This study involved a dynamic assessment (DA) of word attack (decoding) skills. DA is a form of evaluation in which the tester interacts with the participant to affect change in the skills being evaluated. The purpose of the study was to determine the usefulness of this DA for the evaluation of reading disability and the utility of this DA for teaching students strategies to improve their word attack skills.

Twenty-four students drawn from high school, middle school and an adult literacy program participated in the study. Eligibility for participation was based on below average performance (<25^th> percentile) on the Word Attack subtest of the Woodcock Reading Mastery Tests-Revised (Woodcock, 1998). Participants were randomly assigned to either the DA or Practice Only groups.

The DA involved teaching students a single syllable and a multi-syllable strategy for decoding unfamiliar words. Both strategies involved instructing participants to segment spellings of pseudowords, to pronounce the segments, and to blend the segments to pronounce each pseudoword. Increasing feedback was provided until the participant was able to read the pseudoword. Participants were evaluated in terms of their ability to read the list of pseudowords and to utilize the two decoding strategies. Practice Only group participants were given the DA pseudoword list to read, but they were not provided with strategy instruction or feedback.

The results support the value of the DA as an evaluation instrument. Students' performance on the DA was significantly correlated with the pretest pseudoword decoding task, indicating that inclusion of the strategy instruction and feedback did not undermine the test's validity. The DA provided a more comprehensive assessment of participants' strengths and weaknesses than traditional static assessment. It also provided information regarding the type and amount of instruction that was effective for individual participants.

The DA proved to be an effective means of word attack instruction. The DA group decoded significantly more pseudowords on the immediate posttest than the Practice Only group. However, the effectiveness of instruction was not sustained over time as evidenced by no significant difference in decoding performance of the groups on the delayed posttest administered two weeks later.