幸福感的數位變遷：從年齡、時期與世代探討網路使用時間跟幸福感之關聯

Abstract

科技進步推動數位設備普及，深刻影響人們的生活與社會互動。自1998年以來，台灣民眾的網路使用隨時代演變，並與社會背景和科技發展密切相關。過往研究多關注年齡對網路使用時間與幸福感關係的調節效果，但對於時期與世代效應的影響仍缺乏探討。因此，本研究旨在檢驗網路使用時間對幸福感的影響，並進一步探討年齡、時期與世代的調節作用。本研究整合1998至2023年間共19期《台灣社會變遷基本調查》資料，超過兩萬名參與者，運用「階層年齡—時期—世代交叉分類隨機效應模型」（Hierarchical APC-Cross Classified Random Effects Models, HAPC-CCREM）進行分析。結果顯示：（1）控制年齡、時期與世代後，網路使用時間與幸福感呈倒U型趨勢關係。（2）年齡效應在網路使用時間對幸福感之間具顯著調節效果；其中，高齡者對網路使用時間的敏感度較高，倒U型趨勢更為明顯。（3）然而，時期與世代效應在網路使用時間對幸福感之間皆未達顯著調節效果。（4）以快樂感作為替代指標進一步分析，顯示倒U型趨勢及APC效果得到驗證。根據研究發現，網路使用時間對幸福感的影響存在最佳範圍，過度／少皆可能帶來負面效果，且此關係主要受年齡調節。研究結果對於理解數位時代變遷下的心理福祉提供了實證支持，為未來的政策制定與心理健康促進提供參考依據。

Technological progress has promoted the popularization of digital devices and deeply influenced people's life and social interaction patterns. Since 1998, the use of Internet in Taiwanese people has evolved with the times and is closely related to social background and technological development. Previous studies have mostly focused on the moderating effect of age on the relationship between internet use time and well-being, while the influences of period and cohort effects have received limited attention. Therefore, this study aims to examine the effect of internet use time on well-being and further explore the moderating roles of age, period, and cohort. This study integrates data from 19 waves of the "Taiwan Social Change Survey" conducted between 1998 and 2023, over 20,000 participants, this study adopts the Hierarchical Age–Period–Cohort Cross-Classified Random Effects Model (HAPC-CCREM) for analysis. The results show that: (1) After controlling for age, period, and cohort, internet use time shows an inverted U-shaped relationship with well-being. (2) Age significantly moderates the relationship between internet use time and well-being, with older adults being more sensitive to internet use time and the inverted U-shaped trend is more obvious. (3) Period and cohort effects do not significantly moderate the relationship between internet use time and well-being. (4) A supplementary analysis using happiness as an alternative indicator confirms the inverted U-shaped trend and APC effects. Based on the findings, there appears to be an optimal range of internet use time for well-being, while both excessive and insufficient use lead to negative outcomes. This relationship is primarily moderated by age.