The motor learning strategy instrument: interrater reliability within usual and virtual reality physical therapy interventions.

PURPOSE: To evaluate and compare the interrater reliability of the Motor Learning Strategy Rating Instrument (MLSRI) within usual and virtual reality (VR) interventions for children with acquired brain injury. METHODS: Two intervention sessions for each of 11 children (total, 22) were videotaped; sessions were provided by 4 physical therapists. Videotapes were divided into usual and VR components and rated by 2 observers using the MLSRI. A generalizability theory approach was used to determine interrater reliability for each intervention. RESULTS: Interrater reliability for usual interventions was high for the MLSRI total score (g-coefficient, 0.81), whereas it was low for the VR total score (g-coefficient, 0.28); MLSRI category g-coefficients varied from 0.35 to 0.65 for usual and from 0.17 to 0.72 for VR interventions. CONCLUSION: Adequate reliability was achieved within ratings of usual interventions; however, challenges related to MLSRI use to rate VR-based interventions require further evaluation.