The linguistic differences in concept conveying in English and Chinese xMOOC forums

Abstract

Recent studies have found that comments from teaching assistants may encourage interactions in edX-like Massive Open Online Course (xMOOC) forums. However, how concepts from these interactions are conveyed to other xMOOC participants has not received much attention. Therefore, this study focuses on a unidirectional teaching assistant-student xMOOC interaction (TS interaction), a content-related pair including one question from a student and one immediate answer from a teaching assistant. The authors particularly investigate the linguistic features (i.e., concept connectivity, concept concreteness, readability and semantic overlap) of concept conveying in TS interactions with many responses (mTS) and with few responses (fTS). In addition, a language factor (English and Chinese) is also considered. Additionally, the interaction transcripts from science lectures (SL) and political briefings (PB) were used as control groups as two opposite cases of concept conveying. At the concept level, the concept conveying in transcripts were modelled as a graph, and measured by common indicators in graph theory. At the overall level, the concept conveying in transcripts were measured by regular linguistic measuring tools. The results show that interactions with mTS and fTS demonstrate different concept conveying tendencies toward SL and PB in terms of linguistic features in both languages. The results suggest that in both languages, teaching assistants may use mixed concept-conveying strategies to stimulate more follow-up responses in xMOOC forums. These conclusions drawn from TS interactions can be even partially generalized in a larger student-student (SS) interaction dataset.