51.92 49.51
North Carolina 12.66 53.66 35.37 35.28 13.00 52.63 37.21 36.62 12.93 52.77 37.69 37.01
North Dakota 27.90 79.70 57.52 57.22 26.15 76.04 58.32 57.08 26.21 76.27 57.75 56.69
Ohio 15.55 59.46 49.10 47.00 15.67 59.01 47.23 45.52 15.81 58.97 47.20 45.50
Oklahoma 14.91 52.33 35.51 35.43 14.59 51.09 36.62 36.07 14.46 51.18 37.35 36.62
Oregon 14.91 58.31 52.99 49.90 15.45 57.86 53.86 50.63 15.88 58.92 52.17 49.48
Pennsylvania 15.24 60.27 50.42 48.14 15.84 60.80 50.92 48.65 15.79 60.47 51.09 48.74
Rhode Island 16.87 64.11 58.04 54.56 18.03 66.25 55.47 52.93 17.65 65.13 55.89 53.10
South Carolina 13.51 46.54 39.05 37.31 13.55 47.19 36.18 35.19 13.31 47.07 37.15 35.89
South Dakota 19.89 69.38 55.11 52.91 19.81 69.77 51.86 50.55 20.02 69.60 52.00 50.66
Tennessee 13.36 43.32 32.93 32.31 13.25 44.92 33.73 33.11 13.10 44.79 34.67 33.81
Texas 17.98 54.73 44.86 43.15 17.73 54.62 44.39 42.76 17.70 54.66 44.76 43.04
Utah 9.50 29.49 28.86 26.15 10.00 32.31 32.71 29.36 10.62 32.66 30.55 28.07
Vermont 19.34 69.47 60.31 57.02 20.02 68.63 59.46 56.38 19.63 67.87 59.79 56.50
Virginia 10.76 58.64 49.25 46.57 12.26 57.77 47.36 45.14 12.24 57.91 48.40 45.96
Washington 14.90 50.23 52.12 48.07 14.82 50.20 52.48 48.33 15.10 50.60 50.69 47.03
West Virginia 17.33 46.46 31.92 32.42 16.20 48.16 32.00 32.59 16.03 48.09 32.18 32.70
Wisconsin 21.94 68.25 64.33 60.22 21.47 68.97 60.17 57.15 21.46 68.92 60.29 57.24
Wyoming 21.45 64.93 52.83 50.90 20.53 62.03 54.02 51.27 21.00 63.32 52.28 50.22

