Chapter 8
We now review specific findings regarding adult basic skills development from the two evaluations. According to the CASAS developers, a score of 225 on the CASAS reading test signifies high school level performance. Exhibit 8.4 shows that all of the Even Start evaluation groups performed at or above this level. The exhibit also shows that adults in both the first and second Even Start evaluations (NEIS and ESIS) made pretest-posttest #1 gains on the CASAS reading test. Those gains were as large or larger than the gains observed in other pretest-posttest design studies of adult education programs.
Was Even Start responsible for the gains? One way of judging this is to examine the CASAS reading scores of adults who participated in the random assignment In-Depth Study that was conducted in five projects as part of the first Even Start evaluation. The Even Start IDS group did not gain much from pretest to first posttest (one school year) but made a more substantial gain from the first posttest to the second posttest (the second school year). While the IDS control group started out a few points lower at pretest, their growth rate exactly paralleled the Even Start group. This occurred, in part, because adults in the IDS control group also availed themselves of local adult education programs, leading to the conclusion that while Even Start adults do make gains on the CASAS reading test, we cannot necessarily attribute those gains to Even Start.
Exhibit 8.4: CASAS Reading Growth for Different Even Start Samples
|
Exhibit reads: Even Start adults achieved gains on the CASAS reading test, but so did adults in a control group (as part of the first evaluation). |
Even Start adults also showed a pattern of growth on the CASAS math test, which was used only in the second evaluation. Exhibit 8.5 shows that Even Start adults who entered the program scored almost half a standard deviation below the high school level. After one program year, their scores had increased so that they were able to perform high school level work. While it is an impressive gain, we do not have a control group against which to gauge the progress of Even Start adults. Hence, we cannot unambiguously attribute this growth to Even Start.
Exhibit 8.5: CASAS Math Growth for the ESIS Evaluation Sample
|
Exhibit reads: Even Start adults who took the CASAS math test scored below the high school level at pretest, but their scores increased to the high school level at posttest. No control group data are available for this analysis. |
Exhibit 8.6 presents growth data for Even Start adults on the TABE reading and math tests. Both of these measures were used only in the second evaluation. While adults showed clear gains between the pretest and first posttest of about 3 standard deviation units on each test, we have no control group and are unsure of the extent to which Even Start was responsible for these gains. Further, for the small number of adults with a second posttest, there was a decline in scores on the math test. Interestingly, for those few adults with a second posttest, the reading scores were consistently lower, on average, than the reading scores for adults with only waves one and two, while the math scores were close to the mean scores.
Based on these data we are faced with uncertainty about Even Start's effects on adult literacy. Clearly, adults who participate in Even Start make gains on all of the measures that have been used. Gains in math appear to be larger than gains in reading at posttest #1. However, where data are available on adults not in Even Start, they too make gains, possibly because they too take part in adult education programs.
Exhibit 8.6: TABE Reading and Math Growth for the ESIS Evaluation Samples
|
Exhibit reads: Even Start adults achieved gains on the TABE reading and math tests. No control group data are available for this analysis. |
A final piece of information about adult literacy is presented in Exhibit 8.7, which shows the rate at which adults attained a GED while in Even Start. The two "universe" data collections (from the first and second evaluations) showed that 8 percent and 11 percent of Even Start adults attained a GED over a program year. We also examined the pattern of GED attainment for adults who participated for more than one program year. The proportion increased for each subsequent year of continued participation (although only approximately 40 percent of adults continued into a second program year, and 14 percent continued into a third program year). For those who entered in 1994-95, for example, about 8 percent attained a GED, for those who continued in program year 1995-96, about 11 percent did so, and for those who continued for a third program year, 14 percent attained a GED. The data for those who entered in 1995-96 suggest a similar pattern; approximately 7 percent attained a GED in that year, and 13 percent (of the 35 percent of adults who continued into the 1996-97 program year) attained a GED in 1996-97. The random assignment In-Depth Study from the first evaluation shows that adults in Even Start were far more likely to attain a GED than control group adults (22 percent versus 6 percent).
It seems clear that Even Start does help adults get a GED. The next question is "How helpful is a GED?" There is little evidence that a GED can be equated with any particular level of literacy performance or gains (e.g., the New Chance evaluation conducted by Quint et al., 1994). Recent research by Murnane, Willett, and Boudett (1995) shows that attainment of a GED is better in an economic sense than not having a GED but is not as beneficial as having a high school diploma.
The GED credential is an important focus of many Even Start projects and a goal that seems achievable for many Even Start adults within a reasonable amount of time. As mentioned earlier, approximately half of Even Start adults indicate that attaining a GED is indeed a goal (refer to Exhibit 7.15), based upon the number who are working toward GED attainment, who have taken some portion of the tests, or who have attained it during a particular program year. Recent research on adults' literacy skills found comparable literacy levels when contrasting GED examinees to National Adult Literacy Survey (NALS) respondents. This raises some potential questions about the value of a GED versus more traditional high school completion, because GED attainment is not necessarily perceived as equally credible as a traditional high school diploma by prospective employers (Baldwin et al., 1995). At the same time, there also is recent research that documents the value of the GED as a credential in terms of future employment earnings (Murnane, Willett, and Boudett, 1995). However, there is some research that suggests that GED attainment is unlikely in the short term for first-level adult learners who enter adult education programs with less than a 5th-grade education or equivalent (Stites, Wagner, Foley, and St.Pierre, 1996).
Exhibit 8.7: Rate of GED Attainment for Different Even Start Evaluation Samples
|
Exhibit reads: The IDS Even Start adults attained GEDs at a much faster rate than the control group adults. |
This presents a challenge to many of the Even Start projects because approximately half of adult enrollees enter Even Start having completed less than a 10th-grade education, 20 percent having completed up to the 8th-grade, and 6 percent having completed less than a 6th-grade education. While this debate about the value of a GED will undoubtedly continue, we also know that many Even Start participants enroll for several reasons, including furthering their education (and attaining a GED). Even Start projects provide an array of literacy education services, in addition to GED preparation, to help participating families improve their literacy skills.
-###-
 [Chapter 8: Findings from the Two Evaluations] |  |
 [Chapter 9: Conclusions] |