The effects of conversational language on group interaction and group performance in computer-supported collaborative argumentation

This study examined the effects of conversational language (e.g., asking questions, inviting replies, acknowledgments, referencing others by name, closing signatures, º agree, but', greetings, etc.) on the frequency and types of responses posted in reply to given types of messages (e.g., argument, evidence, critique, explanation), and how the resulting response patterns support and inhibit collaborative argumentation in asynchronous online discussions. Using event sequence analysis to analyze messageresponse exchanges in eight online group debates, this study found that (a) arguments elicited 41% more challenges when presented with more conversational language (effect size .32), (b) challenges with more conversational language elicited three to eight times more explanations (effect size .12 to .31), and (c) the number of supporting evidence elicited by challenges was not significantly different from challenges that used more versus less conversational language. Overall, these and other findings from exploratory post-hoc tests show that conversational language can help to produce patterns of interaction that foster high levels of critical discourse, and that some forms of conversational language are more effective in eliciting responses than others.