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.21 Past Month Use of Cigarettes Based on Pooled 1999 and 2000 Data
State Design-Based Estimates SAE with Region SAE Without Region
12-17 18-25 26+ Total 12-17 18-25 26+ Total 12-17 18-25 26+ Total
National 14.16 38.97 24.57 25.34 14.15 38.68 24.60 25.33 14.15 38.69 24.59 25.32
Northeast 14.28 40.17 23.38 24.49 14.36 39.56 23.37 24.41 14.20 39.44 23.83 24.74
Midwest 16.13 44.60 25.90 27.35 16.10 44.33 25.97 27.35 16.21 44.00 25.74 27.15
South 14.72 38.01 25.49 25.98 14.65 37.86 25.59 26.03 14.49 37.94 25.64 26.06
West 11.22 33.67 22.73 22.94 11.23 33.42 22.63 22.83 11.47 33.77 22.34 22.69
Alabama 16.09 34.65 25.86 26.01 15.99 35.86 26.32 26.51 15.81 35.90 26.45 26.60
Alaska 15.80 44.44 24.55 26.31 15.95 43.10 23.29 25.22 16.37 43.83 23.13 25.26
Arizona 13.59 40.43 23.08 24.31 13.49 38.97 23.41 24.36 13.95 39.60 23.04 24.21
Arkansas 20.59 36.13 28.64 28.77 18.89 37.20 28.47 28.60 18.56 37.29 28.35 28.48
California 8.73 29.62 22.07 21.62 8.76 29.65 21.84 21.45 8.85 29.78 21.58 21.29
Colorado 18.13 45.29 22.90 25.35 16.62 43.20 23.40 25.28 17.17 44.01 23.17 25.27
Connecticut 18.87 43.50 21.06 23.40 16.90 42.29 21.74 23.62 16.63 42.18 22.40 24.11
Delaware 17.77 45.50 25.13 26.89 15.55 43.65 25.75 26.91 15.39 43.72 25.78 26.93
District of Columbia 9.05 33.27 25.17 24.61 10.15 32.20 25.23 24.62 9.90 32.42 25.28 24.66
Florida 10.92 34.60 25.16 24.85 11.41 34.87 25.26 25.00 11.30 34.96 25.23 24.98
Georgia 14.54 37.33 27.21 27.23 14.37 37.23 26.34 26.53 14.19 37.29 26.45 26.60
Hawaii 11.95 39.86 22.32 23.40 11.47 38.76 21.38 22.47 12.05 39.63 21.14 22.45
Idaho 10.86 35.30 23.15 23.51 12.31 35.76 23.45 23.98 12.64 36.28 23.14 23.87
Illinois 15.61 43.44 25.57 26.93 15.44 43.13 25.62 26.91 15.54 42.94 25.46 26.77
Indiana 17.40 41.86 26.03 27.26 16.18 41.73 26.72 27.63 16.28 41.31 26.24 27.22
Iowa 17.03 42.70 25.49 26.89 16.76 43.36 24.39 26.11 16.89 42.87 24.02 25.77
Kansas 11.82 39.50 23.55 24.38 13.60 40.19 23.31 24.49 13.66 39.62 22.97 24.16
Kentucky 23.42 47.03 32.57 33.57 22.58 46.89 31.34 32.51 22.42 47.01 31.32 32.49
Louisiana 15.12 40.44 27.79 28.13 15.03 40.02 27.22 27.66 14.92 40.05 27.21 27.63
Maine 15.48 45.99 25.41 26.82 17.04 44.96 24.77 26.34 16.72 44.67 25.23 26.64
Maryland 15.89 34.40 21.41 22.40 14.25 34.89 22.80 23.39 14.04 35.06 22.84 23.42
Massachusetts 15.59 42.50 21.65 23.52 15.69 41.48 22.22 23.84 15.49 41.33 22.84 24.29
Michigan 15.68 43.81 25.14 26.57 15.82 43.54 26.01 27.22 15.88 43.29 25.88 27.09
Minnesota 19.87 49.38 25.84 28.36 19.14 48.88 24.63 27.28 19.34 48.36 24.35 27.02
Mississippi 14.93 33.45 26.85 26.43 15.04 33.61 27.36 26.83 14.79 33.65 27.39 26.83
Missouri 14.06 46.38 30.55 30.88 14.91 45.98 28.62 29.43 15.09 45.62 28.29 29.15
Montana 17.68 41.29 21.91 23.94 17.05 40.90 22.32 24.13 17.47 41.39 22.10 24.07
Nebraska 12.48 44.69 20.10 22.61 13.77 43.52 21.23 23.45 13.91 43.01 20.90 23.14
Nevada 17.31 41.30 30.51 30.52 15.85 40.35 28.98 29.07 16.42 41.27 28.77 29.07
New Hampshire 15.67 41.56 24.68 25.73 16.04 42.65 23.67 25.12 15.76 42.35 24.15 25.43
New Jersey 12.75 41.71 21.11 22.73 12.68 39.79 21.63 22.91 12.58 39.67 22.24 23.37
New Mexico 15.06 42.85 22.76 24.70 15.33 40.45 23.74 25.11 15.80 41.19 23.73 25.26
New York 12.28 36.29 23.95 24.31 12.38 36.20 23.72 24.13 12.25 36.08 24.09 24.39
North Carolina 17.85 45.13 26.34 27.77 17.41 43.80 26.87 27.97 17.29 44.04 27.03 28.11
North Dakota 21.34 47.22 24.46 27.35 20.50 46.50 23.76 26.59 20.61 46.03 23.65 26.46
Ohio 15.83 45.66 28.21 29.21 15.79 44.99 28.28 29.17 15.89 44.87 28.18 29.09
Oklahoma 13.51 43.56 29.91 29.88 15.25 42.66 27.97 28.50 14.95 42.67 28.07 28.54
Oregon 14.42 43.04 23.06 24.70 14.66 41.42 23.99 25.24 15.03 41.99 23.61 25.05
Pennsylvania 16.21 42.32 24.97 26.14 16.55 41.93 24.72 25.93 16.44 41.88 25.08 26.20
Rhode Island 12.94 37.18 27.36 27.02 14.27 37.50 25.99 26.11 14.03 37.25 26.40 26.39
South Carolina 16.47 35.01 23.95 24.54 15.68 35.42 25.51 25.71 15.49 35.46 25.54 25.72
South Dakota 19.36 45.25 21.87 24.86 18.90 45.11 22.44 25.19 18.95 44.60 22.22 24.96
Tennessee 16.88 45.67 26.87 28.28 17.35 44.17 26.87 28.12 17.15 44.38 27.02 28.24
Texas 12.73 34.49 23.12 23.57 12.73 34.67 23.23 23.68 12.59 34.67 23.26 23.68
Utah 8.33 22.37 12.81 13.96 9.82 24.61 17.12 17.46 10.05 24.75 16.91 17.39
Vermont 14.40 44.69 22.61 24.42 14.72 43.94 22.13 24.01 14.49 43.66 22.59 24.30
Virginia 12.85 39.23 22.00 23.19 13.31 38.26 22.90 23.83 13.13 38.41 22.92 23.84
Washington 12.69 32.80 26.02 25.51 13.12 33.88 24.13 24.23 13.44 34.42 23.54 23.87
West Virginia 19.75 46.11 30.03 31.12 20.06 45.35 29.21 30.40 19.90 45.55 29.34 30.51
Wisconsin 17.60 46.43 22.00 24.79 17.58 46.63 23.72 26.09 17.70 45.96 23.46 25.81
Wyoming 16.30 42.58 23.32 25.28 15.79 42.06 23.87 25.54 16.20 42.83 23.67 25.55

