2014 年一只明成化鬥彩雞缸杯以10.8 億新台幣成為瓷器拍賣市場眾所矚目的逸品，引發筆者針對鬥 彩瓷器在拍賣市場上價格變化進行探究的動機。本研究以佳士得與蘇富比兩家國際拍賣公司於2004 年至 2013 年10 年間拍賣鬥彩瓷器的詳細內容與成交紀錄，整理成樣本資料庫共310 筆，運用統計學方法，加 以歸納、整理及分析影響鬥彩瓷器價格的相關因素，並推論鑑價模式。研究架構分別從鬥彩瓷器整體概 況、歷年成交量、超過預估價格比例及總成交金額探討鬥彩瓷器拍賣市場整體概況，並運用統計學方法 中的多元迴歸分析法（Multiple regression），建立可供參考的鑑價模式。 本研究結果得出鬥彩瓷器拍賣市場十年間在超過預估高價的比例上有走高的趨勢，而當年度成交額 的高低視該年度是否有精良藝術品而定，經過2008 年全球金融風暴低檔，鬥彩瓷器市場從谷底逐年爬升 至2011 年鬥彩瓷器市場總成交額的最高峰。進一步以「拍賣公司」、「拍品年代」、「拍品數量」、「拍品尺 寸」、「拍品款識」、「預估低價」、「預估高價」為自變項，探討影響鬥彩瓷器「成交價」依變項，運用統 計學方法的多元迴歸分析，得出由「預估低價」及「拍品數量」為自變數對於「成交價」有81.1%的解釋 力，判定係數R 帄方為.813，迴歸模型方程式如下： Y=67138.017 3 X （拍品數量）  1.825 6 X （預估低價）  34288.950

This study made use of transaction records, as well as detailed descriptions of the content of the transactions, from the sale by auction of Doucai porcelains by the international auction houses Christies and Sotheby’s over the 10-year period between 2004 and 2013. This information was used to create a database with 310 entries. Then, with the aim of deriving a valuation model, statistical methods were used to categorize, collate, and analyze the factors that affect the price of Doucai porcelains. This study begins by using historical transaction volume data, data on total prices paid, and the percentage by which anticipated prices were exceeded to examine the overall state of the Doucai porcelains auction market. Finally, a multiple regression analysis is used to establish a valuation model that may be used as a reference by interested parties in the future. The analysis shows that, in the Doucai porcelains auction market, the percentage by which anticipated prices were exceeded increased slightly over the ten-year period. The transaction volume depended on the degree of excellency of the porcelains. After the global financial crisis of 2008, the transaction volume of the Doucai porcelain market increased to a peak in 2011. This study used statistical analysis to search for correlations between the auction house, the size of the Doucai porcelain, the period in which the Doucai porcelain was created, the quantity of the Doucai porcelain, the presence or absence of inscriptions on the Doucai porcelain, the reserve price, the anticipated highest bid, and the actual transaction price. The multiple regression model shows that the actual transaction price was affected by 81.1% by two factors, the reserve price and the quantity of the Doucai porcelain. The model equation is as follows: Y=67138.017 3 X （the quantity of the Doucai porcelain）  1.825 6 X （the reserve price）  34288.950