Purpose: Educators need high-quality literacy assessments to identify how to best support student in learning to read aloud, however gathering assessment data can be challenging. Because computers have struggled to process speech with human accuracy, schools often measure oral reading fluency through time-consuming teacher-administered assessments or multiple-choice tests with questionable construct validity. In previous projects supported through ED/IES SBIR awards and other sources, the team developed an assessment to measure the aspects of reading, including rapid naming, phonological awareness, phonics, prosody, and comprehension while reading aloud. The project team will develop a prototype of an automated formative assessment software to enable low-cost formative assessment across aspects of reading development. Project Activities: In Phase I, the project team will develop a prototype of an automated oral reading fluency assessment that employs speech recognition software to assess oral reading. At the end of Phase I, a pilot study will be conducted with 8 Kindergarten to grade 3 educators as well as a literacy coach who works with each educator. Students in these classes will use the prototype for automated oral reading fluency assessments. Literacy coaches will support teachers in implementation, data interpretation, and in instructional adjustments using the results from the automated assessment. At the conclusion of the implementation period, educators will the rate the usability and feasibility of the assessment. Analyses will also determine the accuracy and reliability of the automated scoring relative to human scoring.
