Effect of feedback type and self-selected practice schedule on speech fluency during aphasia script treatment

Abstract

Introduction

Aphasia, particularly in persons with nonfluent aphasia (PWNA), disrupts speech fluency. Script treatment—delivered with recursive self-feedback (RSF) or external feedback (EF)—has been shown to improve fluency in PWNA. Evidence from anomia treatment suggests that structured practice schedule (continuous vs. discontinuous) may further influence outcomes. However, it remains unknown how feedback type and PWNA’s self-selected practice schedules—reflecting the variability of real-life routines—interact to influence language outcomes. This study introduces a novel self-selected practice schedule approach and examines its interaction with feedback type on speech fluency in PWNA.

Method

Using a cross-over design, four PWNA engaged in script production with RSF or EF over two intensive treatment blocks. They practiced for two hours daily across two to three weeks per block. The outcome metrics were speaking rate and speech initiation latency. Flexible practice schedules were allowed, where participants selected between continuous (intensive) or discontinuous (spaced) practice across sessions during each treatment block. Approximately 25,000 speech samples, collected across both treatments’ sessions, were analyzed using linear mixed-effects models.

Results

RSF showed a slower gradual increase in speaking rate over time but greater improvement in speech initiation latency than EF over the course of treatment. The discontinuous practice schedule was associated with significantly greater gains in both outcome metrics. Discontinuous practice schedule particularly influenced improvement in speaking rate during script production with RSF.

Conclusion

By introducing self-selected practice scheduling, this work advances more ecologically valid and personalized models of aphasia intervention. Based on our findings, we hypothesize that a discontinuous, spaced practice schedule may be particularly well-suited for RSF and other errorful-based approaches in promoting language recovery in real-world context. Future research with larger and diverse samples of persons with aphasia should test this hypothesis.