Mayer, R. E., Deleeuw, K. E. & Ayres, P. (2007). Creating retroactive and proactive interference in multimedia learning. Applied Cognitive Psychology, 21, 795-809.

When students learn how a physical system works, does it help if they also learn how similar systems work? Some college students (concise group) studied a multimedia lesson that explained how hydraulic brakes work, consisting of narrated animation or annotated illustrations. Others (expanded group) received the same lesson along with multimedia explanations of caliper brakes and air brakes presented either after (Experiment 1) or before (Experiment 2) the explanation of hydraulic brakes. Across the combined experiments, students performed worse on retention (d = 0.37) and transfer tests (d = 0.30) concerning hydraulic brakes if the lesson also contained lessons on caliper and air brakes; within each experiment, the effects were statistically significant in Experiment 1 (d = 0.57 and 0.53, respectively) but not in Experiment 2 (d = 0.23 and 0.17, respectively). Students performed similarly with narrated animations and annotated illustrations. These results suggest that if students are expected to learn about a specific system, using examples about related systems can depress learning, particularly in the form of retroactive interference (in Experiment 1). These results extend the coherence principle, that is, the idea that adding extra material to an explanation can interfere with learning. There was no evidence to support converting static graphics into animation.