As defined by the authors of a recent study on the subject, “discovery learning” involves asking students “to develop individual solutions to a problem before receiving instruction” (all quotes are from the original article.) The rationale for this reversal of the traditional approach lies in the “productive failure” approach according to which “students who engage in problem-solving activities targeting yet-to-be-learned concepts usually generate erroneous or incomplete solution attempts, which can form a basis for developing valid knowledge during subsequent instruction.”
The beneficial effect of discovery learning and the productive failure approach is often attributed to the fact that authentic learning implies a conceptual shift, i.e., a revision by students of their intuitive ideas about a particular question. Without this updating process, new knowledge cannot be integrated into the students prior knowledge, and learning is superficial.
However, discovery learning might not automatically lead to such productive failures without explicit instructions to elaborate on correct and incorrect answers. To test this hypothesis, a team of researchers conducted an experiment on 200 German fifth-graders, asking them to answer a mathematical problem on a topic they had never studied befire (fraction comparisons.) During the problem-solving phase, all students were asked to select the winning team in a throwing contest in which
In the second condition, the students were given 2 erroneous answers ans asked to perform a similar elaboration.
In the third condition, the students were given both correct and incorrect answers, asked to compare them and to elaborate.
Then, the participant were tested on 3 similar problems and asked to explain their reasoning.
Results indicated that the group asked to reflect on erroneous answers did not perform better than the group who had been given the correct answers directly. However, the group prompted to compare correct and incorrect answers did outperform both.
As explained by the researchers: “Students detect flaws in their mental models when comparing the correct and incorrect attempts and consequently revise their mental models based on the correct solutions.” Or to put it otherwise, “confrontation with errors alone does not allow the full potential of learning from errors to unfold.”
Reference: Loibl and Leuders (2019) “How to make failure productive: Fostering learning from errors through elaboration prompts”, Learning and Instruction, 62, pp. 1-10.