This study examined how message-response exchanges produced in the interactions between active learners only, reflective learners only, active-reflective learners and reflective-active learners affected how often active versus reflective learners posted rebuttals to arguments and challenges across four types of exchanges that believed to promote critical discourse (argument-challenge, challenge-counterchallenge, challenge-explain, challenge-evidence) in computer-supported collaborative argumentation (CSCA). This study found that the exchanges between reflective learners produced 44% more responses than in the exchanges between active learners (ES = +0.17). The reflective-reflective exchanges produced 47% more responses than the active-reflective exchanges (ES = +0.18). These results suggest that groups with reflective learners only are likely to produce more critical discourse than groups with active learners only, and the ratio of active-reflective learners within a group can potentially influence overall group performance. These findings illustrate how specific traits of the learner can affect discourse processes in CSCA and provide insights into process-oriented strategies and tools for structuring dialogue and promoting critical inquiry in online discussions.