心理科學之複製危機與轉機：回顧、因應及對體育研究的啟發

The replication crisis and opportunities of psychological science: Review, remedies and the application in physical education

2011年荷蘭心理學家Diederik Stapel被發現資料造假，打響質疑心理學研究真實性的第一槍。2015年開放科學團隊 (Open Science Collaboration) 發表在《Science》的結果顯示心理學實驗的複製率不達五成，突顯出心理科學知識的不穩定性，並令學界反思進而調整出開放的觀點以及新式統計作法，這也影響了現今對於社會及行為科學的研究方法設計以及資料分析方式，尤其近期於體育研究領域中也獲得高度重視。為更了解這全盤影響，本文將先介紹近期複製危機的相關發展脈絡，之後循序討論在研究設計層次上可能遇到的問題，尤其在統計檢定力與樣本數的估計，以及體育研究領域中的狀況。更進一步的，再介紹新式資料分析如「p值曲線」 (p-curve) 對此之意涵與效用。最後，針對複製與註冊研究，以及開放科學的相關內容予以說明，並且分析此在國內外期刊出版演進，並探討對於體育研究之啟發。

 

In 2011, the Dutch psychologist Diederik Stapel has been revealed for fabricating and manipulating data for his research publications. In 2015, the Open Science Collaboration published an article on “Science” journal and then demonstrated that less than 50% of the classical psychological experiments can be replicated, highlights the instability of the knowledge in psychological science, thereby allowing the academic community to reflect and adjust some new methods in statistical analysis and advocate perspective on open science. Furthermore, this reflection also influences the current social and behavior science on research design and data analytic method, especially in the area of recent physical education. In order to understand the full scope of this impact, the current article will be illustrated such background of the current replication crisis, and then discussed some issues in research design in which focus on power analysis and sample sizes estimation, and explained the current status in physical education. Furthermore, the new statistical analysis such as p-curve will be introduced, and then discussed the Registered Replication Report and open science perspective. Finally, their applications in physical education will be discussed.