In 2003 the National Center for Education Statistics (NCES) conducted the National Assessment of Adult Literacy (NAAL) to measure the English literacy skills of the nation's adults. The NAAL was a follow up to the National Adult Literacy Survey (NALS), conducted by NCES in 1992. The 2003 NAAL included interviews of more than 18,500 adults (age 16 and older) who resided in private households across the country. As a part of that interview, participants were asked to perform a variety of reading and math activities using documents from daily life. The overall sample comprised a core national sample supplemented by samples in the 6 states that participated in the State Assessment of Adult Literacy (SAAL). The SAAL was designed to provide direct survey estimates of adult literacy levels for each of the participating states.1 In a similar fashion, the 1992 NALS interviewed more than 24,000 adults in private households. It consisted of a core national sample supplemented by samples in the 11 states that participated in the State Adult Literacy Survey (SALS).2,3
The 2003 and 1992 surveys were designed to provide standard survey estimates—"direct"—estimates of literacy proficiency for the nation as a whole and for major subgroups within the nation defined in terms of race/ethnicity and region, as well as for those states participating in the SAAL or the SALS. Neither survey, however, was designed to provide information for individual counties or for states outside the SAAL and SALS samples. A demand by policymakers and educators for state and county estimates of literacy prompted NCES to use statistical modeling approaches to produce estimates of adults at the lowest literacy level for individual states and counties. These model-based estimates are called "indirect" estimates to distinguish them from the direct survey estimates. The state and county indirect estimates were produced using small area estimation techniques (Rao 2003) that combine data from the surveys with data from other sources, such as the decennial censuses, that provide information on characteristics of states and counties.
The NAAL and NALS produced estimates of Prose, Document, and Quantitative literacy, each reported on a 0 to 500 scale and on four performance levels: Below Basic, Basic, Intermediate, and Proficient. The specific measure chosen to estimate literacy at the state and county levels on this website is the percentage of adults . The literacy of adults who lack BPLS ranges from being unable to read and understand any written information in English to being able to locate easily identifiable information in short, commonplace prose text, but nothing more advanced.4 However, adults who were not able to take the assessment because of a language barrier are included in the indirect estimates and are classified as lacking BPLS on the grounds that they can be considered to be at the lowest level of English literacy. Users should note that the indirect estimates of the percentages of adults lacking BPLS are not comparable to the percentages of adults Below Basic in prose literacy in other NAAL or NALS published results because the latter results exclude adults who were unable to take the assessment because of a language barrier. The national direct estimates of the percentages of adults lacking BPLS are 14.5 percent for the 2003 NAAL and 14.7 percent for the 1992 NALS. In comparison, the national direct estimates of the percentages Below Basic in prose literacy are 13.6 percent for the NAAL and 13.8 percent for the NALS.
Small Area Estimation Method
County indirect estimates were produced applying small area estimation techniques that use a statistical model to relate the estimated percentage of adults lacking Basic Prose Literacy Skills (BPLS) in a county with sample members to predictor variables available from external sources, such as levels of educational attainment obtained from the decennial censuses. On the basis of the observed relationship between the survey county estimates of the percentages lacking BPLS and predictor variables from the external sources, it is possible to estimate the percentage lacking BPLS for any county in the United States using that county's values on the predictor variables.
The primary sources for the predictor variables used to create the indirect estimates were the 2000 Census for the 2003 model and the 1990 Census for the 1992 model. Both models included predictor variables related to educational attainment and race/ethnicity, indicators for census divisions, and SAAL or SALS state assessment indicators. There are some differences between the two models. For instance, extensive model testing resulted in the inclusion of foreign-born status and poverty in the 2003 model only, whereas native-English-speaking status was used in the 1992 model only.
Although care was taken in selecting the predictor variables from a range of available external sources, the ability of the external variables to predict the percentage of adults lacking BPLS by county was limited. (which are similar to traditional confidence intervals) indicate the level of uncertainty (referred to as prediction error) in the indirect estimates. Users need to pay careful attention to the 95 percent credible interval bounds that are provided along with the estimated percentage of adults lacking BPLS to assess the range of uncertainty in the estimates.
The indirect estimates for state were computed as weighted aggregates of the county indirect estimates, where the weights were based on adjusted Census Bureau 2003 postcensal estimated residency counts. For more information on the estimation methods, see Estimation Approach.
A more detailed description can be found in the NAAL Small Area Estimation Technical Report (U.S. Department of Education, National Center for Education Statistics 2008).
Indirect Estimates of Adults Lacking Basic Prose Literacy Skills (BPLS)
This website provides indirect estimates of the percentages of adults lacking BPLS in 2003 and 1992 for each county and state in the United States. It also allows users to compare the percentages of adults lacking BPLS across ANY two states or counties that are specified in advance, and across years for a single state or county. See General Cautions when making multiple comparisons for pairs of states or counties. The website does not allow users to compare estimates at different aggregation levels, such as a state versus the nation or a state versus a county.
For estimates at a higher aggregation level (e.g., national estimates) or state-level estimates for 2003 SAAL or 1992 SALS states, please refer to the main NAAL website.
Note, however, that the state- and county-level indirect estimates of the percentages of adults lacking BPLS reported here are not comparable to the percentages Below Basic in prose literacy in other NAAL or NALS published results (see General Cautions).
For the 2003 NAAL, the national direct estimate of the percentage of adults lacking BPLS is 14.5 percent. For the 1992 NALS, the national direct estimate is 14.7 percent.
1 The 2003 SAAL states were Kentucky, Maryland, Massachusetts, Missouri, New York, and Oklahoma.
2 The 1992 SALS states were California, Illinois, Indiana, Iowa, Louisiana, New Jersey, New York, Ohio, Pennsylvania, Texas, and Washington.
3 In addition to the household sample, both surveys included samples of adults from federal and state prisons. The inmate samples do not contribute to the small area estimates.
4 For more information about performance levels, see White and Dillow (2005) or NAAL's webpage on Performance Levels.