Rater reliability and construct validity of PostureScreen

Kimberly A. Szucs, PhD, OTR/L1,* and Elena V. Donoso Brown, PhD, OTR/L1Department of Occupational Therapy, Duquesne University: 600 Forbes Avenue, Pittsburgh, PA, USA 

[Purpose] Measurement of posture is important for those with a clinical diagnosis as well as researchers aiming to understand the impact of faulty postures on the development of musculoskeletal disorders. A reliable, cost-effective and low tech posture measure may be beneficial for research and clinical applications. The purpose of this study was to determine rater reliability and construct validity of a posture screening mobile application in healthy young adults.

[Subjects and Methods] Pictures of subjects were taken in three standing positions. Two raters independently digitized the static standing posture image twice. The app calculated posture variables, including sagittal and coronal plane translations and angulations. Intra- and inter-rater reliability were calculated using the appropriate ICC models for complete agreement. Construct validity was determined through comparison of known groups using repeated measures ANOVA.

[Results] Intra-rater reliability ranged from 0.71 to 0.99. Inter-rater reliability was good to excellent for all translations. ICCs were stronger for translations versus angulations. The construct validity analysis found that the app was able to detect the change in the four variables selected. [Conclusion] The posture mobile application has demonstrated strong rater reliability and preliminary evidence of construct validity. This application may have utility in clinical and research settings.
Keywords: Reliability, Construct validity, Posture
Link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5788770/