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.22 Relative Absolute Bias for Past Month Use of Marijuana Based on Pooled 1999 and 2000 Data
State Age in Years Total
12-17 18-25 26+
CA (design-based) 7.60 13.94 4.16 5.86
CA1 7.30 14.22 5.26 6.70
CA2 7.69 13.28 3.41 5.22
CA3 7.34 13.53 3.08 4.96
CA4 7.53 12.78 3.34 5.09
Average across 4 substates 7.47 13.45 3.77 5.49
Relative Absolute Bias 1.75 3.54 9.28 6.35
FL (design-based) 6.33 13.31 3.39 4.73
FL1 7.39 14.64 3.59 5.13
FL2 7.51 11.76 3.68 4.90
FL3 5.66 13.22 3.10 4.42
FL4 6.63 13.52 3.71 5.03
Average across 4 substates 6.80 13.28 3.52 4.87
Relative Absolute Bias 7.39 0.19 3.77 3.02
IL (design-based) 8.57 14.31 2.51 4.70
IL1 7.41 17.15 3.02 5.35
IL2 8.27 14.36 2.85 4.93
IL3 6.83 12.49 2.46 4.24
IL4 8.25 13.82 2.66 4.72
Average across 4 substates 7.69 14.45 2.75 4.81
Relative Absolute Bias 10.24 1.01 9.66 2.44
MI (design-based) 7.77 16.64 3.53 5.68
MI1 7.85 18.00 3.40 5.77
MI2 7.90 17.12 3.48 5.72
MI3 7.77 16.17 3.34 5.48
MI4 8.51 16.37 3.37 5.60
Average across 4 substates 8.01 16.92 3.40 5.64
Relative Absolute Bias 3.08 1.69 3.71 0.68
NY (design-based) 6.32 16.77 2.02 4.26
NY1 7.29 14.69 2.54 4.51
NY2 7.06 15.60 2.70 4.72
NY3 7.38 15.26 2.85 4.82
NY4 6.61 15.98 2.40 4.49
Average across 4 substates 7.08 15.38 2.62 4.63
Relative Absolute Bias 12.08 8.26 29.53 8.69
OH (design-based) 6.07 14.31 2.49 4.40
OH1 5.92 13.31 2.21 4.04
OH2 6.83 13.14 2.57 4.38
OH3 7.00 14.75 2.60 4.63
OH4 6.98 14.72 2.37 4.45
Average across 4 substates 6.68 13.98 2.44 4.38
Relative Absolute Bias 10.03 2.31 2.17 0.50
PA (design-based) 5.83 14.16 2.79 4.42
PA1 7.45 14.49 3.13 4.88
PA2 6.33 12.47 2.40 3.96
PA3 6.37 13.12 2.58 4.18
PA4 7.11 15.58 2.88 4.78
Average across 4 substates 6.81 13.91 2.75 4.45
Relative Absolute Bias 16.90 1.75 1.63 0.71
TX (design-based) 6.00 10.41 1.34 3.22
TX1 5.99 9.27 1.63 3.26
TX2 6.01 11.52 1.86 3.77
TX3 6.06 10.75 1.84 3.64
TX4 5.29 10.83 1.77 3.51
Average across 4 substates 5.84 10.59 1.77 3.55
Relative Absolute Bias 2.65 1.79 32.35 10.19
Average Relative Absolute Bias 8.01 2.57 11.51 4.07

