Automatic representation of knowledge structure: enhancing learning through knowledge structure reflection in an online course

Abstract

Summary writing is an important skill that students use throughout their academic careers, writing supports reading and vocabulary skills as well as the acquisition of content knowledge. This exploratory and development-oriented investigation appraises the recently released online writing system, Graphical Interface of Knowledge Structure (GIKS) that provides structural feedback of students’ essays as network graphs for reflection and revision. Is the quality of students’ summary essays better with GIKS relative to some other common approaches? Using the learning materials, treatments, and procedure of a dissertation by Sarwar (Doctoral Thesis, University of Ottawa, 2012) but adapted for this setting, over a three-week period Grade 10 students (n = 180) read one of three physics lesson texts each week, wrote a summary essay of it, and then immediately received one of three counterbalanced treatments including reflection with GIKS, solving physics problems as multiple-choice questions, and viewing video information, and finally students rewrote the summary essay. All three treatments showed pre-to-post essay improvement in the central concepts subgraph structure that almost exactly matched the results obtained in the previous dissertation. GIKS with reflection obtained the largest improvement due to the largest increase in relevant links and the largest decrease for irrelevant links. The different treatments led to different knowledge structures in a regular way. These findings confirm those of Sarwar (2012) and support the use of GIKS as immediate focused formative feedback that supports summary writing in online settings.