Q1. What does NAAL measure?

1. The 2003 National Assessment of Adult Literacy (NAAL) examined the English-language literacy among adults ages 16 and older living in the United States. The assessment focused on authentic, everyday tasks American adults regularly encounter. NAAL defines literacy as, "the ability to use printed and written information to function in society, to achieve one's goals, and to develop one's knowledge and potential." It assessed three literacy areas: prose, document and quantitative. A nationally representative sample of more than 19,000 adults was assessed in their homes, including a prison sample of nearly 1,200 inmates.

Q2. Why does NAAL measure literacy in English only?

1. A 1988 congressional mandate and the 1991 National Literacy Act require us to evaluate the status and progress of adults' English literacy. The reason we do not also assess literacy in other languages is primarily because given the very low prevalence of non-English speakers in the United States, the costs of doing so would be prohibitive. The 2000 Census shows that Spanish is the predominant language (about 57 percent) among those who speak a language other than English. Among those who do not speak English well or at all, about 71 percent speak Spanish. We accommodate these Spanish-speaking adults in the NAAL screener, background questionnaire, core assessment, and Adult Literacy Supplemental Assessment (ALSA) through a bilingual (English/Spanish) interviewer.

Q3. How did you determine the three literacy domains-prose, document, and quantitative-to be measured?

1. Literacy is a multifaceted set of skills that adults bring to bear throughout their lives to meet a host of personal, social, and workplace needs involving retrieval of information from printed or written material. It was from this perspective that the National Assessment of Adult Literacy and its predecessors were developed.
   
   Reading is a primary facet of literacy, as it enables adults to use printed and written information to function in society, to achieve their goals, and to develop their knowledge and potential. The structure of prose passages, such as editorials, is qualitatively different from the structure associated with documents such as charts and forms. Quantitative literacy was also included since adults are often required to use printed information to perform basic arithmetic calculations (such as reading a label on a can of paint and determining how much paint to buy for a 20- by 30-foot room). In the end, three scales were developed-a prose literacy scale, a document literacy scale, and a quantitative literacy scale.

Q4. Why were the 2003 literacy performance levels developed?

1. The NationalCenter for Education Statistics (NCES) originally used five literacy levels to report the 1992 results. In preparation for reporting on adult literacy performance in 2003, we asked the National Research Council (NRC) to evaluate the original 1992 literacy levels and recommend a set of performance levels for 2003 reporting, if necessary.
   
   The main reason we asked NRC to evaluate the 1992 levels was the inability to make distinctions among those grouped in Level 1, the lowest level of literacy. It was not clear whether relatively poor performance was due to lack of basic (word-level) reading skills or lack of higher-level literacy skills. Available data did not permit detailed analysis of Level 1 performance or effective response to the news reports.

Q5. How is the term nonliterate in English defined?

1. The Nonliterate in English category identifies all adults lacking literacy in English. Non-native speakers of English may be literate in their first language even though their English skills may be poor. This is why we use the term nonliterate in English as opposed to illiterate. The category includes two groups of adults: (1) adults who have English literacy skills so low that they are unable to participate in the assessment at all; and (2) adults who are unable to take the main part of the assessment because it is too difficult for them, but who are able to take the alternative assessment specifically designed for the least-literate adults.

Q6. Setting aside procedural differences, how do the 2003 levels differ from the 1992 levels?

1. The major difference between the 2003 and the 1992 levels is the increased information about performance at the lowest level of literacy than the 2003 levels afford. Because the 1992 Level 1 was so broad, it did not inform the public whether poor performance on NAAL was due to a lack of basic reading skills or a lack of higher-level literacy skills. The new Nonliterate in English category identifies all adults lacking the minimum basic reading skills necessary to participate in the main NAAL. While some of these individuals are able to identify letters, numbers, and simple words and phrases, most are unable to read and comprehend connected text in English-NAAL's definition of literacy.

   We can further distinguish adults in the Nonliterate in English category. One group of adults within the Nonliterate in English category was unable to participate at all in the assessment because of literacy-related reasons. And, since no scores could be estimated for these adults and we had no other information about them, they were not included in the NAAL literacy results showing the performance of various groups of adults. Another group within this category was able to participate in the alternative assessment for the least-literate adults. While some of these adults were able to identify letters, numbers, and single words and phrases, most (like those who could not attempt the assessment) were unable to read and understand connected text.

   The Nonliterate in English category can be reported for the 2003 NAAL but not for the 1992 National Adult Literacy Survey (NALS). This is because the 1992 assessment did not include an alternative assessment for the least literate adults, and it is not possible to identify a group of 1992 participants corresponding to those who took the alternative assessment in 2003. Although the 1992 assessment also began with a set of easy tasks, these tasks were fewer and different from the ones used in 2003. In 2003, moreover, questions for the easy tasks could be offered in Spanish, whereas only English was used in 1992.

Q7. Who was included in the 2003 NAAL sample?

1. The 2003 NAAL sample represents all individuals age 16 and older that live in households or prisons in the United States (as well as in a dormitory, fraternity or sorority dwelling if the student was not at home on break and if the school was located in or around the 160 locations around the US with NAAL interviewers). Based on the Census 2000 data, approximately 99 percent of adults over age 16 lived either in households (about 97 percent), in college dormitories (about one percent), or in prisons (about one percent). The 2003 NAAL results are based on a total sample of 19,300 adults, excluding those who were unable to take the assessment because of literacy-related reasons. The sample includes 5,700 adults from states that chose to participate in the state assessments and 1,200 from federal and state prisons. We did not assess adults who lived in institutions other than college dormitories or prisons, such as military barracks, retirement homes, assisted-living homes, and homeless shelters. This is consistent with the approach used in 1992.

