應用分量迴歸模型分析電影票房之影響因素

Abstract

有別於過去的研究大多採用最小平方法來分析電影相關變數對總票房的影響，由於最小平方法是以平均值為條件的觀察，估計的控制變數，美國電影票房只是「平均值」趨勢的概念，無法觀察到電影票房高低不同時，各變數對其影響所呈現的不同變化。 本研究透過「分量迴歸法」，得到傳統OLS方法估計的迴歸模型沒有辦法得到的實證結果。研究中以電影特徵變數、發行相關變數及評價變數此三類別不同的電影相關變數做為解釋變數，探討其對於不同分量上美國電影票房績效及產品生命週期的影響。研究結果發現，製片預算並非對於所有電影之總票房及首週票房都有正向的結果，反而只有在總票房分量10%及分量90%會呈現正向顯著的影響; 首週螢幕數對於每個分量之首週票房皆有正向影響，尤其對於高分量之首週票房有更加顯著正向效果; 參與評比美國觀眾人數對於中高分量之總票房有正向顯著的結果，另外參與評比之美國觀眾人數會對於中高分量之市場潛量有明顯的正向影響。 最後針對分量迴歸的結果，提出行銷意涵與未來的研究方向。

The OLS model was the most frequently used method to analyze the influential factors to box office of motion pictures in the previous literature, because of the OLS model taked the mean as the condition observation, the American movie box office was only “the mean” concept, it could not observe the different influence between the movie box office and the various variables. This research used the “Quantile Regression” method to obtain the result which the traditional OLS method can’t obtain. There are three category different independent variables in this study including movie characteristic variables, distribution-related variables and evaluated variables. We probed into the independent variables regarding the different quantile in the American movie box office achievements and the product life cycle influence. The study discovered that, the movie budget didn’t have the positive result relating all total box office and the first week box office, instead only the quantile 10% and 90% total box office appeared the significantly influence; the first week screens were positive influence related to the first week box office, especially the high quantile first week box office has a remarkable effect; American audience population were significantly related to the high quantile box office, in addition, American audience population also has obvious influence to high quantile potential market demand. According to the research result, proposed the marketing meaning and the future research direction.