The development of a test of overarm throwing: an application of generalizability theory.

Although much effort has been directed toward an understanding of sequential progressions between and within motor skills, little evidence is available to indicate that teachers have incorporated this material into instructional practice. The fact that pupil progress has been evaluated primarily through movement achievement (distance, velocity), rather than through movement content (analysis of form), has been postulated as a reason for this non-use. It was thus the purpose of this study to develop a reliable, objective and practical method of measuring one particular manipulative skill--that of overarm throwing. As such, previously developed sequential progressions were analyzed and modified to produce a measure which did not require the use of a camera. A system of rating throwing performance was adopted which allows for separate rating of the three components: "foot placement", "body rotation", and "arm action". Each component depicts five hierarchically arranged levels, or "steps", and performance is rated on a scale ranging from 1-5. The throwing test was administered to 78 subjects by three trained observers. Reliability analyses, through the use of generalizability theory, showed high objectivity (r's of approximately .90) and good day-to-day reliability (r's of .75 to .93) for most testing conditions. It was concluded that this test is a reliable, valid and practical test for the analysis of overarm throwing in children.