Network mapping of asynchronous student interactions in online discussion boards

Abstract

The recent movement towards asynchronous learning activities in higher education has transformed student and educator interactions. However, as the efficacy of such approaches are unknown, many educators must rely on personal experience and peer anecdotes regarding student engagement without underlying data to inform assessment. As educators transition from pandemic‐driven pedagogy to sustainable models of course design, the use of course‐based analytics can help instructors understand the ways in which students are interacting with asynchronous activities and each other.To develop fluency in reading and understanding scientific journal articles and to provide an interactive forum for discussions between students, weekly discussion boards were assigned in a graduate‐level neuroscience course. Based on an assigned article, each student was required to post an original response to the reading and reply to two posts made by their peers. These discussion board interactions were analyzed to form a social network map of asynchronous student interactions. Of the 27 students initially enrolled in the course, 24 completed it. The 3 students that did not complete the course were included as nodes and edges but excluded from statistical analysis.During the 9‐week semester (9 original posts, 18 required peer replies), students responded well to instructions about ensuring that the majority of posts received peer interaction, with 91.5% of all original posts receiving at least one reply and 28.7% of posts receiving 3 or more replies. Prestige or indegrees, measured as the average number of replies each student received, was 17.96 (SD 6.39). Centrality or outdegrees, measured as the average number of replies each student made, was 17.92 (SD 1.64). Students connected to an average of 45% (SD 0.07) of their peers. There was no correlation between student quiz performance and either prestige or centrality. Students that tended to post earlier saw a non‐significant trend to receive more peer replies than those who tended to post later. This trend became significant (p < 0.001) when analyzing the correlation between absolute post order and number of replies rather than the average post order of each student.These findings provide evidence regarding student engagement behaviors as they pertain to online discussion boards. Students generally did not exceed the mandated minimums for full points, and in some cases, self‐reported through end‐of‐semester feedback that they strategically engaged either to maximize earned points or to minimize overall effort. Collectively, in their feedback, the students provided deep and nuanced feedback on the activity experience. Overall, while many students noted the time required and initial difficulty in regularly reading scientific literature, 92% of responses indicated that the student found value and utility in reading and discussing the scientific papers in this format, supporting the use of this activity in future versions of this course.