Sponsors page...
A longitudinal study of the effects of the SpeechEasy device in naturalistic environments


Ryan Pollard, J.B. Ellis, P.R. Ramig, D. Finan


Purpose: The SpeechEasy is a prosthetic altered auditory feedback (AAF) device designed to improve fluency in persons who stutter. The purpose of this project was to measure and describe the effects of the SpeechEasy over time in authentic settings and during activities of daily living.

Method: A multiple single-subject, ABA, repeated measures design was used. 11 adult subjects received custom-fit ITC devices to wear for 4 months. SpeechEasy fitting protocols were followed for establishing optimal settings and providing training to optimize the effects of AAF. After a 1-month baseline period, subjects wore SpeechEasy devices for 4 months, followed by a 1-month withdrawal period when the device was removed. Speech samples were taken every 2 weeks during 3 conditions: oral reading, conversation, asking a question to a stranger. This allowed for a total of 15 samples to be collected over 6 months for each subject. Percent stuttered syllables (%SS) scores were calculated and significant fluency improvement was defined as a %SS score falling at least 2 SD below the mean of that subject's baseline. Several questionnaires were administered for pre/post-treatment comparisons. Qualitative data was gathered through weekly written logs, exit interviews, and speech naturalness ratings.

Results: Currently, 5 subjects have completed the study, 4 are nearing the end of the study, and 2 subjects chose to drop out. For all subjects but one (S3), the SpeechEasy appears to provide some level of sustained benefit in at least one speaking condition. Variable results have been found across speaking conditions, as the reading condition (63% of samples show significant improvement) has shown more improvement than the conversation (43%) and question (27%) conditions. Group data thus far reveals a significant (p<.05) treatment effect for the reading condition, but not for the conversation or question conditions. An unexpected finding has been that discrepancies may exist between fluency counts and subjects' perceptions of the device's effectiveness. For instance, some subjects (S3, S8) experienced almost no measurable fluency improvement during formulated speech, but nonetheless reported a positive experience from wearing the device. Alternately, some subjects (S4, S11) showed significant fluency enhancement but reported that they did not like the device. Overall, we report highly variable results across subjects. 7/10 (70%) of our subjects report that they would continue to wear the device if cost was not a concern, while 4/10 (40%) plan to buy/have bought the device at a substantially reduced price after completing the project. Qualitative analysis of the weekly logs and exit interviews reveals several emerging trends. The most commonly reported advantages are: 1) speaking on the telephone, 2) increase in confidence. The most commonly reported drawbacks are: 1) background noise, 2) some subjects report the device does not help them.

Discussion: Our results do not support previous research reporting group stuttering reductions of 75%-85% under AAF (e.g., Stuart et al., 2006). However, our data suggests that the SpeechEasy is effective at reducing stuttering and/or increasing confidence for some who stutter, and thus has merit as a clinical tool for the management of stuttering.