Note: Relative Absolute Bias = 100 × abs(Average SAE over 4 substates - Large State design-based estimate) / Large State design-based estimate.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.23 Relative Absolute Bias for Past Year Use of Cocaine Based on Pooled 1999 and 2000 Data
State Age in Years Total
12-17 18-25 26+
CA (design-based) 2.05 4.79 1.29 1.85
CA1 1.86 4.21 0.99 1.52
CA2 1.82 4.81 1.00 1.61
CA3 2.20 5.43 1.63 2.21
CA4 2.13 4.58 1.06 1.65
Average across 4 substates 2.00 4.76 1.17 1.75
Relative Absolute Bias 2.19 0.65 9.24 5.36
FL (design-based) 1.52 5.96 1.18 1.73
FL1 1.72 5.56 1.49 1.95
FL2 1.53 4.13 1.01 1.40
FL3 1.55 5.18 1.20 1.66
FL4 1.42 4.58 1.10 1.50
Average across 4 substates 1.55 4.86 1.20 1.63
Relative Absolute Bias 2.18 18.43 1.60 5.78
IL (design-based) 0.96 3.93 1.11 1.47
IL1 1.32 4.42 1.06 1.53
IL2 1.49 4.88 1.19 1.71
IL3 1.61 4.50 0.98 1.52
IL4 1.24 3.66 1.07 1.43
Average across 4 substates 1.41 4.36 1.08 1.55
Relative Absolute Bias 47.60 10.97 2.88 5.43
MI (design-based) 1.02 5.04 0.86 1.42
MI1 1.27 5.32 1.22 1.76
MI2 1.46 4.34 1.04 1.51
MI3 1.38 4.54 1.00 1.50
MI4 1.53 4.68 1.25 1.72
Average across 4 substates 1.41 4.72 1.13 1.62
Relative Absolute Bias 38.23 6.34 30.39 14.07
NY (design-based) 1.18 3.87 1.01 1.37
NY1 1.31 4.08 1.07 1.46
NY2 1.41 4.70 1.08 1.56
NY3 1.71 4.52 1.05 1.54
NY4 1.41 3.90 1.20 1.55
Average across 4 substates 1.46 4.30 1.10 1.53
Relative Absolute Bias 23.10 11.10 9.67 11.32
OH (design-based) 0.78 4.98 0.92 1.43
OH1 1.26 4.72 1.18 1.65
OH2 1.32 4.32 1.03 1.49
OH3 1.40 5.03 1.04 1.59
OH4 1.31 4.64 1.05 1.54
Average across 4 substates 1.32 4.68 1.07 1.57
Relative Absolute Bias 69.04 6.15 17.00 9.45
PA (design-based) 1.18 4.39 1.00 1.41
PA1 1.50 4.13 1.01 1.42
PA2 1.24 4.33 0.82 1.27
PA3 1.55 4.73 1.18 1.63
PA4 1.59 5.10 1.06 1.58
Average across 4 substates 1.47 4.57 1.02 1.48
Relative Absolute Bias 25.11 4.01 1.71 4.45
TX (design-based) 2.66 6.10 0.83 1.82
TX1 2.48 5.43 1.04 1.85
TX2 2.15 5.69 1.19 1.96
TX3 2.29 5.34 1.21 1.94
TX4 2.35 5.73 1.26 2.04
Average across 4 substates 2.32 5.54 1.17 1.95
Relative Absolute Bias 12.90 9.07 41.21 7.17
Average Relative Absolute Bias 27.54 8.34 14.21 7.88

