Research on Learning Behavior Pattern Analysis of Expertise Development Facilitated by a Visualization-Learning Tool
Bian WU,Xiaoqing GU
DOI: https://doi.org/10.3969/j.issn.1007-2179.2014.05.008
2014-01-01
Abstract:In various professional domains such as medicine, law, education, and computer engineering, how those domain novices can grow up to become experts remains unclear. Researchers agreed that a learner should keep deepe-ning his or her domain knowledge,understanding and sharpening problem-solving skills ( Ericsson, 2008; Lajoie, 2003 ) . They argued that professional expertise development lies in the interaction between two complex cognitive activi-ties, i. e. , knowledge construction and problem solving. However, these two activities are implicit, which often becom-esan obstacle for evaluation and facilitation of expertise development. To overcome this challenge, a computer-suppor-ted visualization tool can help learners externalize cognitive activities during learning, which in turn, is an effective way to facilitate expertise development and also a new perspective to assess professional expertise. As a typical example of learning analytics, a learner's behavior mode in a visualization-learning environment can help shed light on the development of professional expertise and provide effective scaffolds. Therefore, the focus of this study was to design an innovative approach to investigate the learning behavior pattern of expertise development. First, this study designed a visualization tool called dual-mapping tool that includes a concept map to represent knowledge construction process and an argument map to represent problem-solving process during learning. Second, as learning analytics method is concerned, this study proposed to use a sequential analysis approach to analyze visualization-based learning process data and explore the learning patterns in support of expertise development. Clinical diagnostic problem solving in medi-cal education was selected as an example to justify the effectiveness of this learning analytics approach on visualization learning. This study recruited 14 medical students from a medical school in Shanghai for one-month online learning during their spare time and introduced the dual-mapping learning tool in support of their medical problem-based learning ( PBL) . Participants were required to solve four authentic kidney-related problems online and constructed four dual-maps respectively. Dual-mapping learning products were rated by domain experts and classified into either good or bad performance groups. Further, online visualization learning behavior was analyzed to explore the learning behavior patterns and to compare the differences in learning patternsbetween good and bad performance group. The study found that learners with good learningperformance showeda unique learning pattern of a three-step cycle, i. e. , concept elicitation-hypothesis generation-reasoning& justification, which suggests their better abilities in knowledge summarization and problem-solving-oriented knowledge construction during learning. The finding suggests this spe-cific visualization-learning behavior pattern may facilitate expertise development. Implications and directions for fur-ther study were discussed at the end.