The purposes of this study were to: (a) examine the impact of the Silverman and Solmon article (1998) on how researchers handle the unit of analysis issue in their field-based intervention research in physical education in the United States and summarize statistical approaches that have been used to analyze the data, and (b) provide recommendations for future field-based intervention research and related statistical analysis.
We identified and coded 50 peer-reviewed, field-based intervention research articles with a coding template, published in Journal of Teaching in Physical Education and Research Quarterly for Exercise and Sport from 1998 to January 2016.
Our findings showed that 60% of the articles disregarded the unit of analysis and 80% of the articles applied the interventions to classes/groups, but used individual students as unit of analysis. Eight statistical modeling and analysis approaches were used to address the unit of analysis issue.
These findings provide first empirical evidence that the Silverman and Solmon 1998 article had limited impact on how researchers handle unit of analysis in their field-based intervention research in physical education. This suggests that the issue of unit of analysis remains largely unsolved. To address this problem, two experimental designs and corresponding statistical analysis methods were recommended.