Effects of AI-Driven Real-Time Captioning as an Orthographic Scaffold for Literacy Acquisition Among Deaf Learners in Plateau State, Nigeria

Abstract

The literacy gap between deaf and hard-of-hearing (DHH) learners and hearing peers remains crucial in resource-constrained settings like Plateau State, Nigeria. Although human sign language interpretation gives conceptual access, DHH learners sometimes fail to develop direct orthographic representations of difficult English vocabulary due to fast fingerspelling and visual fatigue. This study examined AI-driven real-time captioning as a collaborative orthographic scaffold alongside human sign language interpreters. Using a quantitative quasi-experimental pretest-posttest control group design with delayed retention, 30 Primary Five pupils at Plateau State School for the Deaf, Bassa, were cluster-randomised into two intact groups: Experimental (AI captioning + sign interpretation; n=15) and Control (sign interpretation only; n=15). The eight-week intervention includes pretest, immediate posttest, and four-week delayed retention evaluation. Data were obtained using a validated 40-item English Literacy and Vocabulary Assessment (ELVA) and evaluated with independent-samples t-tests and ANCOVA. The Experimental Group attained considerably better posttest scores (M=74.20, SD=6.85) than the Control Group (M=58.40, SD=7.92), t(28)=5.84, p< 0.001, Cohen’s d=2.13. ANCOVA confirmed a substantial intervention effect, F(1,27)=38.65, p< 0.001, partial η²= 0.589, adjusting for pretest scores. Delayed retention was likewise considerably higher for the Experimental Group (M=71.80, SD=7.10) than the Control Group (M=52.60, SD=8.15), t(28)≈6.88, p< 0.001, Cohen’s d≈2.51. Technical evaluation recorded an average Word Error Rate of 8.4% employing edge-computing hardware and FM lapel mics, circumventing internet instability. Findings imply simultaneous visual-verbal captioning and visual-nonverbal signing increase orthographic-lexical mapping without replacing sign language. Recommendations address policy, edge-hardware adoption, and teacher ICT competence training in Nigerian special schools.