雖然初探研究支持學生試卷建置對知識統整、精緻化與編寫題目技巧提升的教育潛能，但其學習效益仍有賴進一步研究驗證，鑑此，本研究嘗試以教學實驗方式探究學生試卷建置之學習效益。研究目的除探討參與試卷建置線上學習活動學生在自然科學習之前後認知、情意與行為表現的差異情形，並比較學生試卷建置與學生出題兩種方式對自然科學習成就、認知以及情意表現的差異效果。配對t-test分析結果顯示，試卷建置活動能顯著提升學生自然科學習態度與動機，但對認知與後設認知策略成效未達顯著水準；此外，試卷建置活動能促發參與學生的自我指導學習行為。最後，共變數分析結果發現，學生試卷建置與學生出題在自然科學習成就、認知及情意表現皆未有顯著差異，顯示兩者具相當的學習效果。文末，針對本研究結果與發現，提出本研究貢獻與教學、未來研究建議。

While the learning potential of student-constructed tests (SCT) for the promotion of knowledge integration and elaboration has been suggested, its learning effects warrant further empirical examination. A single-group pretest-posttest experimental research design was adopted to examine if the performance of students exposed to the SCT condition was enhanced, while a control-group pretest-posttest experimental research design was used to examine the comparative effects of SCT and student generated questions (SQG) on science learning. An online learning system was adopted to support the associated activities involved in the respective groups. The results from the paired t-tests found significant differences in student attitudes toward science and science learning motivation. Yet, no significant differences were found in student use of cognitive and metacognitive strategies. Furthermore, SCT induced the majority of the participants into self-regulatory behavior. Finally, the results from the ANCOVA found no significant differences in any of the examined variables between the SQG and SCT groups. Suggestions for instructional implementations and future study are provided.