以結構方程模型的統計檢定力計算及樣本數決定運用於臺灣運動管理學報的正確性之研究

On the Application of Statistical Power Calculation and Sample Size Determination of SEM in the Journal of Taiwan Society for Sport Management

結構方程模式（SEM）有逐年越來越多被應用的趨勢。在統計檢定方面，仍需加強報導及檢測能力。本研究藉由Mac Callum（1996）等人RMSEA估計法計算統計檢定力與樣本數。運用R語言程式，分析「臺灣體育運動管理學報」有關於SEM論文的應用概況及統計檢定力與樣本數決定。研究結果發現SEM的應用大部分的模式統計檢定力與樣本數還算理想，但仍需有加強。文末針對本次研究發現，提供未來研究在研究設計、SEM統計檢定力計算及樣本數決定運用程序等的實用建議，以作為體育領域學者應用SEM研究的參考。

Structural Equation Model (SEM) have experienced increased research applications over the past few years. With regard to statistical testing, there is some room for improvement in both reporting and testing. This research adopted Root Mean Square Error of Approximation (RMSEA), developed by Mac Callum et al. (1996), for the calculation of statistical power and sample size, and used R programming language to analyze how SEM has been employed, as well as the statistical power and sample size, in the research presented in the Journal of Taiwan Society for Sport Management. The findings showed that the application of statistical power calculation and sample size determination of SEM in the journal is arguably satisfactory, albeit there is still some room for improvement. Based on the findings,suggestions regarding research design and the application of statistical power calculation and sample size determination of SEM were proposed in the conclusion as a reference for scholars in the sports field.