Answers are grounded in this paper and cite their sources.
Background Text-to-speech (TTS) technology supports reading confidence and speed for people with aphasia (PWA) but does not universally improve comprehension. This may reflect selection of a less-than-optimal TTS output rate. To evaluate this, researchers need procedures to identify a rate best matching reading capability. This study’s purpose was evaluating whether online eye-tracking procedures are effective for determining a personally optimized TTS rate and examining reading speed, comprehension, and efficiency as variables for the selection process.Methods Three PWA and three neurotypical adults read passages in four conditions. Subjective, online judgment about eye fixations and TTS word production synchronicity was compared with retrospective, objective analysis using eye-tracking data. Close modality synchronization combined with comprehension accuracy and comfort/acceptability ratings guided personally optimized TTS rate selection.Results Determining a personally optimized TTS rate required 6–20 trials per participant. Affirmative close synchronization judgments usually aligned with objective measurements exceeding 50% of fixated words. TTS system limitations prevented using habitual reading speed as a starting point for determining an optimized rate for two participants. Reading efficiency was advantageous for identifying optimized rates for neurotypical adults. Examining reading speed and comprehension accuracy separately was better for PWA because substantial differences between optimized and default TTS rates obscured comprehension accuracy changes.Conclusions Findings substantiated procedural feasibility with noted modifications. Online monitoring was adequate for judging modality synchronization and determining personally optimized rates when combined with comfort/acceptability ratings and comprehension scores. Reading speed and comprehension variables were superior to reading efficiency for determining optimized TTS rates for PWA.
No takes yet. Share an insight, caveat, or question.
Knollman-Porter et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: