| 中文摘要 |
目的：本研究目的在瞭解臺灣體育與運動學術研究中，應用題項包裹法於結構方程模式分析中的現況，並探討可能衍生的相關爭議性問題，進而提出相關建議供國內相關領域研究者參考。方法：本研究採用內容分析法，針對2008～2012年間國內體育學門一級期刊進行分析，包含大專體育學刊、體育學報、臺灣體育運動管理學報，以及臺灣運動心理學報四種學術期刊。結果：研究結果顯示，在45篇探討潛在變項路徑關係之研究中，33篇進一步採用題項包裹方法，其採用的包裹法均為內部一致性方法；然而僅有24.2％文章提供使用包裹法之理由或依據；在單維結構假設部分，48.5％的研究在其研究設定外，另有採用統計程序加以檢驗；而有21.2％報告題項包裹之信度值；11（33.3％）篇研究執行並列出題項包裹之常態性檢定結果。此外，研究者亦發現，部分研究者在其研究中曾利用事後模式修正的方式，對研究之理論假設模式進行探索導向的模式校正。結論：2008～2012年間，國內採用題項包裹法的研究數量比例偏高，然而在相關概念依據及檢定程序上仍有未盡之處，爰此，本研究在整合相關研究之建議後，進一步提出一可行之題項包裹執行過程的檢驗程序架構。Purpose: The aim of this study was to examine the controversial practice of using parcels of items as manifest variables in structural equation modeling (SEM) procedures for physical education and sport research. Methods: Content analysis was applied to analyze research data that collected from top four prestigious Chinese journals published from Taiwan for the period of 2008~2012, including Journal of Physical Education in Higher Education, Physical Education Journal, Journal of TASSM and Bulletin of Sport and Exercise Psychology of Taiwan. Results: Thirty-three of 45 research articles with SEM that created item parcels in the test of model structural relationships between the latent variables and internal consistency approach was the most common method for parceling. Only 24.2 % papers reported the reason of using parcels, 48.5 % studies explicitly tested the dimensionality of the full item set, 21.2 % articles presented the reliability of parcels, and 11 (33.3 %) studies demonstrated the test of parceling normality in each research. In addition, some research used post modification indices to modify their proposed models. Conclusions: During 2008~2012, many research utilized item parceling in data analysis procedure, but lacked for reporting rational evidences to support their actions. Therefore, based on our findings of this study, the researchers provided a framework and recommendations for guiding researchers for improved use of parcels in SEM. |
本站僅提供期刊文獻檢索。
