Title

如何利用全國性民調推估地方民意?多層次貝式定理估計模型與分層加權的應用

Translated Titles

Using National Surveys to Estimate Local Public Opinions: An Application of Bayesian Multilevel Regression Model with Post-stratification

DOI

10.6683/TPSR.200606.10(1).5-38

Authors

俞振華(Eric Chen-Hua Yu);蔡佳泓(Chia-Hung Tsai)

Key Words

多層次貝式定理估計模型 ; 事後分層加權 ; 總和分佈 ; 選舉預測 ; 總統大選 ; Bayesian multilevel model ; post-stratification ; joint distribution ; election forecast ; presidential election

PublicationName

台灣政治學刊

Volume or Term/Year and Month of Publication

10卷1期(2006 / 06 / 01)

Page #

5 - 38

Content Language

繁體中文

Chinese Abstract

過去台灣學術性的民意調查研究主要關注的焦點在於推論全國性民意或全國性投票取向。受到樣本數的限制,研究者很難利用此類型的全國性民調資料推估地方民意(例如:以縣、市或鄉、鎮為單位的選民投票意向)。本研究透過多層次的貝式定理估計模型(Bayesian Multilevel Model),並以人口普查資料進行事後分層加權的方式(Post-stratification),發展出一套利用全國民調推估地方民意的模式。此模式主要包含以下兩個步驟:首先將全國人口依據基本特徵如性別、年齡、及教育程度等分成各個類型(Cell),並利用少數的全國性個體樣本輔以各區域(或稱次層級)總體特徵模擬出各區域中每一種人口類型的態度或意向。其次,配合人口普查資料我們可以得知各區域內其人口類型的總和分佈(Joint Distribution),並據以推估各區域或次層級的總體態度與意向。為驗證此模式的實用性,本研究進一步利用2004年總統大選前四週內約二千二百個民調樣本(其中僅有不及一千五百個樣本明白表明支持人選)及2000年台灣人口普查資料,成功地推估台灣本島地區二十三個縣市總統候選人的得票率。與各縣市實際的總統大選得票率相比,推估值和實際值的平均差距不到二個百分點。換言之,儘管各縣市在全國性民調資料中的平均樣本數不到一百個,利用本研究所發展的模式仍可準確地推估各縣市的總體民意。

English Abstract

Survey research in Taiwan almost always focuses on studying public opinion in the national level. Due to the potential small n problem for less populous sub-geographic units, scholars have not utilized national survey data to estimate public opinions in any local level (e.g., county, town, or village). To overcome this sample size limitation, we use a method combining the multilevel modeling approach with the population information for post-stratification. We apply it to a set of pre-election national surveys of the 2004 Taiwan presidential election (sample size<2200) and produce county-level estimates of vote choice. Specifically, we first construct a multilevel logistic regression model to estimate the mean of the vote choice variable given demographics and county of residence. Second, we post-stratify on all the variables in the model by using the joint population distribution of the demographic variables within each county. Comparing our estimates with the actual county-level election outcomes, the average absolute error is less than 2 percentage points.

