Abstract
This study investigates the possibility of utilizing online learning data to design face-to-face activities in a flipped classroom. We focus on heterogeneous group formation for effective collaborative learning. Fifty-three undergraduate students (18 males, 35 females) participated in this study, and 8 students (3 males, 5 females) among them joined post-study interviews. For this study, a total of 6 student characteristics were used: three demographic characteristics obtained from a simple survey and three academic characteristics captured from online learning data. We define three demographic group heterogeneity variables and three academic group heterogeneity variables, where each variable is calculated using the corresponding student characteristic. In this way, each heterogeneity variables represents a degree of diversity within the group. Then, a two-stage hierarchical regression analysis was conducted to identify the significant group heterogeneity variables that influence face-to-face group achievement. The results show that the academic group heterogeneity variables, which were derived from the online learning data, accounted for a significant proportion of the variance in the group achievement when the demographic group heterogeneity variables were controlled. The interviews also reveal that the academic group heterogeneity indeed affected group interaction and learning outcome. These findings highlight that online learning data can be utilized to obtain relevant information for effective face-to-face activity design in a flipped classroom. Based on the results, we discuss the advantages of this data utilization approach and other implications for face-to-face activity design.
Original language | English |
---|---|
Pages (from-to) | 2055-2071 |
Number of pages | 17 |
Journal | Educational Technology Research and Development |
Volume | 68 |
Issue number | 5 |
DOIs | |
State | Published - 1 Oct 2020 |
Keywords
- Activity design
- Flipped classroom
- Group heterogeneity
- Learning analytics
- Online learning data