Effects of Pairing Methods Based on Digital Textbook Logs and Learner Artifacts in Conceptual Modeling Exercises

Introduction

From the beginning of 2020, educational sites were forced to conduct online lessons due to COVID19. Lectures and exercises were conducted using various digital technologies, such as Zoom and Microsoft Teams. In addition, digital textbook systems have been introduced into schools and universities to improve education and learning by collecting and analyzing educational big data. Many educational researchers have focused on analyzing the digital textbook logs in order to enhance teaching and learning (Mohammad et al., 2018; Mouri et al., 2016 and 2018; Shimada et al., 2017; Ogata et al., 2017).

Exercises are also being conducted after teaching lectures, such as database design in computer science departments. Conceptual modeling activities are inevitable when designing a database as they help grasp a whole system, such as relations among entities and an attribute in an entity. Tanaka et. al. (2016) developed a conceptual modeling tool called KIfU 3.0 to support such activities. By using the system, learners can create entities, attributes, and relations based on system requirements. In addition, the system can collect operation logs such as “Create,” “Edit,” “Save,” and “Undo.” They reported that the system is useful in carrying out conceptual modeling activities, but the problem of the differences in the understanding levels among the learners tends to increase. To solve the problem regarding the differences in the understanding levels among the learners, many researchers have focused on introducing group work to exercises (Lage et al., 2000; Mori et al., 2009). Ike et al. (2018) proposed a pairing algorithm in the group work of the conceptual modeling activity using operation logs collected in KIfU 3.0. Although they showed that the heterogeneous pairing method with different characteristics is effective for enhancing learning effects, the following aspects are yet to be explored: (1) proposing an effective pairing method based on both digital textbook logs and conceptual modeling activity logs and (2) verifying the effects of (1).

A pairing method utilizing digital textbook logs and conceptual modeling activity logs is proposed for this study. An experiment was conducted to measure the effectiveness of the proposed pairing method in comparison with the random pairing method.

Research Related To Digital Textbook Systems For Learning Analytics

In recent years, digital textbook systems have been introduced to schools and universities in many countries. Most pilot studies on digital textbook systems focused on learning effects and educational digitalization by introducing a digital textbook system. In the field of educational big data and learning analytics, researchers have focused on analyzing and visualizing digital textbook logs to improve teaching and learning. The research team of Ogata et al. tackled introducing digital textbooks at the university level and analyzed digital textbook logs (Ogata et al., 2015; Mouri et. al., 2021; Majumdar et. al., 2021). The purposes of this research are as follows: Analyzing and visualizing digital textbook logs to improve learning materials (Mouri, Yin, & Uosaki, 2018; Mouri & Yin, 2017) and identifying students who are likely to fail or drop out (Okubo & Yamashita et al., 2017). Based on the results of the analysis, the method can find the points to be improved in the digital textbooks. By giving the visualization results to teachers and students, the latter could change their learning to obtain better learning achievements.