透過您的圖書館登入
IP:3.145.97.248
  • 期刊

中文網路成癮量表之等級反應模式分析

An IRT Study of Graded Response Model on the Chinese Internet Addiction Scale

摘要


試題反應理論(item response theory, IRT)具有試題與能力不變的心理計量特性,能克服傳統測量模式中樣本依賴的限制,置於同一量尺的能力與試題參數可相互比較。而本研究旨在應用多元計分的「等級反應模式」進行中文網路成癮量表品質與網路成癮潛在特質的分析,以提升臨床診斷上的效率。研究樣本來自北部三所綜合型大學,以網路問卷方式蒐集資料,有效樣本為987名。 研究結果發現,中文網路成癮量表品質符合本質上單向度之假設,證實測量單一潛在構念;在測量恆等性的檢驗方面,共有7題在男性與女性群體中產生差異試題功能的情形,其中以「網路成癮耐受性」的試題居多,顯示這類試題具有個別差異,須要仔細審視;試題各選項的閩難度可以涵蓋輕微至嚴重之網路過度使用情形,而其中的第15、16、19、24題第三階閾難度偏高,可提供臨床診斷上的依據。最後,本研究提出數項建議供未來研究改進之參考。

並列摘要


Unlike classic test theory (CTT), item response theory (IRT) has the benefit of population-free item and trait estimation that could be compared to each other since persons and items are placed on a common scale. Before comparing groups of individuals with regard to their trait level, one must confirm that the numerical values under consideration are on the same measurement scale, and that is ”measurement invariance”. Differential item functioning (DIF), defined as when the probability of endorsing an item depends on both individual's trait level and other factors such as gender, is usually used to assure measurement invariance and pure anchor items for the purpose of linking several examinee's group membership. Based on graded response model, this study is aimed to analyze the quality of Chinese Internet Addiction Scale (CIAS) and facilitate clinic diagnosis efficiently. By conducting a questionnaire survey on a website, nine hundred and eighty seven samples were selected from students in three universi ties of northern Taiwan. The findings are as follow. First, CIAS is congruent with essentially unidimensionality and fit the assumption of IRT. Second, there are seven of the total items with DIF and the remaining items with equivalence are used as anchor items to link gender groups. Besides, most of the DIF items belong to the domain of ”Internet addiction tolerance”. Third, the item discrimination parameters are so large that most items are strongly related to the latent trait and the category difficulties would cover all the range of Internet addiction levels. We also find there are four items which have large values in the third category difficulty and conclude that these items could be efficient for clinic diagnosis. Finally, researcher provides several suggestions for the future.

參考文獻


王文中(2004)。Rasch測量理論與其在教育和心理之應用。教育與心理研究。27(4),637-694。
林珊如、黃宏宇()。
林珊如、黃宏宇(2004)。網路敵意及網路沈迷-兩種去社會抑制行為之相關研究。教育與心理研究。27(2),325-352。
林珊如、蔡今中(1999)。國科會「教育政策研析:網路科技對教育的影響」計畫成果。國立台灣師範大學。
陳淑惠(1999)。國科會「教育政策研析:網路科技對教育的影響」計畫成果。國立台灣師範大學。

被引用紀錄


蔡雅惠(2011)。桃園市某國中學生自尊與網路成癮關係之研究〔碩士論文,國立臺灣師範大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0021-1610201315231948

延伸閱讀