Measuring Activity Level with Actometers: Reliability, Validity, and Arm Length

EATON, WARREN O. Measuring Activity Level with Actometers: Reliability, Validity, and Arm Length. CHILD DEVELOPMENT, 1983, 54,720-726. The gross motor activity of 27 3and 4-yearolds was measured using teacher ratings, parent responses to the activity scale of the Colorado Childhood Temperament Inventory, and 13 uncalibrated actometers worn singly on the wrist over a period of 1 month in a free play setting. Activity scores from single actometers had low but meaningful levels of reliability; activity scores composited across multiple actometers had high reliability and correlated strongly with parent and teacher measures, even when age and sex effects were partialled. The possibility that limb length is a confounding factor of wrist-placed actometers was also evaluated, and though a nonsignificant trend in this direction was found, validity correlations were essentially unchanged when arm-length effects were removed. The limb-length confound does not appear serious, and when used appropriately actometers provide reliable and valid information on individual differences in children's gross motor activity levels.