A study on students’ interactive characteristics in online-personalized tutorship

Online question-and-answering (Q&A) is very important for students during online learning. Study on the interactive characteristics of online Q&A dialogues can provide useful references for teachers to design high-quality online Q&A strategies. In this study, for 31420 online Q&A dialogue records from an intelligent tutoring system, clustering was conducted based on three dimensions of dialogue frequency, duration, and content, and by defining and calculating the sentence entropy to measure discourses quality. Online Q&A dialogue presents three different types of interaction categories: low, moderate and high, and the higher interaction category corresponds to the poorer students’ knowledge level. The different categories of interactions reflect three states of Q&A dialogue sequences, namely, “little instruction, well understanding”, “precise instruction”, and “active discussion”. Teacher’s discourse quality in the first Q&A activity of students can significantly affect their subsequent interaction states. Those findings can be useful for teachers to design high-quality Q&A strategies.