本文利用 2004 年立委 TEDS 的實證資料，嘗試解答下列研究問題：（1）民眾面臨重大問題時，其諮詢對象為何？（2）社會網絡中的政治討論情形為何？（3）政治討論是否會影響人們的投票參與？研究獲致的發現如下。首先，當人們面對重大問題時，有配偶者與無配偶者尋求諮商的社會網絡有所不同。有配偶者依賴的討論對象是建立在一個以家庭中最親近成員為核心，而逐步向外擴展的網狀系絡；其諮詢對象依次是配偶、直系親屬、旁系親屬，最後才是非親屬關係的朋友、同事等。至於多數無配偶者的討論對象分佈於兩大類：首為父母，次為朋友。旁系親屬則落居第三位。其次，社會網絡中的政治討論呈現出幾種不同的面貌。第一、有配偶者的政治討論頻率高於無配偶者，這可能與前者年齡較大有關。第二、諮詢對象的親疏程度與意見同意度有高度正相關。第三、大多數民眾認為討論對象對於政治事務是介於有點瞭解與不太瞭解之間。最後，有關政治討論效果的研究。在無配偶者中，相較於對照組，討論壓力使人對投票卻步；討論所帶來的鼓勵則讓決定投票者較早選擇其支持對象。至於有配偶者，討論壓力使人傾向不去投票的效果，也幾近達到統計上的顯著水準。 In the argument of classic democratic theories, active political discussion is one of the motives for crafting democracy. By applying the data of 2004 TEDS, this study attempts to explore three questions: (1) who will be the major discussants when people urge advices? (2) how people discuss political issues within their social networks; (3) how political discussions affect voting participation.There are several findings.?First, the social networks of the married people are based on family members, including spouse, parents, children, brothers, and sisters.In contrast, sources of the unmarried people's social networks are mainly from parents and friends.Second, the married people are more interested in discussing political issues than the unmarried people.This maybe due to the life cycle effect. According to the literature, the elder or married people will pay much attention to public affairs. In addition, most people are likely to choose their discussants who share consistent political opinions.? Finally, the pressures or encouragements brought by political discussions impose significant effects on voting participation, although the extent of their influences depends on different cases.In conclusion, this study suggests that the data collected by panel studies are necessary for further research.