研究目的：流行消費的主要目的已不是為了滿足基本需求，而在於追求情感性的與社會性的目標。本研究從社會互動的觀點探討青少年階段的同儕與家庭關係如何影響大學生的流行消費行為，並進一步檢視性別效應與寂寞感的中介作用。研究方法：我們使用「台灣青少年成長歷程研究」針對中學(第一至六波)以及大二(第八波)階段的長期追蹤問卷調查資料，篩選就讀一般或技職體系的大學生做為分析樣本，共計1,070位，並運用多元迴歸分析法進行統計檢定。研究結果：本研究有兩個主要發現：（一）友誼受創事件和家庭受創事件有提高流行消費支出的作用，但家庭受創事件的影響僅限於女性，對於男性的影響不顯著；（二）寂寞感對於受創事件和流行消費行為之間的連結具有中介作用，但寂寞感的中介作用是有性別效應的，只有女性受中介因素的影響。研究結論：本研究顯示負面的社會關係對於大學生流行消費行為的影響效果與影響路徑因性別因素而不同。往後的消費研究可就性別因素進行更有系統的探討。

Purpose: For the younger generation, fashion consumption is focused on social and emotional aspects rather than on basic needs. The current study investigates how family and peer interactions during adolescence affect fashion consumption among college students. It further examines whether the associations vary according to gender and whether they are mediated by a sense of loneliness. From the perspective of social interaction, we hypothesize that (1) peer or family traumatic events are associated with greater fashion consumption and the association is mediated by a sense of loneliness, (2) the associations among traumatic events, the sense of loneliness, and fashion consumption is stronger for female college students than for males. Methods: The empirical data used to examine the hypotheses were taken from Taiwan Youth Project (TYP), a longitudinal survey research project conducted by the Institute of Sociology, Academic Sinica. The TYP survey, launched in 2000, applies a multistage stratified-cluster random sampling framework to obtain school-based representative samples. The sampled students were seventh graders when the first-wave survey was administered. Annual follow-up surveys continued until 2009. The data analyzed in this study include reports collected at both high school (waves 1 to 6, from March 2000 to March 2005) and college stage (wave 8, June 2007). The sample size was 1,070. Two-stage multiple regression analysis was employed to investigate gender effects on the association between social relationships and fashion consumption. Two separate multiple regression analyses were then used to further examine whether the association between social relationships and fashion consumption is mediated by a sense of loneliness. Results: There were two major findings. First, both friendship and family traumatic events promote expenditure on fashion consumption as hypothesized, but the effect of family traumatic events was significant only for female and not male students. Secondly, the effect of friendship and family traumatic events on expenditure on fashion consumption is mediated by loneliness. The mediating effect of lonliness, however, was only significant among female but not male students. Specifically, female college students who had ever suffered negative friendships were more likely to have a stronger sense of loneliness and then to make greater expenditures on fashion consumption. Conclusions: The social and emotional influence model proposed by this study seems fitting for females rather than for males. It indicates research on fashion consumption need to pay attention on gender effects.