Topic Category 社會科學 > 政治學
Reference
  1. Tsai, Chia-Hung(2005).Bayesian Inference in Binomial Logistic Regression: A Case Study of the 2002 Taipei Mayoral Election.Journal of Social Science and Philosophy,17(1),103-23.
    連結:
  2. 許勝懋、洪永泰(2003)。『2001年台灣選舉與民主化調查研究』抽樣設計:分層效果、抽樣誤差與設計效果之分析。選舉研究,10(2),59-91。
    連結:
  3. 黃紀、張佑宗(2003)。樣本代表性檢定與最小差異加權:以2001年台灣選舉與民主化調查為例。選舉研究,10(2),1-35。
    連結:
  4. Abramowitz, Alan I.(1988).An Improved Model for Predicting Presidential Election Outcomes.PS. Political Science & Politics,21,843-47.
  5. Clinton,Joshua,Simon Jackman,Douglas Rivers.(2004).The Statistical Analysis of Roll Call Data.American Political Science Review,98(2),355-70.
  6. Bugs. R.: Functions for Calling Bugs from R
  7. Gelman, Andrew.(2004).Unpublished manuscript.
  8. Gelman,Andrew,John B. Carlin,Hal S. Stem,Donald B. Rubin.(2004).Bayesian Data Analysis.London:Chapman & Hall.
  9. Gelman,Andrew,Thomas C. Little.(1997).Postratification into Many Categories Using Hierarchical Logistic Regression.Survey Methodology,23,127-35.
  10. Gill, Jeff.(2002).Chapman & Hall/CRC,Boca Raton.
  11. Goldstein, Harvey.(1995).Multilevel Statistical Models.New York:John Wiley.
  12. Jackman, Simon.(2000).Estimation and Inference via Bayesian Simulation: An Introduction to Markov Chain Monte Carlo.American Journal of Political Science,44,369-98.
  13. Jackman,Simon,Douglous Rivers(2001).Paper presented at the Annual Meeting of the American Political Science Association, San Francisco, CA..
  14. Kelley, Stanley, Jr.,Thad W. Mirer.(1974).The Simple Act of Voting.American Political Science Review,68(2),572-91.
  15. King, Gary(1997).A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data.Princeton, NJ:Princeton University Press.
  16. Kramer, Gerald.(1971).Short-Term Fluctuations in U. S. Voting Behavior, 1896-1964.American Political Science Review,65(1),131-43.
  17. Lavine, Michael.(1999).What Is Bayesian Statistics and Why Everything Else Is Wrong.The Journal of Undergraduate Mathematics and Its Applications,20,165-74.
  18. Lewis-Beck,Michael S.,Tom W. Rice.(1992).Forecasting Elections.Washington, DC:CQ Press.
  19. Martin, Andrew.(2001).Congressional Decision Making and the Separation of Powers.American Political Science Review,95(2),361-78.
  20. Park, David.,Ph. D. diss(2004).Department of Political Science Columbia University, New York, NY.
  21. Park,David,Andrew Gelman,Joseph Bafumi(2004).Bayesian Multilevel Estimation with Poststratification: State-Level Estimates from National Polls.Political Analysis,12,375-85.
  22. Quinn, Kevin M.,Andrew D. Martin,Andrew B. Whitford.(1999).Voter Choice in Multi-Party Democracies: A Test of Competing Theories and Models.American Journal of Political Science,43(4),1231-47.
  23. Vienna, Austria: R foundation for Statistical Computing
  24. Rao, J. N. K.(2003).Small Area Estimation.New York:Wiley InterScience.
  25. Rosenstone, Steven J.(1983).Forecasting Presidential Elections.New Haven:Yale University Press.
  26. Spiegelhalter, D.,A. Thomas,N. Best.(1999).WinBugs Version 1.4.Cambridge, UK:MRC Biostatistics Unit.
  27. Tsai, Chia-Hung.(2000).American Voter Responses to International Political Events and Economic Conditions: 1920-1996.Journal of European and American Studies,30(3),143-91.
  28. Tufte, Edward R.(1978).Political Control of the Economy. Princeton.NJ:Princeton University Press.
  29. Western,Bruce,Simon Jackman.(1994).Bayesian Inference for Comparative Research.American Political Science Review,88(2),412-23.
  30. Yu, Eric C. H.(2004).Paper presented at the 2004 Annual Meeting of Northeastern Political Science Association, Boston.
  31. 洪永泰(1994)。選舉預測:一個以整體資料為輔助工具的模型。選舉研究,1(1),93-110。
  32. 洪永泰、謝復生、盛杏湲(2000)。政治學的範圍與方法。台北:五南圖書出版公司。
  33. 盛杏湲(1998)。選民的投票決定與選舉預測。選舉研究,5(1),37-76。
  34. 盛治仁(2000)。總統選舉預測探討一未表態選民與情感溫度計。選舉研究,7(2),75-107。
  35. 莊文忠(2000)。選舉預測與策略性投票:以八十九年總統選舉為例。理論與政策,14(2),55-92。
  36. 黃紀(2005)。多層抽樣設計芻議。「台灣選舉與民主化調查研究方法工作坊」學術研討會,台北:
  37. 黃紀(2001)。一致與分裂投票:方法論之探討。人文及社會科學集刊,13(5),541-574。
  38. 黃紀、張益超、徐永明、黃紀(2001)。政治分析的層次。台北:韋伯文化事業出版社。
  39. 劉文卿(1995)。台北市長選舉之基因預測模型。選舉研究,2(1),1-16。
  40. 劉念夏(1996)。一九九六年總統大選選舉預測:民意調查中未表態選民投票行為規則假設的提出與驗證。選舉研究,3(2),131-156。
  41. 劉義周(1996)。選舉預測:一組簡單理論的檢驗。選舉研究,3(2),107-130。
  42. 鮑彤(1999)。選舉地盤:候選人票源凝聚程度之分析。台北:國立政治大學政治學碩士論文。
Times Cited
  1. 黃羿菁(2007)。Logit模型預測2006年高雄市長選舉。淡江大學管理科學研究所碩士班學位論文。2007。1-49。 
  2. 魏瑞慧(2008)。層級貝氏模型應用於信用卡顧客消費行為分析之研究。臺北大學統計學系學位論文。2008。1-62。