Does the detection of misunderstanding lead to its revision?

When dealing with complex conceptual systems, low-prior- knowledge learners develop fragmentary and incorrect understanding. To learn complex topics deeply, these learners have to (a) monitor understanding to detect flaws and (b) generate explanations to revise and repair the flaws. In this research we explored if the detection of a flaw in understanding does lead to the generation of a revising explanation or, alternatively, if there is some independence between the two processes. In two studies participants with low prior domain knowledge learned about plate tectonics from a multimedia presentation. In Study 1 participants in two conditions received support for detection and their performance was compared to that of participants in two control conditions, one receiving no support and another one receiving support for both detection and revision. Performance on retention, transfer, and distortions revealed that the conditions with support for detection were not different from that with no support. The condition with support for both detection and revision was significantly better than all others. In Study 2 participants thought aloud while studying from the presentation. Self-generated detections were not correlated with revising explanations or with performance on retention or transfer measures. Overall, the results indicate that the detection of flaws in understanding does not necessarily lead to the generation of explanations to repair the flaws, meaning that this latter process is very difficult to execute and calls for instructional support.