Control and collaboration in multimedia learning: Is there a split-interaction?

Work presented to obtain the « Diplôme d'études approfondies en psychologie cognitive expérimentale » Abstract This work presents an empirical research on the comprehension and learning from animated pictures explaining the functioning of dynamic systems (tectonic and astronomic phenomenon). In a previous study we found that learners studying in pairs benefited more of animated presentations than individual learners. But, participants had lower results when the interface contained interactive devices. We explained the results as a possible flawing of the learning process caused by the concurrent presence of two types of interaction (with the peer and with the screen). In order to further investigate this so called " split-interaction effect " , we designed the present experiment. Two factors were used; the first one was the effect of user's control on the pace of the animation (low vs. high control). The second factor was the learning setting, depending if participants learned the experimental material individually or collaboratively. Results were inconclusive and not in favour of our split-interaction hypothesis. Participants in collaborative and low control condition obtained lower retention scores than other groups. This result goes against the split-interaction effect. Other analysis were led and showed that our factors, but not individual cognitive abilities, influenced the way participants explored the materials. Nevertheless the differences in the material exploration did not led to differences of understanding. These results question the existence of a split-interaction effect but also the importance of delivery features as opposed to instructional design questions.