Longitudinal Data Show Little Gain After the First Post-Test. Only two data bases were found that presented repeated testing of adults who stayed in literacy programs for a year or more. In both data sets, an initial year or so of gain was observed, and then repeated assessments indicated that the rate of gain slowed and became almost level. Students in New York City who were in programs for up to three years gained 2.0 years of skill, from the 1.5 grade level to the 3.5 grade level. A year and a half of that gain was made between the pretest and the first post-test given a year later.

It should be noted that all test scores presented were average scores for programs. Averages conceal the large differences that may occur among the individuals in the programs. Many of the programs may have included adult students who developed quite a bit of skill. Similarly, however, there may have been many students who made little or no gain.

A general finding is that typical program evaluation reports do not provide much information about variance in scores; standard deviations are typically not reported nor are the distributions of scores on pre- and post-tests. Importantly, no studies or reports were found in which concern was expressed about the level of gain reported and what might be done to increase adults' learning as measured by standardized tests during the program. More likely, when low test score gains were reported, the standardized tests were criticized as invalid indicators of what was actually learned in the program. Generally, no alternative indicators of growth in achievement were reported.

Hopefully, the extensive review of literacy assessments presented in this Compendium, with the major findings outlined above, will lead to new approaches both to improving the extent of growth that adults achieve both in and out of adult literacy programs, and to new, cost-effective measures that can inform policymakers about general literacy trends among adults.