標題: 問卷選項排序分析Ranking responses for questionnaire 作者: 張楷威Chang, Kai-Wei王秀瑛Wang, Hsiuying統計學研究所 關鍵字: 問卷調查;複選題;卡方檢定;列聯表;選項排名;Questionnaire;Multiple response question;Chi-square test;Contingency table;Ranking response 公開日期: 2012 摘要: 問卷調查在很多研究之中是一項常用的工具，但問卷格式的設計方式是依據研究者有興趣的目的去設計。複選題是問卷常見的題型。近年來，很多研究提出了一些模型和方法針對複選題問題的資料作分析。Wang(2008)提出一些方法來檢定複選題當中任兩個選項被選到的機率是否相等。另一種題目是把每個選項給予1到5等級的去評分，在本篇文章中提出結合複選題排序的方法及5等級評分的分數來找出選項的排名。然而，並不是所有的選項都能顯示出其背後因子的重要性。在本文中提出在排序之前可利用卡方列聯表檢定(Chi-squared test for contingency table)和歐式距離(Euclidean distance)找出重要因子，且排除不重要的因子。並根據模擬結果，來驗證卡方列聯表檢定此方法是可行的。Questionnaire is a common tool for surveying in many areas. The design of questionnaire is according to the goal that researchers are interested in. A multiple response question is a commonly used question in a questionnaire. Recently, many studies proposed models and approaches for analyzing data of a multiple response question. Wang (2008) proposed methodologies for testing the equality of selected probabilities for two responses. Another question is to require a respondent to select a score for each item. In this study, we propose to combine the principle of ranking responses and scores to rank the responses. However, not all of responses are shown to be important because of their hidden factors. In this study, we propose using chi-squared test on contingency table and Euclidean distance to determine the important factors and remove unimportant factors. According to the simulation study, we show that our method is a feasible method in finding the important factors. URI: http://140.113.39.130/cdrfb3/record/nctu/#GT070052614http://hdl.handle.net/11536/71897 Appears in Collections: Thesis