Note: Relative Absolute Bias = 100 × abs(Average SAE over 4 substates - Large State design-based estimate) / Large State design-based estimate.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.24 Relative Absolute Bias for Past Month "Binge" Alcohol Use Based on Pooled 1999 and 2000 Data
State Age in Years Total
12-17 18-25 26+
CA (design-based) 9.12 32.46 18.58 19.42
CA1 9.01 31.04 18.40 19.08
CA2 8.50 30.35 17.77 18.45
CA3 9.77 35.11 19.18 20.30
CA4 9.35 32.13 18.83 19.59
Average across 4 substates 9.16 32.16 18.55 19.36
Relative Absolute Bias 0.40 0.93 0.18 0.32
FL (design-based) 7.93 35.02 17.72 18.67
FL1 8.56 34.49 17.41 18.42
FL2 9.49 32.97 17.82 18.67
FL3 8.30 34.20 17.06 18.08
FL4 8.83 37.06 18.13 19.29
Average across 4 substates 8.79 34.68 17.60 18.62
Relative Absolute Bias 10.94 0.97 0.68 0.28
IL (design-based) 11.53 41.83 21.43 23.13
IL1 9.98 41.10 19.09 21.08
IL2 12.38 43.38 22.76 24.44
IL3 10.39 40.60 20.69 22.28
IL4 11.25 41.42 21.46 23.07
Average across 4 substates 11.00 41.62 21.00 22.72
Relative Absolute Bias 4.60 0.50 2.00 1.77
MI (design-based) 10.88 42.23 19.08 21.23
MI1 9.65 40.35 18.37 20.31
MI2 11.00 41.60 19.54 21.51
MI3 10.68 41.86 20.57 22.30
MI4 12.03 40.64 19.72 21.63
Average across 4 substates 10.84 41.12 19.55 21.44
Relative Absolute Bias 0.38 2.64 2.47 1.00
NY (design-based) 10.14 39.47 18.61 20.33
NY1 8.99 34.89 18.74 19.76
NY2 9.35 42.29 18.54 20.54
NY3 10.66 40.54 19.07 20.87
NY4 10.68 38.23 18.62 20.24
Average across 4 substates 9.92 38.99 18.74 20.35
Relative Absolute Bias 2.25 1.22 0.72 0.11
OH (design-based) 9.97 41.73 20.32 22.04
OH1 9.86 42.08 19.89 21.74
OH2 11.03 42.39 19.08 21.28
OH3 10.31 40.68 20.32 21.93
OH4 10.48 41.52 20.50 22.20
Average across 4 substates 10.42 41.67 19.95 21.79
Relative Absolute Bias 4.48 0.15 1.84 1.13
PA (design-based) 9.30 42.13 20.55 21.97
PA1 11.17 44.23 20.02 21.99
PA2 9.94 39.78 19.15 20.66
PA3 10.03 42.50 20.54 22.08
PA4 9.64 41.15 18.97 20.65
Average across 4 substates 10.20 41.92 19.67 21.35
Relative Absolute Bias 9.67 0.50 4.26 2.84
TX (design-based) 11.07 35.62 20.08 21.31
TX1 10.66 34.17 19.50 20.62
TX2 10.31 40.01 20.94 22.50
TX3 11.64 32.77 19.33 20.40
TX4 10.53 37.29 20.84 22.04
Average across 4 substates 10.78 36.06 20.15 21.39
Relative Absolute Bias 2.59 1.24 0.35 0.39
Average Relative Absolute Bias 4.41 1.02 1.56 0.98

