Date of Award


Document Type

Scholarly Project

Degree Name

Master of Physical Therapy (MPT)


Physical Therapy

First Advisor

Peggy Mohr




Background and Purpose: Throwing is a meaningful play activity that encourages social interaction, develops the sequencing of motor skills, and signals the onset oftoddlerhood. Little published data clearly documents the three-dimensional kinematics ofthe progression of the overarm throwing motion in children. The purpose of this study is twofold: 1) To explore the biomechanical differences in the overarm throwing techniques of children between the ages of two and seven, and 2) To compare the kinematics of dominant versus non-dominant arm throws in children between the ages of two and seven.

Methods: 1) Nine children, 6 males and 3 females, were classified into 3 age groups (2-3 years, 4-5 years, 6-7 years) and then videotaped while performing 6 overarm throws, 3 with each arm. 2) The Peak Motus Software was used to digitize and analyze the reflective markers on the resultant video and create data sets for each child. Five variables were analyzed: tyPe of throw, time of throw, ball velocity, maximum shoulder abduction, and elbow flexion at release.

Results: 1) Older children consistently used the more mature 'dynamic' and 'sequentially-linked' throwing techniques, while the younger children were variable in their approaches. 2) The 6-7 year-olds took the longest time to throw and had the greatest time difference between arms (0.23 sec). 3) Ball velocity increased with age, with the greatest mean difference between arms (2.68 mls) occurring in the 6-7 year-olds. 4) The same group also averaged the largest amount of shoulder abduction (dominant = 60.8° Inon-dominant = 70.9°). 5) The 4-5 year-olds demonstrated the greatest amount of elbow flexion at ix release, and the largest mean difference (8.6°) between arms.

Conclusion: In general, arm dominance seems to begin affecting the quality of throw between the ages of 5 and 6. Studies done with large numbers of subjects and equal sample sizes are needed to obtain statistically significant results which can be generalized to the normal population.