The Bilingual Assessment of Phonological Sensitivity (BAPS): Efficient Assessment with Branching Subtests.

• In Event: Poster Session 04
  In Poster Session: PS 04 Topic - Language, Communication

Abstract

Phonological sensitivity, the ability to manipulate the sound structure of oral language, is a critical skill predicting later reading outcomes for young children, including English-Spanish DLLs (cf. Anthony et al., 2007; Melby-Lervåg, et al., 2012; NELP, 2008; Powell & Atkinson, 2020). Broader understanding of the role of phonological sensitivity for English-Spanish bilingual children has been hindered by a lack of validated assessments that offer direct comparison of skills across languages and are efficient, reliable, and usable by researchers and educational professionals. The goal of this study is to present the current version of the Bilingual Assessment of Phonological Sensitivity (BAPS), a comprehensive assessment of Spanish-English DLLs’ phonological sensitivity. We focus on our branching subtest approach used to improve measurement efficiency by selecting subtests based on scores on previous subtests while simplifying administration and ensuring content balance between the English and Spanish versions by having a limited number of branches.
Method: Children (n=465) ages 3;6 to 5;11 took several subtests of multiple-choice (Matching) and open-ended (Production) items in both language versions of the BAPS. Rasch measurement analyses were performed using Winsteps (Linacre, 2018) to identify items that were effectively measuring phonological sensitivity across the continuum of development. We selected subtests to include on the branching version if: most items on the subtest fit the Rasch model well in both languages, the subtest had similar average item difficulty in both languages, and inclusion of the subtest ensured each branch yielded information about children’s phonological sensitivity equivalent to a reliability of at least .8. We also identified cut points at each branching point to determine which branch a test administration should follow to maximize information about the child’s phonological sensitivity.
Results: Children start with Blending Matching, which involves blending syllables and phonemes to form words. This subtest consists of items effective at assessing phonological sensitivity for a range of children centered near the mean level of phonological sensitivity. Children who score lower on Blending Matching receive Segmenting Syllables Production. Those who score higher on Blending Matching receive Blending Production, followed by either Initial Sound Production or Rhyming Production. For all potential branches, exactly two skills are assessed, minimizing instructions by targeting the test to each individual child. The branching structure of the BAPS is identical for both languages to ensure comparability of content.
Conclusions: The branching version of the BAPS offers an effective way to efficiently measure phonological sensitivity in both English and Spanish. Using a branching approach allows for the measurement of phonological sensitivity with high reliability while minimizing the number of items administered and instructions given. The BAPS can generally be administered in about 10 minutes in each language. Because the BAPS assesses children across the continuum of phonological sensitivity skills, educators can use the test results to identify strengths and individualize instruction as children begin to learn reading.

Authors

• Ryan Bowles, Michigan State University
  Presenting Author

• Sarah Goodwin, Michigan State University
  Non-Presenting Author

• Bethany Keffala, Teachers College, Columbia University
  Non-Presenting Author

• Carol Scheffner Hammer, Columbia University
  Non-Presenting Author

• Shelley Scarpino, Bloomsburg University of Pennsylvania
  Non-Presenting Author