Note: Relative Absolute Bias = 100 × abs(Average SAE over 4 substates - Large State design-based estimate) / Large State design-based estimate.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.25 Relative Absolute Bias for Past Month Use of Cigarettes Based on Pooled 1999 and 2000 Data
State Age in Years Total
12-17 18-25 26+
CA (design-based) 8.73 29.62 22.07 21.62
CA1 9.53 30.29 22.31 21.99
CA2 8.96 32.22 21.72 21.74
CA3 9.30 30.84 21.28 21.25
CA4 8.87 29.25 20.82 20.65
Average across 4 substates 9.16 30.65 21.53 21.41
Relative Absolute Bias 4.93 3.49 2.41 0.98
FL (design-based) 10.92 34.60 25.16 24.85
FL1 10.67 35.75 24.02 24.04
FL2 12.75 34.25 25.95 25.62
FL3 10.10 35.28 25.34 24.99
FL4 12.85 36.80 24.37 24.63
Average across 4 substates 11.59 35.52 24.92 24.82
Relative Absolute Bias 6.16 2.65 0.96 0.13
IL (design-based) 15.61 43.44 25.57 26.93
IL1 13.45 41.65 24.49 25.64
IL2 16.27 44.36 27.65 28.70
IL3 16.16 42.55 24.36 25.94
IL4 14.76 40.70 24.66 25.78
Average across 4 substates 15.16 42.31 25.29 26.51
Relative Absolute Bias 2.87 2.60 1.11 1.53
MI (design-based) 15.68 43.81 25.14 26.57
MI1 16.08 44.79 25.40 26.94
MI2 16.06 42.04 26.92 27.74
MI3 14.70 42.18 26.85 27.57
MI4 16.79 42.27 26.36 27.42
Average across 4 substates 15.91 42.82 26.38 27.42
Relative Absolute Bias 1.45 2.27 4.93 3.17
NY (design-based) 12.28 36.29 23.95 24.31
NY1 12.89 37.88 25.07 25.44
NY2 12.52 35.60 23.19 23.65
NY3 11.36 33.66 23.54 23.58
NY4 11.98 38.06 24.53 24.94
Average across 4 substates 12.19 36.30 24.08 24.40
Relative Absolute Bias 0.76 0.03 0.54 0.38
OH (design-based) 15.83 45.66 28.21 29.21
OH1 16.20 44.45 26.39 27.69
OH2 14.48 43.82 26.18 27.27
OH3 17.37 47.51 29.15 30.32
OH4 16.18 42.92 29.07 29.54
Average across 4 substates 16.06 44.67 27.70 28.71
Relative Absolute Bias 1.45 2.16 1.83 1.71
PA (design-based) 16.21 42.32 24.97 26.14
PA1 16.94 39.47 23.81 24.97
PA2 17.27 42.45 26.64 27.57
PA3 14.71 40.80 24.96 25.81
PA4 16.54 44.22 25.88 27.11
Average across 4 substates 16.36 41.74 25.32 26.37
Relative Absolute Bias 0.97 1.39 1.44 0.87
TX (design-based) 12.73 34.49 23.12 23.57
TX1 13.11 35.49 23.15 23.78
TX2 11.83 36.65 23.12 23.78
TX3 12.43 32.79 22.03 22.48
TX4 12.17 35.52 25.17 25.17
Average across 4 substates 12.39 35.11 23.37 23.80
Relative Absolute Bias 2.74 1.79 1.07 0.98
Average Relative Absolute Bias 2.67 2.05 1.79 1.22

Note: Relative Absolute Bias = 100 × abs(Average SAE over 4 substates - Large State design-based estimate) / Large State design-based estimate.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999 and 2000.

