Citizens Trust for Public Opinion Survey Methods by Government- A Comparative Study of Public Opinion Analysis by Telephone and Internet

民意是一個相當模糊的概念，民意測量的結果和真實民意的狀態是否相同沒有人能夠解答，而政府經常以民意作為政策規劃的基礎，測量方法的可信度便成為民主政府在政策制定上需要加以考量重點之一，當民眾不信任民意調查工具時，政策規劃的過程也會遭到民眾的質疑。然而，過去針對民眾對於民意蒐集分析方法信任程度的相關研究相當缺乏，本研究除了整理人口變項的影響文獻，也將可能影響民意調查的政治行為與政治態度作為自變數納入分析模型中，試圖探尋具有影響力的重要自變數。 在網路堀起的現代，網路民意漸漸被政府重視，網路輿情分析的存在是為了因應網路民意發展而產生的新技術。本研究將網路輿情分析視為民意蒐集與分析的工具之一，並利用民意調查相關的文獻作為基礎，探究網路信任、網路政治效能感等變數影響網路輿情分析信任程度的情況。 本研究使用網路民意調查作為資料蒐集方法，問卷設計上以簡單的圖示與表格介紹民意調查與網路輿情分析的流程，再以結構化問卷方式測量受訪者的網路使用、網路信任、政治參與、政治信任、網路政治效能感、機構信任、對民意調查的信任程度題組與對網路輿情分析的信任題組。再以多元迴歸分析建立變數間影響力的模型，綜合文獻與研究結果探討民眾對於不同調查方法的信任程度差異。 研究發現，年齡越高對於兩種民意分析方法的技術信任程度越低，政治信任程度會顯著影響民意調查的信任，對於網路輿情分析的信任則無顯著影響，而網路政治效能感越高，對於網路輿情分析的技術信任則會越高。而政治參與程度高的民眾對於民意調查支持程度越高，對於網路輿情分析的支持程度越低。政治機構信任對於兩種分析方法的信任都是具有顯著影響力的變數。 從研究結果可知，網路快速傳播的特性造就了其在議題領導上的優勢，如果政府欲進行網路輿情分析進行網路民意的探戡與分析，甚至是以網路上的民意作為政策行銷的依據，藉由增進民眾對於網路信任、政治信任、網路政治效能感等政治態度，來提昇民眾對網路輿情分析的信任，且提昇民眾的政治參與程度並增加政策資訊的透明度，消除民眾對於網路隱私權的疑慮，可能會有效提昇民眾對網路輿情分析的支持，政府如果能夠與研究機構進行民意的研究，民眾將較相信且支持。Public opinion is an ambiguous concept. There is no answer to the question whether opinion polls genuinely reflect public opinion or not; thus, reliability of the survey methods of measurement becomes a significant issue. This is particularly important to government as its policy-making process heavily relies on the results of opinion poll. If citizens don’t trust the survey methods, they inevitably have significant doubt on the public policy. However, there are few empirical studies engaging in the issues of the trust for public opinion surveys. This study, accordingly, first examines the influence of demographic variabes on citizen trust in two diffrenet suvery methods. It then examines the effect of political behavior and attitude to identify the most influential independent variables to explain citizen trust for the survey methods. In addition to the existing telephone survey (TS), there is now a new method called internet public opinion analysis (IPOA). The study compares both survey methods in terms of internet trust, political participation, political trust, internet political efficacy and house trust by using multiple regression models. The key findings show that, among all demographics status, the elder citizens have lower trust in both survey methods. Despite of its statistical significance for TS, political trust is relatively insignificant to IPOA. Internet political efficacy and political participation are also significant factors to IPOA. Similarly, political house trust is significant to both TS and IPOA. Based on the key findings, political issues spread rapidly on the internet, and online public opinion becomes more important. If government attempts to utilize IPOA as one of the survey methods for public policies, it is suggested to increase citizen’s internet trust, political trust, internet political efficacy. In addition, political participation may increase citizen’s trust and support for IPOA. By increasing policy transparency, decreasing citizen’s doubts on privacy and cooperating with academic institutions, government would be possibly able to reinforce citizen’s trust in IPOA