知識結構網路表徵分析之研究－－高一基礎力學為例

Analysis of Knowledge Structure Network Representations: an Example for Tenth-grade Fundamental Mechanics

本研究旨在應用徑路搜尋法建構分析高一學生力學知識結構，探究知識結構指數對學習成就之預估效力，並深入探究知識結構網路表徵之組型差異及節點活化情形，以提供教學診斷和教學設計之參考。本研究對象為高雄地區高中一年級學生共計148人進行知識結構評量，配合圖解與放聲思考深入分析知識結構網路表徵。主要研究工具有「力學知識結構測驗」、「基礎物理力學成就測驗」、「放聲思考力學解題分析」。主要研究發現為：1.知識結構量化指數能有效預估學生學習成就，其中以PFC指數對於力學學習成就預估效力最高（R=.635，R2＝.404）；2.高成就群學生其知識結構以上位概念為中心且具組階層性，並呈現聯結相關概念的樹狀結構，而這些組型差異影響其物理解題時概念提取與節點活化路徑，進而影響其解題表現。

The purpose of this study is to apply the Pathfinder Method to analyze the mechanics knowledge structure of tenth-grade students. This study investigates the predictive efficacy of learning achievements by knowledge structure indices. It also analyzes both the patterns difference in knowledge structure network representations and the node activation in problem solving. The subjects included 148 tenth-grade students from Kaohsiung district schools. The tools employed in this study included a mechanics knowledge structure assessment, a foundational mechanics achievement test, and an assessment of the analysis of problem solving with thinking aloud. The main findings were: (1) The knowledge structure indices can effectively predict learning achievement; among the three indices the one most able to predict students' learning achievement was found to be the PFC index (R=.635，R2 ＝.404). (2) The knowledge structures of the high-achiever group were found to be organized with a super-ordinate concept in the center, and compared with the low-achiever group were more hierarchical, displaying tree structures that connected related concepts by concept definition. These variations influenced students' ability to extract concepts and their node activation path, and thus influenced their performance in physics problem-solving. These results are thought to have relevance for curriculum design and pedagogy.