網路同溫層與黨性強度：定群追蹤資料的實證分析

Internet Echo Chambers and the Strength of Partisanship: An Empirical Analysis of Panel Data

以往文獻針對同溫層與政治態度關係的探索，多集中於單向因果關係的討論。然而，無論是從兩者的概念意涵思之，或者觀察現實的政治現象，同溫層與政治態度彼此具雙向的互為因果關係，應是有極大可能性。為了驗證此一看法，本文選擇網路同溫層、黨性強度為觀察變數，嘗試釐清二者的因果關係與影響力大小。本文提出兩個研究假設：（一）網路同溫層、黨性強度二者具有互為因果的關係；（二）相較於網路同溫層對於黨性強度的影響力，黨性強度對於網路同溫層的影響力較大。本研究使用次級資料分析法，資料來自「『台灣政經傳播研究』多年期研究規劃：2018–2021年民眾定群追蹤面訪調查資料」。統計方法上運用定群追蹤資料迴歸，深入分析四波面訪資料；結果發現，本文所提出的兩個研究假設皆獲得證實。本文此一在研究資料與統計方法上的改進，使得以往文獻對於同溫層與政治態度孰為因果的爭議得以獲得解答，因而也對相關理論內容的補正有所貢獻。

Previous research exploring the relationship between echo chambers and political attitudes has predominantly focused on unidirectional links. However, considering the conceptual implications of both factors and observing the real-world political dynamics, there is a strong possibility of a bidirectional relationship between echo chambers and political attitudes. To validate this perspective, this study selected the variables of online echo chambers and the strength of partisanship to examine the reciprocal causal relationship and their respective magnitudes of influence. Two research hypotheses were proposed. First, online echo chambers and the strength of partisanship have a mutually causal relationship. Second, the strength of partisanship has a stronger impact on online echo chambers than vice versa. Employing a secondary data analysis approach, data adopted in this study were derived from the Taiwan Institute for Governance and Communication Research (TIGCR) from 2018 to 2021. A regression analysis was applied to the four-wave panel data obtained via face-to-face interviews. The results confirmed the research hypotheses. The methodological improvements presented via the data analyses in this study contribute to resolving a longstanding debate in the literature on the causal relationship between echo chambers and political attitudes. Consequently, this study provides valuable insights for the relevant theoretical content.