sizes were greater when outcomes were assessed with standardized tests than when evaluated
with experimental measures, which suggests that observed gains in writing performance
following an intervention were not restricted to measures that closely matched the intervention
parameters. Although writing strategy interventions were found to yield rather large gains in
writing performance, they produced weaker effects on students’ writing knowledge, self-efficacy
beliefs, and attitudes about writing (effect sizes ranged from .40 to .64, or small to moderate, on
associated measures). In addition, Gersten and Baker reported that generalization and
maintenance of treatment effects were inconsistent across studies: the majority of students
appeared to have difficulty transferring what they learned to novel situations and the impact of
writing interventions noticeably diminished over time (also see De La Paz, in press; Troia,