Examining school, student, and measurement effects on first grade students' demonstration of the alphabetic principle
Basaraba, Deni Lee, 1981-
MetadataShow full item record
Basaraba, Deni Lee, 1981-
Learning to read and successfully decode words is complex, requiring the integration of critical component skills such as phonological awareness, alphabetic understanding, and phonological recoding. As foundational skills required for reading with automaticity, researchers recommend that explicit instruction of these skills begin early, particularly for students at risk. One commonly used measure to examine students' alphabetic understanding and phonological recoding skills is DIBELS Nonsense Word Fluency (NWF), a pseudo-word reading measure composed of vowel-consonant and consonant-vowel-consonant words. One purpose of this study was to examine the effects of school-level and individual student-level predictors on students' overall performance on NWF in the spring of grade 1 as evidenced by their total Correct Letter Sounds (CLS) and Words Read as Whole Units Correctly (WRWUC) scores. A series of hierarchical linear models were estimated to investigate the contributions of three student-level predictors (English Learner status, fall of grade 1 Phoneme Segmentation Fluency raw scores, and fall of grade 1 NWF scores) and two school-level predictors (percentage of students eligible for free and reduced lunch and percentage of incoming at-risk kindergarteners) in explaining the variance observed in NWF scores A second purpose was to estimate the item difficulties of the first 20 pseudo-words for comparability of difficulty, controlling for student-level covariates. A series of hierarchical generalized linear models were estimated to investigate the contribution of student-level predictors while controlling for school effects. Participants were 1,111 first grade students enrolled in 14 elementary schools participating in the Oregon Reading First initiative. Results indicated that fall of grade 1 NWF raw and quadratic scores were the only statistically significant student-level predictors of CLS and WRWUC scores in the fully specified Level 1 model. The relation between school-level predictors and spring of grade 1 NWF performance complicated interpretation, but both school-level predictors were also significant. Additionally, results of the item difficulty estimates reveal significant student-level effects on item difficulties, providing evidence that item parameters are not equal for the first 20 pseudo-words on DIBELS NWF. The effects were particularly strong for English Learners. Implications for practice and directions for future research are discussed.