An Examination of the Predictive Validity of Early Literacy Measures for Korean English Language Learners
- Author(s): Nam, Jeanie Eunjoo
- Advisor(s): Vanderwood, Michael L
- et al.
The primary aim of this study was to examine the predictive validity of early literacy measures with Korean English language learners (ELLs) representing varying levels of English language proficiency. First-grade Korean ELLs (N = 102) were screened in the winter using measures of Phoneme Segmentation Fluency (PSF), Nonsense Word Fluency (NWF), and Word Identification Fluency (WIF). Spring reading criterion measures included Oral Reading Fluency (ORF) and the Woodcock Reading Mastery Tests (WRMT). Among the winter screening measures, WIF was found to be most correlated with end-of-the year reading outcomes for the larger sample, as well for students aggregated by language proficiency groups. While moderate to high correlations were also found for winter NWF with respect to spring outcomes, correlations were much smaller in magnitude for winter PSF. Results of hierarchical regression models, with PSF, NWF, and WIF entered respectively, also found that the addition of WIF resulted in a large and significant change in R2 with respect to both spring ORF and WRMT scores. Final models accounted for a total of 75.6% of the variance in spring ORF scores and 60.4% of the variance in WRMT scores. Furthermore, of the winter screening measures, only winter WIF was found to be significant at the p < 0.05 level in both of the final models. Finally, an examination of the diagnostic accuracy of PSF and NWF at-risk cutoffs established by the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) yielded negative predictive values at or above 85% for all tests with respect to spring outcomes; however, positive predictive values were significantly lower, ranging from 17% to 47%. Although past studies have found that many of the same assessments tools that have been used with native English speakers can be used with ELLs, the results of this study indicate that there is a need for the continued development of decision rules and appropriate assessment tools that are sensitive to the unique cultural and linguistic backgrounds of ELLs.