Evaluating the Effectiveness of Hybrid and Traditional Speech Therapy for /ɹ/ and /s/ Sound Production

Syracuse University23 enrolled

Overview

The purpose of this research study is to evaluate the effectiveness of a hybrid Artificial Intelligence-assisted speech therapy model, which combines the use of an AI speech therapist (Jessica) with traditional speech therapy sessions led by a Speech-Language Pathologist (SLP). This study aims to determine whether the hybrid model improves articulation in children more effectively than traditional SLP-only therapy. Over the course of 8 weeks, participants will receive weekly therapy sessions, and those in the hybrid therapy group will also practice with Jessica between sessions. The study will also gather feedback from parents and speech therapists to assess their experiences and satisfaction with treatment.

The RCT compares two intervention conditions over eight weeks: (1) a hybrid treatment group, where children receive one weekly "treatment as usual" clinician-led tele-practice session PLUS an additional 60 minutes of home practice with Jessica, and (2) a traditional treatment group, where children receive one weekly "treatment as usual" clinician-led tele-practice session. Children from both groups will be assigned to work on either the /ɹ/ or /s/ sound for the duration of the study, depending on their clinical needs. Children are randomized to conditions and stratified by clinicians. The investigators expect to enroll 100 children, 50 in each condition, aged 5-17 from the following states where study-trained SLPs hold valid clinical licenses: CO, NC, WA, KS, MS, IN, MN, MT, AK, NV, OH, VA, FL, GA, AZ, LA, NY, NJ, MI. Participants must be children aged 5;0 to 17;11 who speak American English as a primary language, have no significant cognitive, neurological, or structural speech impairments, score at or below the 10th percentile on the Goldman-Fristoe Test of Articulation-3 (GFTA 3), have a desire to change their production of /ɹ/ or /s/, and have access to broadband internet for telepractice. Participants with autism, Down syndrome, cerebral palsy, intellectual disabilities, epilepsy, significant hearing loss, ADHD, Tourette's, OCD, or orthodontic appliances that block the roof of the mouth are excluded from this clinical trial phase. The investigators hypothesize that the hybrid treatment group will show greater improvement in speech sound learning compared to the traditional group, as measured by expert perceptual ratings of unpracticed words. Additionally, it is expected that children in the hybrid group will achieve mastery of targeted speech sounds in fewer sessions, making speech therapy more efficient and cost-effective. If successful, this study could provide strong evidence supporting the integration of AI-driven home practice into standard speech therapy protocols, leading to broader accessibility and improved outcomes for children with SSDs. Furthermore, an analysis of treatment fidelity and participant engagement with Jessica will offer insights into the practical implementation of AI-assisted speech therapy in real-world clinical settings.

Interventions

Traditional TherapyBEHAVIORAL

Hybrid therapyBEHAVIORAL

The child would receive standard speech therapy, meeting with the SLP once a week for therapy sessions for 8 weeks. These sessions will focus on articulation practice, where your child will work on speech sounds with direct feedback from the SLP. In addition to these weekly sessions, the child will practice using the AI-based speech therapy tool, Jessica, at home between sessions for 60 minutes per week. The tool will provide real-time feedback on the child's articulation and track their progress. The child will be asked to practice specific speech exercises, such as producing target sounds (e.g., /r/, /s/) in words, phrases, or sentences.

Eligibility

Sex: ALLMin age: 5 YearsMax age: 17 Years

Medical Language ↔ Plain English

Inclusion Criteria: \- Response to be filled out by the Parent after sign up. * Does the child speak American English as one of their primary languages: Yes * At what age did the child first begin to learn English? Age 4 or younger * Child's current age: \> 5 years or \< 17 years 8 months * Does the child have any known history of any of the following: autism spectrum disorder, Down Syndrome, cerebral palsy, intellectual disability, permanent hearing loss, epilepsy/antiepileptic medication, or brain injury/neurosurgery/stroke: NO * Does the child have a diagnosis of ADD/ADHD, Tourette's Syndrome, or OCD? NO * Does the child currently have orthodontic appliances that block the roof of the mouth (e.g., palate expanders)? NO * Does the child currently have a cleft palate, fluency disorder (e.g., stuttering), or voice disorder? NO * Is one of your goals for your child to work on saying the "r" or "s" sounds? YES Exclusion Criteria: * EXCLUSIONARY Response to be filled out by the Parent after sign up. * Does the child speak American English as one of their primary languages: No * At what age did the child first begin to learn English? Age 4 or older * Child's current age: \<5 years or \> 17 years 8 months * Does the child have any known history of any of the following: autism spectrum disorder, Down Syndrome, cerebral palsy, intellectual disability, permanent hearing loss, epilepsy/antiepileptic medication, or brain injury/neurosurgery/stroke: YES * Does the child have a diagnosis of ADD/ADHD, Tourette's Syndrome, or OCD? YES * Does the child currently have orthodontic appliances that block the roof of the mouth (e.g., palate expanders)? YES * Does the child currently have a cleft palate, fluency disorder (e.g., stuttering), or voice disorder? YES * Is one of your goals for your child to work on saying the "r" or "s" sounds? NO

Outcomes

Primary Outcomes

Motor Learning - Percent Improvement on Outcome Words/Phrases

Change scores from baseline to week 8 (i.e., after same amount of clinician treatment). Change scores are calculated from expert listener-averaged perceptual ratings of /ɹ/ or /s/ in pre-defined word lists. These ratings/words are solicited from the Outcome Word List. The ratings are obtained using Perceptual Rating Scale-5 point. Note: The Outcome Word List is not necessarily an untreated word list. Untreated words are ideal; however, it is not possible to ensure that words are untreated in a treatment-as-usual paradigm. At a minimum, this word list will be a measure of "retention and generalization".

Time frame: From enrollment to the end of treatment at 8 weeks

Secondary Outcomes

Time to Mastery of Practiced Words

Group-level percent mastery for in-session clinician ratings of /ɹ/ or /s/ in practiced words from weeks 1-8. The target words are solicited from the in-session practiced words and the percent mastery rating is taken from the SLP session notes. Because SLPs rating of session accuracy could possibly be biased by their knowledge of if the child is working with Jessica or not, it is CRITICAL that SLPs MUST NOT know a participant's group assignment.