Statistical Design/Analysis for Intervention Research on Intact Classes in PE

Friday, April 4, 2014: 10:15 AM
127 (Convention Center)
Ping Xiang, Texas A&M University, College Station, TX and Weidong Li, The Ohio State University, columbus, OH
Background/Purpose: N/A

Method: N/A

Analysis/Results: N/A

Conclusions: N/A

Intervention research is always considered valuable in any field of study, as it attempts to answer the cause and effect issue, an issue critical to advancing our knowledge of any phenomenon of interest. Different from intervention research in the fields such as medicine, exercise physiology and motor neuroscience, intervention research in physical education pedagogy is often characterized by the use of intact classes rather than randomization for the treatment of individual students. Additionally, the treatment is generally implemented at the level of class, but the variables of interest are measured at the student level.  These characteristics are primarily due to the fact that physical education pedagogy research is primarily field research and must be conducted in the “real world” – physical education classes, to inform the practice of physical education in schools. They also require researchers to use appropriate statistical design and units of analysis to ensure the faith that readers can place in the results of such intervention research. As the first presentation of this tutorial, therefore, its purposes were to clarify the concepts of statistical design, an experiment unit, and a unit of analysis in intervention research and describe inappropriate use of them and resulting consequences.

Previous Abstract | Next Abstract >>