Assessing dyslexia in second language learners by means of a computerized dynamic assessment of decoding ability

Purpose: A dynamic assessment (DA) of decoding has been found to be a promising tool for dyslexia assessment with adult immigrants (Elbro et al., 2012) and for predicting reading difficulties in kindergarten (Gellert & Elbro, 2018). The purpose of this study was to investigate the utility of a computerized extended version of this DA in identifying dyslexic school-age second language learners and predicting development in decoding of students beyond age 10.
Method: DA of decoding was administered along with tests of Danish vocabulary, reading words and nonwords, spelling nonwords, and pseudo‑homophone identification to 231 L2 learners of Danish and 195 native speakers aged 10-25. Teachers pre-classified students with or suspected of dyslexia versus normal readers. A year later, students’ decoding skills were reassessed.
Results: Danish vocabulary correlated weakly with DA of decoding, slightly more with a nonword reading test constructed for L2 learners, and moderately with traditional dyslexia assessments. ROC analyses showed that for L2 students with limited vocabulary, DA results provided a better fit to teachers’ classifications than traditional dyslexia assessments. In combination with the nonword reading test for L2 learners, DA provided an AUC of .86. Follow-up data will be analyzed.
Conclusions: Among students aged 10-25, a computerized DA of decoding was found to be less dependent on vocabulary than traditional dyslexia assessments and provided a better fit to teachers’ reading-related classifications of L2 learners with limited vocabulary. These results indicate that the DA is a useful tool for identifying L2 learners with or at risk of dyslexia.