Improving health literacy: a Web application for evaluating text-to-speech engines

Area of Science:

• Health Informatics
• Human-Computer Interaction
• Digital Health Literacy

Background:

• The internet is a primary source for health information, but accessibility barriers exist for individuals with lower health literacy.
• Text-to-speech (TTS) technology offers a potential solution by supplementing web content with audio, creating a virtual surrogate reader.
• Objective quality assessment of various TTS engines is challenging, hindering informed selection for patient-oriented web applications.

Purpose of the Study:

• To develop and pilot an open-source web application for evaluating TTS engine performance.
• To objectively measure user reaction times, subjective quality ratings, and task accuracy across different TTS-generated audio files.
• To facilitate informed comparisons and selection of TTS engines for improving health information accessibility.

Main Methods:

• Developed an open-source web application to assess TTS engine quality.
• Collected data on user reaction times, subjective quality ratings, and task completion accuracy.
• Compared audio files generated by three TTS engines with varying price points.

Main Results:

• Significant differences were observed in subjective quality ratings among the evaluated TTS engines.
• No significant differences were found in user reaction times or task accuracy between the TTS engines.
• The developed web application was successfully piloted, demonstrating its utility in TTS engine evaluation.

Conclusions:

• TTS engines differ in perceived quality, which is a crucial factor for health information accessibility.
• Current TTS technology, despite quality variations, may not significantly impact basic task performance or reaction times.
• Future research should explore complex tasks, usability, and applications in health promotion for vulnerable populations.