ASL CLeaR: STEM Education Tools for Deaf Students
Jeanne Reis, Erin T. Solovey, Jon Henner, Kathleen Johnson, Robert Hoffmeister · 2015 · ASSETS '15: Proceedings of the 17th International ACM SIGACCESS Conference on Computers & Accessibility · doi:10.1145/2700648.2811343
Summary
This demonstration paper introduces ASL CLeaR (American Sign Language STEM Concept Learning Resource), a free online educational platform designed to address a critical gap in STEM education for deaf and hard of hearing students. The authors identify a fundamental problem: deaf children often experience delayed first language acquisition due to limited early exposure to sign language, which cascades into educational disadvantages. Compounding this, many STEM concepts lack standardized ASL signs, forcing students to learn different signs for the same concept across classrooms—imagine teachers using different terms for "photosynthesis" without explanation. Signs created ad-hoc by non-native signers are frequently grammatically incorrect, resulting in learning materials that are neither comprehensible nor conceptually accurate. ASL CLeaR addresses these challenges through three integrated components. Topic Labs bring together Deaf native ASL signers who are also STEM and education experts to develop curriculum-aligned content. These experts create and film micro-lectures and dictionary entries, ensuring grammatical correctness and conceptual accuracy. Each term includes a stand-alone sign, an ASL definition, and usage examples, all tagged with linguistic metadata. The Vocabulary Evaluation Tool (VET) enables nationwide evaluation of proposed STEM signs by L1 ASL domain experts, using an entirely ASL-based interface to prevent reliance on English labels. A small pilot with five evaluators found 16 of 20 terms met or exceeded acceptance benchmarks. The third component is a novel ASL-based search function that allows users to find signs by selecting visual features (handshapes, movements) rather than English text—analogous to how alphabetic search works for spoken languages.
Key findings
The paper provides a comparative analysis of seven existing U.S. online ASL STEM resources, revealing significant gaps that ASL CLeaR addresses. No existing resource offers ASL-based navigation and search combined with native ASL expert vetting, contextual usage examples, and human signers. While some platforms provide signed definitions, many impose English syntax on ASL vocabulary rather than using grammatically correct ASL. DeafTEC, for example, offers signed STEM definitions but these do not consistently follow ASL grammatical rules. Another platform provides ASL-based search but its sign corpus has not been vetted by native Deaf STEM experts. The pilot evaluation demonstrates promise for vocabulary acceptance, with 80% of tested terms (16/20) achieving benchmark scores (M≥4 on the evaluation scale) from native ASL STEM experts. The paper emphasizes that providing STEM terms in at least three ASL contexts is essential to prevent the denotation/connotation confusion that arises from single English word labeling—for instance, the sign for "table" could mean furniture, a chart, or a level (as in water table). This contextual approach is particularly important for early language-deprived deaf children who cannot rely on contextual knowledge to disambiguate meanings.
Relevance
This work highlights a critical but often overlooked accessibility challenge: the lack of standardized, high-quality educational content in native sign languages for STEM subjects. For accessibility practitioners, the paper demonstrates that true accessibility for deaf learners goes far beyond captioning or translation—it requires content created by and for native signers, with attention to linguistic accuracy and cultural appropriateness. The visuocentric design principles and "deaf geographies" approach to web design offer valuable lessons for creating interfaces that respect visual language users rather than forcing hearing-centric interaction patterns. The emphasis on L1 expert involvement throughout content creation and evaluation provides a model for participatory design in deaf education technology.
Tags: deaf · hard of hearing · American Sign Language · STEM education · educational technology · bilingual education · vocabulary learning · visual language