Table B.26 Ratio of Average Widths for Pooled 1999 and 2000 Data
State Age in Years Total
12-17 18-25 26+
Past Month Use of Marijuana
CA 0.76 0.71 0.75 0.76
FL 0.72 0.76 0.77 0.81
IL 0.62 0.70 0.79 0.74
MI 0.72 0.81 0.73 0.80
NY 0.79 0.70 0.91 0.85
OH 0.67 0.64 0.62 0.67
PA 0.71 0.65 0.65 0.71
TX 0.72 0.72 0.67 0.75
Average 0.71 0.71 0.74 0.76
Past Year Use of Cocaine
CA 0.70 0.66 0.52 0.58
FL 0.53 0.60 0.60 0.64
IL 0.65 0.66 0.46 0.54
MI 0.54 0.58 0.59 0.65
NY 0.46 0.71 0.75 0.79
OH 0.60 0.62 0.61 0.68
PA 0.61 0.59 0.50 0.57
TX 0.62 0.65 0.72 0.71
Average 0.59 0.63 0.59 0.65
Past Month "Binge" Alcohol Use
CA 0.82 0.76 0.77 0.81
FL 0.71 0.63 0.72 0.73
IL 0.64 0.66 0.70 0.69
MI 0.69 0.75 0.71 0.71
NY 0.74 0.60 0.76 0.77
OH 0.85 0.60 0.75 0.72
PA 0.75 0.59 0.70 0.69
TX 0.79 0.71 0.70 0.72
Average 0.75 0.66 0.73 0.73
Past Month Use of Cigarettes
CA 0.82 0.84 0.65 0.66
FL 0.71 0.74 0.86 0.86
IL 0.67 0.83 0.69 0.69
MI 0.79 0.71 0.73 0.72
NY 0.64 0.76 0.82 0.82
OH 0.72 0.81 0.75 0.75
PA 0.72 0.69 0.81 0.78
TX 0.72 0.74 0.68 0.66
Average 0.72 0.77 0.75 0.74

Note: Ratio = Average width of model-based PIs for substates / Average width of design-based CIs for substates.

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 2000.

Table B.27 1999 NHSDA Weighted Screening and Interview Response Rates, by State
State Screening Response Rate Interview Response Rate Overall Response Rate State Screening Response Rate Interview Response Rate Overall Response Rate
Total 89.63 68.55 61.44 Missouri 91.32 73.59 67.21
Alabama 92.60 71.36 66.08 Montana 92.76 76.39 70.86
Alaska 91.07 77.20 70.31 Nebraska 89.99 72.05 64.84
Arizona 94.43 65.87 62.21 Nevada 79.89 63.05 50.37
Arkansas 95.71 80.45 77.00 New Hampshire 85.36 69.87 59.65
California 87.47 64.12 56.08 New Jersey 89.65 65.24 58.48
Colorado 91.62 65.84 60.32 New Mexico 96.12 77.77 74.75
Connecticut 85.62 58.60 50.17 New York 84.28 59.98 50.55
Delaware 87.13 58.36 50.85 North Carolina 92.87 71.84 66.72
District of Columbia 93.35 79.93 74.61 North Dakota 89.89 77.48 69.65
Florida 89.94 68.20 61.33 Ohio 90.35 67.78 61.24
Georgia 90.47 66.97 60.59 Oklahoma 91.58 67.79 62.08
Hawaii 89.11 67.61 60.25 Oregon 85.20 71.57 60.98
Idaho 92.93 75.45 70.11 Pennsylvania 92.34 68.99 63.71
Illinois 87.35 63.74 55.68 Rhode Island 86.68 66.72 57.83
Indiana 91.68 73.06 66.98 South Carolina 91.96 65.92 60.61
Iowa 92.44 69.69 64.41 South Dakota 94.35 76.14 71.84
Kansas 90.59 72.89 66.03 Tennessee 90.92 67.70 61.56
Kentucky 92.36 73.75 68.12 Texas 92.57 75.12 69.54
Louisiana 94.81 76.97 72.98 Utah 93.16 81.70 76.11
Maine 89.96 75.18 67.63 Vermont 90.26 74.49 67.24
Maryland 87.78 64.66 56.76 Virginia 89.84 66.28 59.55
Massachusetts 80.59 61.82 49.82 Washington 86.49 75.06 64.92
Michigan 88.21 66.54 58.70 West Virginia 95.59 74.31 71.03
Minnesota 89.46 77.72 69.53 Wisconsin 90.19 73.05 65.89
Mississippi 94.51 82.77 78.23 Wyoming 93.79 72.62 68.11

