AMI, as part of a Small Business Innovation Research (SBIR) grant from the Institute of Education Sciences of the U.S. Department of Education is developing and commercializing SkillCheck, a new innovative, artificial intelligence (AI) based, analytical technology for reading. Designed for students in grades 1 through 4, SkillCheck is an automated diagnostic system of oral reading fluency (ORF) performance and an intelligent practice tool to improve foundational reading skills to achieve fluent oral reading. This Phase 2 project follows the successful completion of a R&D feasibility study funded by a SBIR grant where AMI demonstrated the technical feasibility of the product. See the press release here.
In short, SkillCheck will provide better diagnostic information for teachers to help address the skills students need to become fluent readers without additional testing. For students, SkillCheck will provide engaging practice opportunities to enable them to become confident, fluent readers.
The goal of SkillCheck is to provide teachers with comprehensive data about a student’s reading ability without the need for retesting. With 10-15% of students in Grades 1-4 reading below grade level, the burden of further testing on teachers is problematic. SkillCheck will automatically analyze a student’s reading skills from recordings of the passage read alouds to guide intervention.
How does it work?
SkillCheck uses spoken language and natural language processing to efficiently analyze ORF performances and diagnose readers’ strengths and weaknesses in foundational skills such as sight words, decoding/phonics, and miscues. Initially, AMI will implement SkillCheck into the benchmark module of AMI’s flagship product for oral reading fluency assessment, Moby.Read. It is anticipated that SkillCheck will be ready for more general use in the 2023/2024 school year.
SkillCheck focuses on six broad areas of reading skills: Reading Readiness, Sight Words, Decoding Skills/Phonics, Attention to Text (i.e. Miscue Analysis), Rhythm & Punctuation. Detailed data from each of these categories provides insight into specific areas a student may need remedial work in. For example, a detailed analysis of which decoding skills (e.g., cvc words, closed syllables, etc.) a reader may have difficulty with, along with the corresponding audio links etc enables a teacher to easily individualize instruction without additional testing. SkillCheck will also include a sequence of practice packets that are targeted for the skills a reader needs practice with to attain fluency in oral reading.
After diagnostic results are available, the intelligent practice tool part of SkillCheck presents a sequence of practice packets that are targeted for the skills individual readers need practice with to attain fluency in oral reading.
AMI is collaborating with WestEd, a national leader in educational research, development, and service, on the development of SkillCheck. As a research partner, WestEd will provide content validation, conduct focus groups, pilot content and conduct user testing.
Bernstein, J. (upcoming: January 2022). Machine learning methods behind AI applications in Education. Annual IES Principal Investigators Meeting: Advancing Education and Inclusion in the Education Sciences.
Bernstein, J., Suzuki, M., Cheng, J., Yang C., Ciancio, D., Brenner, D. (2020). Automated Estimation of Foundational Reading Skills from Recorded Passage Read-Alouds. California Educational Research Association (CERA) 99th Annual Conference, Anaheim, CA. View Slides