2023 ASEE Annual Conference & Exposition

Computer-supported collaborative learning (CSCL) is a widely known and used teaching and learning approach in higher education online and hybrid scenarios. When planning an online or hybrid session from a didactic point of view, in most cases only the maximum capabilities of the planed CSCL-tool are considered. However, participants differ in their interindividual tool usage, e.g., webcam usage, due to personal or technical reasons. In result, a CSCL-session planned on a web-conferencing platform can unintentionally turn into a session on a spectrum from videoconferencing (all participants use their webcam and microphone) over audioconferencing (participants refusing webcam usage) to synchronous text-chat (webcam- and microphone-refusal). In worst case this can cause misleading conclusions about the didactic match between tool and task with negative effect on teaching and learning. To consider the users’ interindividual tool usage, we conducted an online experiment with 45 undergraduate students building 15 three-student groups performing a murder mystery based on hidden-profile-paradigm. The murder mystery task simulated a typical CSCL-session task in higher education, where an ad-hoc small-group of three students collaborate by pooling shared and unshared knowledge into a common solution. This data-driven approach targeted to evaluate potential differences in (a) task performance, (b) interaction-process quality, and (c) mental workload between CSCL-sessions intentionally planned to be performed on a web-conferencing platform by the lecturer for didactic reasons but participants refuse to use webcam or microphone. The gained insights should serve to define CSCL related policies and practices, which are conducive to learning. Despite from theory and prior empirical findings deduced expected differences, no evidence of superiority of one of the three experimental conditions (videoconferencing, audioconferencing, and synchronous text-chat) could be observed in this contribution. Possible reasons for this result, limitations of this study, and practical implications are discussed.