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 1999.

Table B.28 2000 NHSDA Weighted Screening and Interview Response Rates, by State
State Screening Response Rate Interview Response Rate Overall Response Rate State Screening Response Rate Interview Response Rate Overall Response Rate
Total 92.84 73.93 68.64 Missouri 92.25 70.80 65.31
Alabama 95.50 77.98 74.47 Montana 94.91 80.21 76.13
Alaska 95.43 80.24 76.58 Nebraska 93.13 74.58 69.46
Arizona 92.99 73.78 68.61 Nevada 92.08 74.44 68.54
Arkansas 97.19 81.00 78.73 New Hampshire 92.41 75.12 69.42
California 90.99 69.50 63.24 New Jersey 91.96 66.56 61.21
Colorado 94.84 75.26 71.37 New Mexico 97.43 80.80 78.72
Connecticut 89.83 71.36 64.10 New York 88.78 73.73 65.46
Delaware 92.91 68.25 63.42 North Carolina 94.51 73.19 69.17
District of Columbia 93.50 85.56 80.00 North Dakota 94.43 79.46 75.03
Florida 94.64 75.73 71.67 Ohio 94.89 75.79 71.92
Georgia 92.95 69.76 64.84 Oklahoma 93.06 74.85 69.66
Hawaii 91.95 78.45 72.14 Oregon 91.87 73.91 67.90
Idaho 93.94 74.45 69.94 Pennsylvania 94.37 73.50 69.36
Illinois 88.71 65.59 58.19 Rhode Island 91.26 74.11 67.63
Indiana 92.62 73.87 68.42 South Carolina 94.69 77.84 73.71
Iowa 94.78 80.00 75.83 South Dakota 95.15 76.67 72.95
Kansas 92.28 73.45 67.79 Tennessee 90.25 72.45 65.39
Kentucky 95.79 84.14 80.59 Texas 94.72 78.12 74.00
Louisiana 95.04 80.81 76.80 Utah 95.11 83.44 79.36
Maine 92.39 78.46 72.49 Vermont 92.62 80.80 74.83
Maryland 94.88 76.88 72.94 Virginia 91.44 75.18 68.75
Massachusetts 89.77 66.45 59.65 Washington 93.59 75.45 70.61
Michigan 93.19 73.18 68.20 West Virginia 95.19 78.17 74.41
Minnesota 94.66 80.62 76.32 Wisconsin 94.33 75.06 70.81
Mississippi 93.60 79.14 74.07 Wyoming 95.41 76.61 73.09

Source: SAMHSA, Office of Applied Studies, National Household Survey on Drug Abuse, 2000.

1 The eight large sample States are California, Florida, Illinois, Michigan, New York, Ohio, Pennsylvania, and Texas.

2 The panel included William Bell of the U.S. Bureau of the Census; Partha Lahiri of the University of Nebraska; Balgobin Nandram of Worcester Polytechnic Institute and the National Center for Health Statistics; Wesley Schaible, formerly Associate Commissioner for Research and Evaluation at the Bureau of Labor Statistics; J.N.K. Rao of Carleton University; and Alan Zaslavsky of Harvard University. Other attendees involved in the development or discussion were Ralph Folsom, Judith Lessler, Avinash Singh, and Akhil Vaish of RTI and Joe Gfroerer and Doug Wright of SAMHSA.

Go to the Table of Contents

This page was last updated on December 30, 2008.