AI-driven digital accessibility research disproportionately favors visual impairments, neglecting other disabilities.

Why It Matters

This imbalance in research focus leads to the development of tools and systems that may not be universally beneficial, potentially exacerbating existing digital divides. Designers and engineers must recognize this gap to ensure their solutions are inclusive and cater to the full spectrum of human diversity and ability.

Key Finding

The study found that most research on AI for digital accessibility concentrates on visual impairments, with very little attention paid to other disabilities, and that current systems often fail to meet accessibility standards, highlighting a critical need for more inclusive design practices.

Key Findings

• Predominant focus on AI for visual impairments.
• Significant underrepresentation of research for speech, hearing, autism spectrum disorder, neurological, and motor impairments.
• Lack of adherence to accessibility standards in current systems.
• Urgent need for a fundamental shift in designing accessible solutions.
• Accessible AI is vital to prevent exclusion and discrimination.

Research Evidence

Aim: To analyze the current landscape of AI-driven digital accessibility research and identify disparities in focus across different disability types.

Method: Bibliometric analysis and systematic review

Procedure: The researchers conducted a comprehensive search of academic literature, analyzed publication trends, and systematically reviewed existing research on AI and digital accessibility to identify patterns and gaps.

Context: Digital accessibility and artificial intelligence applications

Design Principle

Universal design principles must be applied holistically to AI-driven digital solutions, ensuring equitable access and usability for the widest range of users, irrespective of their abilities.

How to Apply

When initiating a design project involving AI and digital interfaces, conduct a thorough needs assessment that explicitly includes individuals with diverse disabilities, especially those less commonly addressed in current research.

Limitations

The analysis is based on existing published literature, which may not capture all ongoing or unpublished research efforts. The focus on 'AI-driven' accessibility might exclude other important accessibility research.

Student Guide (IB Design Technology)

Simple Explanation: Most studies about AI making digital things easier to use focus on people who can't see well. There's not enough research for people who have trouble with hearing, speaking, or moving, or who are on the autism spectrum. This means current AI tools might not help everyone equally, and we need to design better for all.

Why This Matters: This research highlights that focusing on only one type of user can lead to designs that don't work for many people. For your design project, it's crucial to consider diverse user needs to create truly inclusive and effective solutions.

Critical Thinking: Given the identified research gap, how can designers proactively advocate for and implement accessibility features for underrepresented disability groups in AI-powered digital products, even with limited existing research?

IA-Ready Paragraph: The current landscape of AI-driven digital accessibility research exhibits a significant bias towards visual impairments, with a critical underrepresentation of studies addressing the needs of individuals with speech, hearing, neurological, and motor impairments (Chemnad & Othman, 2024). This imbalance necessitates a deliberate effort in design practice to ensure that technological advancements promote equitable access for all users, moving beyond a narrow focus to embrace universal design principles.

Project Tips

• When researching a design problem, look for studies that cover a wide range of user needs, not just the most common ones.
• Consider how your design could unintentionally exclude certain user groups and proactively find solutions.

How to Use in IA

• Reference this study when discussing the importance of user research and the need to consider diverse user groups in your design process, especially when exploring AI applications.

Examiner Tips

• Demonstrate an awareness of research gaps and biases in the field you are investigating, and explain how your design project aims to address these.

Independent Variable: Disability type (e.g., visual, auditory, motor, speech, neurological, autism spectrum disorder)

Dependent Variable: Volume and focus of AI-driven digital accessibility research

Controlled Variables: Type of publication (e.g., journal articles, conference papers), search databases used, time period of research

Strengths

• Provides a broad overview of the research landscape using quantitative bibliometric methods.
• Systematic review ensures a structured and comprehensive analysis of the literature.

Critical Questions

• What are the underlying reasons for the research bias towards visual impairments in AI accessibility?
• How can the design community actively bridge the identified research gaps to create more inclusive AI solutions?

Extended Essay Application

• An Extended Essay could investigate the ethical implications of biased AI accessibility research and propose a framework for developing more equitable AI solutions for diverse disability needs.

Source

Digital accessibility in the era of artificial intelligence—Bibliometric analysis and systematic review · Frontiers in Artificial Intelligence · 2024 · 10.3389/frai.2024.1349668