適用於研究生與大學生之 線上自我調節量表開發－因素分析應用

Abstract

本研究係針對一份新開發的量表進行研究，該量表名稱為「線上自我調節量表」(Online Self-Regulation Scale, ORSR)，共包含 15 題，用來量測造成線上學習者表現不理想的因素，以及線上學習者所使用的自我調策略。本研究主要透過因素分析的方式，瞭解該量表之因素結構，並檢驗其聚斂效度、區辨效度以及效標關聯效度。 本研究共包含 352 位大學生與研究生，資料收集分為兩波進行，以第一波收集得來的資料進行探索性因素分析，藉以找出潛在的量表結構，之後再整合兩波資料進行驗證性因素分析，並利用驗證性因素分析結果，進行後續的量表效度檢驗。為檢驗量表的效標關聯效度，另外收集了參與者在網路自我效能量表(Internet Self-Efficacy Scale, ISES)和線上資訊搜尋策略量表 (Online Information Searching Strategies Inventory, OISSI)上的得分，以及他們的學業成績。 探索性因素分析結果顯示，OSRS 共包含四個構念，按照題意分別命名為知覺網路濫用、自我調節策略、罪惡感以及環境干擾；然而，根據探索性因素分析之結果，OSRS 尚有需要修正之處，故在題目修正後，進行第二波資料收集，之後再以整合的資料進行驗證性因素分析。驗證性因素分結果指出，假設的三因素模型為一良好適配的模型，代表OSRS 其實是包含了三個構念，此三構念分別為：知覺網路濫用、自我調節策略和環境干擾。接著，再利用驗證性因素分析所得結果，進一步檢驗 OSRS 的因素負載量、構念間的相關係數、OSRS 與 ISES、OISSI、學生學業表現間的相關係數。分析結果顯示，OSRS 具有合理的聚斂效度，且在區辨效度方面表現理想，同時，與其他線上學習相關的量表以及學業表現之間，具有顯著相關，顯示該量表的效標關聯效度亦頗為不錯。文末列出幾項研究限制，並提供建議予後續研究者，並針對 OSRS 未來在教育領域的應用，給予發展方向。

In this study, a newly-invented instrument, Online Self-Regulation Scale (OSRS) was used for measuring online learners’ scores on reasons for undesired learning outcome and self-regulated strategies. The purpose of the present study was to investigate the factor structure of the 15-item OSRS, and examine its convergent validity, discriminant validity and criterion validity. There were two waves of data collection. Data collected in the first wave was used to perform EFA. The combination data of the first and second wave was used to perform CFA. There were 352 participants in total. With the CFA results, further examination of scale’s validity could be executed. Participants’ scores on Internet Self-efficacy Scale (ISES) and Online Information Searching Strategy Inventory (OISSI) and academic performance were gathered for analyzing criterion validity. The EFA results revealed that there were four factors in OSRS, which were named as perception of Internet misuse, self-regulated strategies, guilty and environmental interruption. However, some revisions were needed. Then, based on the EFA results and the adjustment, CFA was performed to analyze the combination data. The proposed CFA model was consisted of three factors, which were perception of Internet misuse, self-regulated strategies and environmental interruption. CFA results suggested that the three-factor model was an adequate model. By examining values of factor loading, correlations between constructs of OSRS, and correlations between constructs of OSRS, ISES, OISSI and students’ academic performance, OSRS was regarded as having adequate convergent validity, good discriminant validity and fair criterion validity. In the end, educational implication, limitation and suggestion were provided.