Reliability and Structural and Construct Validity of the Functional Strength Measurement in Children Aged 4 to 10 Years

Background. Adequate muscle strength, power, and endurance are important in children's daily activities and sports. Various instruments have been developed for the assessment of muscle function; each measures different aspects. The Functional Strength Measurement (FSM) was developed to measure performance in activities in which strength is required. Objective. The study objective was to establish the test-retest reliability and structural and construct validity of the FSM. Design. A cross-sectional descriptive study was conducted. Methods. The performance of 474 children with typical development on the FSM was examined. Test-retest reliability (n=47) was calculated with the intraclass correlation coefficient (2.1A) for agreement. Structural validity was examined with exploratory factor analysis, and internal consistency was established with the Cronbach alpha. Construct validity was determined by calculating correlations between FSM scores and scores obtained with a handheld dynamometer (HHD) (n=252) (convergent validity) and between FSM scores and scores on the Movement Assessment Battery for Children-2 (MABC-2) (n=77) (discriminant validity). Results. The test-retest reliability of the FSM total score ranged from .91 to .94. The structural validity revealed one dimension, containing all 8 FSM items. The Cronbach alpha was .74. The convergent validity with the HHD ranged from .42 to .74. The discriminant validity with MABC-2 items revealed correlations that were generally lower than .39, and most of the correlations were not significant. Exploratory factor analysis of a combined data set (FSM, HHD, and, MABC-2; n=77) revealed 2 factors: muscle strength/power and muscle endurance with an agility component. Limitations. Discriminant validity was measured only in children aged 4 to 6 years. Conclusions. The FSM, a norm-referenced test for measuring functional strength in children aged 4 to 10 years, has good test-retest reliability and good construct validity.