學生的學習成效是教育品質的展現,學習策略被視為影響學習成效之關鍵因素,但甚少研究針對學習策略與學習成效的動態變化進行分析。本研究以嶺東科大數位媒體設計系二年級學生為研究樣本,針對60 個有效固定研究樣本,以潛在成長曲線模式(Latent Growth Curve Modeling, LGM)為研究方法,分析四次多媒體互動作品成績,並依其學習策略分群比較,探討學習策略對學習成效的動態影響。研究結果發現:1. 學習成效成長並非完整一直線,會隨著時間而變動,整體則有下降的趨勢。2.不同學習策略對動態學習成效無顯著差異,不同學習策略在學習成效上有一致的成長趨勢,但變化的程度則有高低之分。3.深度學習者與策略學習者初始的成績高者,其成績下降愈大。4.成績表現策略性學習者的分數最高、深度學習者次之,表面學習者最低。5.表面學習者之起始狀態對成長速率無預測作用。對教學實務的建議:教師應採取形成性評量,更能呈現真實的學習成效。學習評量具體考量認知、情意與技能三部份,並給予時間彈性,可增進深度學習。
Learning strategy is one of the key factors of learning efficiency. However, very little research is engaged in analyzing the dynamic changes of the learning strategies and learning outcomes. The purpose of this research is to explore how three kinds of learning strategy (deep, surface, and strategic) affect assessment process, and focuses on sophomore students in the Department of Digital Content Design at Ling Tung University, a technology university in Taiwan. Sixty effective samples are collected and analyzed by Latent Growth Curve Modeling (LGM). The process includes: to analyze graded results of four multimedia projects, to compare clustering in learning strategy, and to evaluate dynamic impact of learning strategy on learning efficiency. The discoveries are: 1. the growth curve line of learning efficiency is not always straight. It changes over time. In the end, the curve goes down; 2. Different learning strategies do not show an apparent impact on dynamic learning. Different learning strategy displays a consistent growth trend in learning efficiency. 3. The strategic and deep learners earn scores that are higher initially that then decline greatly later. 4. Strategic learners earn the highest scores, the deep learners second, and the surface learners the lowest. 5. The initial growth rate cannot predict learning efficiency. For teaching practice, the researchers recommend that the teachers should first adopt the formative assessment to present real learning outcomes. Secondly, the teachers should consider the cog nitive, affective and skills assessment, and adapt the class teaching to enhance deep learning.