近年來台灣網路成癮悪化的程度，已與網路成癮情況嚴重的南韓不相上下，而深受網路成癮之害的是上網人口比例最高的青少年，網路成癮折損個人向上提升的能量及動力。為了深入了解造成青少年網路成癮的原因，本計畫以2011 年3 月間在台北及花蓮兩地所搜集2083 高中職學生網上行為研究資料，進行網路成癮分析，透過客觀條件（上網時間、居住地點、性別）及主觀條件（偏好變數、依附變數），來探討什麼樣背景的青少年較容易流連忘返於網路？什麼樣類型的網站比較容易讓人廢寢忘食？什麼樣的管教方式較能遏止青少年沈迷於網路？什麼樣的人際關係讓網路具有致命的吸引力？青少年是發展人際關係，培養社交技巧的關鍵時期，網際網路匿名特質，讓人際互動增添神秘感，挑起探險的刺激感，也延伸實體社交的場域，但若過分沈溺於虛擬人際關係而難以自拔，所造的影響難以評估。網路同儕與實體同儕較勁之下，對網路成癮所能發揮的影響力？青少年若在現實生活中找不到紓解壓力的出口，網路是否成為最佳的避難所？本計畫除了使用敍述、平均數、交叉、卡方等方法來分析網路成癮相關指標變數，並利用HLM 多層次線性模型來檢驗14 所學校樣本資料，是否具有群聚效果或交互作用？檢驗結果顯示同一所學校的學生間相關性極低，大部分的變異來自學生本身，故採用個體層次多元線性迴歸來進行分析，結果顯示常利用網路來紓解壓力，結交網路朋友數量多於現實生活朋友的都會（台北）地區的青少年,可能成為網路成癮的機率較大。

In recent years, the level of Internet addiction in Taiwan has reached a similar levelthat is in South Korea, which has a serious Internet addiction situation. Teenager is thesingle highest population of the total Internet population; it is also the target victims ofInternet addiction. Internet addiction affects the motivation and energy of personalgrowth in youth. In order to extensively understand the relationship between theinter-personal human interaction and Internet addiction, this research collected data ofthe online behaviors of 2,083 high school students’ in Taipei City and Hualien County.The main goals include: To explore what kind of personalities are more drawn to theInternet? The types of online activities that are more likely to become addictive? Whatkind of parenting style can curb the addiction? Moreover, the types of Internetinteraction with peers can that create addiction? Adolescence is critical period fordeveloping inter-personal relationships and social skills. The anonymous nature of theInternet increases the mystery of inter-personal interaction, provokes the excitement ofadventure, and also extends the scope of ordinary social life. It is difficult to extricateoneself after becoming addicted to virtual relationships. The impact of the Internetaddiction is difficult to assess. This research applies descriptive, average,cross-tabulation and chi-square analysis to analyze basic patterns of the relationshipsbetween the indicator variables and Internet addiction. HLM (Hierarchical LinearModeling) is also adopted to assess whether aggregations or interactions exist in thesample data of students from 14 high schools. The results indicate that the correlationwithin the students of the same school is relative low. Most of the variation comes fromthe students themselves. Therefore, the multiple linear regression analysis is appliedinstead. Those results show that independent variables such as region, Internet peers, andemotional stress are critical predictors on Internet addiction.