The Importance of Practice in Reading. Several assessments of adult literacy skills and practices (e.g., reading books, magazines, newspapers) show that years of education, amount of practice, and increased skill go together. Highly literate adults have more education and engage in more reading practices. An important implication for adult literacy programs is that efforts should be made to get adults to engage in larger amounts of reading in various types of literacy practices both inside and outside of programs. Extensive free-reading practice may be as important as direct instruction in producing higher levels of literacy.

Part I: Assessing Adult Literacy Skills

Similarity of Literacy Items Across Time. To a large extent, the tasks found on literacy assessments operationally define what various experts, advisors, teachers of adults, and test developers think about the nature of literacy. Despite the considerable debates about the suitability of different types of literacy tasks that are assessed in national surveys, such as whether to use multiple-choice or constructed response items, "academic" or "real world" items, etc., when the actual items are examined that various advisory boards have approved over several decades, there appears to be a remarkable similarity among items over the last 75 years.

Similarity of Findings Across Time. Beginning with the World War I assessments in 1917, a number of trends have remained salient across time.

1. Higher educated adults perform better than less educated adults.
   
   
2. Younger adults perform better than older adults.
   
   
3. Adults in western, eastern, and northern states perform better than those in southern states.
   
   
4. Whites perform better than African-Americans or non-native language speakers.
   
   
5. Adults in professional, managerial, and clerical occupations perform better than adults in laborer, agricultural, and other relatively unskilled occupations.
   
   
6. Higher income groups perform better than groups with lower incomes or on welfare.
   
   
7. Higher educated adults engage in a greater amount of reading of books, magazines, and news papers.

An implication of these findings for policymakers and researchers is that any number of literacy assessments can be used to rank order the adult population with regard to verbal "intelligence," "aptitude," or "literacy" and the same general trends can be followed. This suggests that, if the concern is simply to identify adults who are high, medium or low in literacy skills, consideration should be given to research for identifying the most cost-effective methods of assessing adult literacy skills. For instance, the very simple vocabulary and paragraph comprehension tests of the Armed Forces Qualification Test (AFQT) produce about the same distribution of adults in five literacy levels as does the more extensive (and expensive) National Adult Literacy Survey (NALS).