Use AI-based text analysis (word embeddings plus classifiers) to screen math assessment items for accessibility before deployment¶

Strategy · All strategies
Evidence · no claims cited

Description¶

This strategy implements the study's AI approach: apply word embedding methods to the text of math assessment items and train classifiers to predict whether items will measure achievement equally well for students with visual impairments. The brief states "The researchers introduced three word embedding methods and three classifiers to predict item quality for accessible assessments," positioning this as "a new way to better present math content online to improve accessibility and increase the opportunity to learn math for students with VI."

Design Implications¶

Context¶

Requirements¶

• Labeled item-quality data from accessible and standard assessment administrations
• Word embedding methods and classifiers trained on item text

Constraints¶

• The brief does not report prediction performance metrics; effectiveness of the screening approach is not quantified here

Target Learners¶

• students with visual impairments who use screen readers

Target Learning Goals¶

• improving accessibility of online math content
• increasing opportunity to learn math

Related Strategies¶

Examples¶

-

Key Sources¶

• Kang Xue, Elizabeth Barker. (2022). Using Artificial Intelligence (AI) to improve math accessibility for students with visual impairments. https://www.nwea.org/resources/using-artificial-intelligence-ai-to-improve-math-accessibility-for-students-with-visual-impairments/