選舉研究：重新計票之理論探討

台灣2004總統大選，無論過程亦或結果均引起台灣社會廣大的爭議。對於兩方候選人得票差距僅 2 萬 9 千多票，在統計上，這差距是否可以解釋為顯著差異？本文提出重新驗票數學模型，此模型為考慮有廢票情況的一種三項式選舉模型，並利用台灣2004年總統大選的選舉資料，做資料分析與探討。文中之理論模型主要是架構在 Harris（ 1988 ）的基礎上，在假設所有計票誤差來自人為無心的疏失，和重新驗票能夠修正所有人為與心疏失，我們適度推廣其理論，使之能符合台灣的特殊選舉環境。

關鍵字

重新驗票；總統大選；三項式選舉模型

並列摘要

Serious debates ensued after the 2004 Taiwan presidential election and as a result a formal recounting process was implemented. In this paper, we proposed a mathematical model for the recounting. Our model is an extension of the one proposed by Harris (1988). Our basic assumption is that errors in misclassifying votes are due to unthought negligence. Denote by d the difference of the votes received by the two candidates. The distribution of d will be investigated. It is shown that under the basic assumption the recounting will not change the result of the election.

並列關鍵字

Election recounting ； multinomial election model ； presidential election

參考文獻

重新驗票能翻盤嗎

Google Scholar

預測驗票的結果

Google Scholar

Dempster, A. P.,Laird, N. M.,Rubin, D. B.(1977).Maximum Likelihood from incomplete data via the EM algorithm (with discussion).Journal of the Royal Statistical Society B.39,1-38.