Predicting reading ability for bilingual Latino children using dynamic assessment.

Area of Science:

• Educational Psychology
• Developmental Psychology
• Linguistics

Background:

• Bilingual children, particularly those at risk for language impairment, often face challenges in early literacy development.
• Accurate prediction of reading difficulties is crucial for timely and effective intervention.
• Dynamic assessment offers a potential alternative to traditional methods for identifying learning potential.

Purpose of the Study:

• To investigate the predictive validity of a dynamic assessment for identifying future reading difficulties in bilingual Latino children.
• To determine if dynamic assessment measures (gain scores, residuum gain, modifiability) can predict first-grade reading outcomes.
• To evaluate the classification accuracy of the dynamic assessment in this specific population.

Main Methods:

• A cohort of 63 bilingual Latino kindergarteners at risk for language impairment participated.
• Participants completed a dynamic assessment nonsense-word recoding task yielding pretest-posttest gain, residuum gain, and modifiability scores.
• First-grade reading measures included word identification, decoding, and reading fluency.

Main Results:

• The dynamic assessment demonstrated high classification accuracy for predicting first-grade reading abilities.
• Sensitivity and specificity were above 80% for all three criterion reading measures.
• Notably, 100% sensitivity was achieved for two of the three first-grade reading measures.

Conclusions:

• The dynamic assessment shows significant promise for predicting word-level reading ability in bilingual Latino children.
• This approach can aid in the early identification of children at risk for reading difficulties.
• The findings support the utility of dynamic assessment in diverse early literacy contexts.