Does automatic captioning provide the same access as an interpreter?

Abstract

Interpreter shortages have led some school districts to substitute Automatic Speech Recognition (ASR) captioning technology in place of qualified interpreters for deaf students. This mixed methods case study examined whether Zoom’s embedded automatic captioning provides equivalent educational access to sign language interpreting for ASL-using deaf learners. Four deaf adults with high English literacy participated in a mixed-methods case study comparing comprehension outcomes across two conditions: video-based instruction with a live ASL interpreter and video-based instruction with Zoom’s real-time automatic captions. Comprehension was assessed using pre- and post-tests across six video stimuli representing multiple content areas and K-12 grade levels. Both conditions produced learning gains; however, the interpreting condition yielded substantially greater improvements (mean gain = 41.7%, Cohen’s d = 3.15) compared to the captioning condition (mean gain = 18.7%, d = 1.13), with the between-condition difference reaching statistical significance (p = .024). Participants consistently reported that captions were too rapid, prone to disorienting autocorrections, and imposed unmanageable dual-attention demands. Findings suggest that automatic captioning, even under acoustically controlled conditions, does not provide equivalent access to ASL interpreting for deaf students and may substantially diminish learning opportunities. Policy implications for Free and Appropriate Public Education compliance and educational accommodation planning are discussed.