Q8. What accommodations are offered for non-native speakers of English?

1. Accommodations that are inherent in the design of NAAL aid non-native speakers as well as others. For example, administration was one-on-one; respondents received additional time if they needed it, within reason, and they may use whatever aids they usually use to work with written materials (e.g., a magnifying glass). In addition, we provided several language accommodations tailored specifically for non-native speakers of English. First, the NAAL background questionnaire is administered in English or Spanish, as preferred by the participant. Second, the general instructions and specific questions for the core screening tasks can be given in either English or Spanish, and the general instructions are given orally. Third, adults who are unable to answer a minimum number of literacy screening questions are administered an alternative assessment in which questions are asked orally in either English or Spanish. Nevertheless, printed materials that participants were asked to read were in English only, since NAAL is intended to be an assessment of English literacy. Participants with a native language other than English or Spanish were encouraged to attempt the easy literacy questions even if they could not complete the NAAL background questionnaire. Finally, (although this is not an accommodation per se) we will report separate results for non-English-speaking adults (in addition to combining their score with the overall score) and compare their results with that of native speakers of English. This way, the unique needs of this population may better be inferred.

Q9. What accommodations are offered for adults with disabilities?

1. With the passage of the Americans with Disabilities Act in 1990, and then the 1991 Individuals with Disabilities Education Act, assessment programs at all levels have been encouraged to consider the issues of inclusion and accommodations for those with disabilities. The NAAL data collection procedures accommodate such adults. This is especially important because a large percentage of adults with disabilities have low literacy skills.
   
   Accommodations are inherent in the design of NAAL. NAAL is administered in the respondent's home on a one-on-one basis, which may help reduce distractions of group administration in unfamiliar locations. Further, all respondents receive additional time if they need it and they are encouraged to use whatever aids they usually use to work with written materials (e.g., a magnifying glass). Finally, although this is not an accommodation per se, reasons for not completing tasks are recorded; this type of information helps researchers understand relationships between literacy and disabilities.

Q10. Do NAAL literacy scores correspond to grade levels?

1. Grade-level measures are misleading when applied to adults and are more suitable for children who are developing literacy skills than for adults with a wide range of literacy skills. The diversity and range of skill levels and experience represented by NAAL participants are quite different from the progressive levels of literacy attainment represented by the grade levels of developing young people. NAAL literacy scores summarize performance on various types of tasks of varying difficulty that are encountered by adults of all walks of life in their daily lives. NAAL participants bring to literacy tasks a full range of backgrounds, experiences, and skill levels; some skills are well developed, and others are underdeveloped. Some people are good readers but perform poorly on quantitative tasks, for example.
   
   The National Center for Education Statistics has no plans to establish grade-level equivalents for the NAAL literacy scores or levels. However, we recognize the need to communicate what the reported NAAL scores mean in terms of practical performance. To this end, the NAAL performance levels, for example, enable us to tell our target audiences what a score of 175, for example, means in a couple of ways. First, the levels tell us that someone with a score of 175 falls in the Below Basic level-somewhat analogous to reporting the percentages of students receiving various letter grades, such as an A or B. In addition to helping us group adults with similar scores, the levels help us report the typical performance and capabilities of specified proportions of the adult population falling within various ranges of literacy ability.

Q11. Is there any way to average scores across prose, document, and quantitative literacy so that there is just one score for literacy?

1. NCES is developing a new, combined 0-500 scale that will give overall literacy scores to be used in future reports. However, we cannot simply average together the prose, document, and quantitative scores into a single literacy score. Instead, the items from all three scales must be rescaled and reanalyzed together. Although the results of all three assessments are displayed on 0-500 scales, the scales were developed separately and a given score on one scale-225 on the prose literacy scale, for example, does not imply the same level of performance as 225 on another scale.
   
   Furthermore, the performance levels (Below Basic, Basic, Intermediate, and Proficient) were also developed separately for each of the three kinds of literacy. For example, the proportion of adults in the Below Basic level on the quantitative literacy scale is larger than the proportion in this level on the other two scales. So the percentages of adults at each level cannot be averaged to obtain overall performance-level results. Future reports will provide overall literacy composite score results, but not overall performance-level results.

Q12. Does NAAL provide results for all states and counties?

1. The NAAL sample size is not large enough to provide valid literacy estimates for individual states, except for states that funded an increase in their state sample sizes. However, we understand the need that all states and counties have for adult literacy estimates for their jurisdictions, particularly regarding the lowest literacy level. We are currently developing projections of adult literacy for all states and counties, and we plan to post these projections on the NAAL website in the future.
   
   In contrast to the direct estimates produced for NAAL, these projections are based, in part, on demographic data from the 2000 Population Census. The Census demographic information includes variables that are related to the literacy of adults at the lowest literacy levels, such as education, immigration status, race/ethnicity, occupation and employment status, income, and English language capability. A thorough evaluation of the model will be conducted to assess its adequacy, including comparing the model-based projections with direct estimates for the states that paid for an increase in their state sample sizes in 2003

Back to Pre-briefing webcast