Limitations in sampling and research design in Fast studies have hampe
red understanding of adult education's impact on the acquisition of ba
sic skills. The present study used a selection bias weighting adjustme
nt and ordinary regression techniques in a path analysis approach to i
dentifying variables which directly and indirectly influence reading a
chievement in adult literacy programs. A major focus of thr study was
estimating the effect of student persistence (hours of instruction) on
literacy outcomes in English as a Second Language (ESL), Adult Basic
Education (ABE), and Adult Secondary Education (ASE). Persistence was
found to contribute positively to reading achievement only in ESL; neg
ative persistence effects were actually observed for ABE classroom and
lab instruction. Considering both direct and indirect effects, the da
ta suggest that adult literacy education can generally be improved by
greater investment in full-time staff. It was concluded that an emphas
is on student persistence in ABE and ASE instruction may be misguided.