AI in Assessment for Students with Disabilities

Academic Insights

• AI can improve accessibility in large-scale STEM assessments of disabled students.
• Risks include bias, construct distortion, and over-automation.
• Fair assessment requires human oversight at all stages.
• AI should support accommodation, not redefine what is being assessed.
• Ethical alignment with disability standards is essential.

Apply This Now

Review whether AI tools change the construct being measured.

Add This in Your Lesson

Provide multiple ways for students to demonstrate understanding.

Avoid This Mistake

Assuming AI-based accommodations are automatically fair.

Keywords

Disability inclusion; STEM assessment; AI fairness; Accessibility