問卷設計：一項對民意調查的挑戰

Abstract

隨著我國自1980年代中期以來的民主化，台灣民眾的意見逐漸受到重視，各式各樣的民意調查也出現在政治菁英的言談與國人每日的生活中。乍看之下，民意調查很容易，然而，實則對民意調查研究者而言，如何設計出適當的問卷以得到受訪者真實的答案，卻是常常必須面對的艱難挑戰。本文指出民意調查問卷設計之所以困難有三個主要原因：第一，民意調查研究的對象是人，而人的想法往往複雜抽象，反覆無常，且人也會自我防衛與自我合理化，甚至人會隱瞞欺騙，因此如何以問卷測量出民眾真正的意見並不容易。第二，問卷設計的內容僅僅是些微的差異，就會造成調查結果的極大不同。第三，若民意調查研究者不加留意，可能會犯下一些設計問卷易犯的嚴重錯誤。這些都在在提醒民意調查研究者在設計問卷時必須格外謹慎。

Responding to the democratization from the beginning of mid-1980s, Taiwanese public opinions have been taken more seriously than before. Various polls have occupied an important part in political elites` discussions and people`s daily life. On the face of it, to conduct a poll is simple. However, it is not simple at all. The pollsters always have to face a formidable challenge: How to get the true answers from respondents. It is a formidable challenge because of three major reasons: First, the respondents of polls are people. People`s opinions are always complicated and intangible. People are sentimental, fickle, self-defensed, and self-rationalized. People may even cheat sometimes. Second, it is more difficult because a slight difference in questionnaire may make a radical difference in answer. Third, pollsters may commit some serious errors if they are neglect. All of these remind pollsters to design questionnaires with special care.