終身學習的理念驅使社會人士持續在人生和職場上追求成長，在人生上追求更豐富的生活，在職場上保有穩定的競爭力。根據吳明烈、李藹慈 (2020)發現，數位學習在近年已成為成人在職場學習的主流方式。本研究使用學術調查研究資料庫(SRDA)中 2016年至2020年之「成人教育調查」問卷結果，並應用二元選擇方法以及最小平方法估計式，分析成人之個人特徵與其參與數位學習的關聯性效果。實證研究結果顯示，數位學習的參與度以及參與時數受到婚姻狀態、年齡、性別、居住地、教育程度、年收入的影響。研究發現，有伴侶的學習者參與數位學習的傾向明顯高於無伴侶者。有伴侶的學習者因為時間與空間受到限制，數位學習的彈性課程使其更能有效安排學習時間。隨年齡增長，參與數位學習的機會增加，但部分長者對科技設備的不熟悉可能導致對線上學習感到焦慮。居住在非六都地區的居民參與數位學習較少，可能受到偏遠地區網路設施不穩定的影響。學歷方面，專科學歷以上的個體參與數位學習的機率較高，這可歸因於知識分子對學術環境較為熟悉。此外，高收入群體更傾向於選擇數位學習，且使用時數相對較高，顯示高收入者傾向透過多元學習方式提升知識。;With the concept of lifelong learning, individuals pursue personal and professional growth continuously in order to enrich their lives and maintain their competitive advantages. Wu & Li (2020) find that the e-learning has become the mainstream recently. Based on the report of Adult Education Survey during 2016 to 2020 in the Statistical Research Data Archive (SRDA), we apply the Probit and Ordinary Least Squares models for investigating the influential factors of e-learning participants and weekly learning hours dedicated by individuals. Our empirical findings show that the online learning participation and participation hours are influenced by marital status, age, gender, residence, education level, and income. The individuals with partner who have the time and space constrain exhibit a greater inclination towards e-learning. Because of the flexibility of online courses, the learners with partner are able to manage their time effectively. Participation in online learning increases with age but older learners may experience anxiety due to unfamiliarity with technology. The residents who live in non- municipality areas participate less in digital learning due to some remote areas where the networks are not stable or available. Those with a college degree or above have a greater probability of participating in online learning as they are more familiar with the academic environments. Furthermore, higher-income group prefers e-learning, and dedicates more time to it. High-income group tends to improve their knowledge through diverse learning methods.
