We’ve known for quite a while now that (direct) instruction is important for learning, especially for children who have less background on the matter or who have troubles catching on (and this can be for various reasons). This is not an easy task, as this can be not the prime wish of both teacher and pupil. Discovery learning can be much more fun, but the effects can be counterproductive (check also here).
This new study I found via @dylanwilliam discusses the element of when instruction should be done. The results of this study contrast with a growing literature suggesting that instruction should be delayed (often in favor for discovery learning). The researchers see 3 reasons why this can be sometimes a bad idea:
- the type of instruction might matter: when instruction is solely conceptual in nature, it can be useful to provide the instruction first.
- the type of self-explanation prompts may also influence the timing of instruction: providing conceptual instruction first may be particularly effective if learners are supported in using this information via the use of conceptual self-explanation prompts.
- the activation of misconceptions during problem solving may impact the timing of instruction: if problem solving is difficult due to the activation of misconceptions, prior instruction is likely necessary to make children’s effort and struggle more productive.
(paraphrased from the article)
Abstract of the research:
The sequencing of learning materials greatly influences the knowledge that learners construct. Recently, learning theorists have focused on the sequencing of instruction in relation to solving related problems. The general consensus suggests explicit instruction should be provided; however, when to provide instruction remains unclear.
We tested the impact of conceptual instruction preceding or following mathematics problem solving to determine when conceptual instruction should or should not be delayed. We also examined the learning processes supported to inform theories of learning more broadly.
We worked with 122 second- and third-grade children.
In a randomized experiment, children received instruction on the concept of math equivalence either before or after being asked to solve and explain challenging equivalence problems with feedback.
Providing conceptual instruction first resulted in greater procedural knowledge and conceptual knowledge of equation structures than delaying instruction until after problem solving. Prior conceptual instruction enhanced problem solving by increasing the quality of explanations and attempted procedures.
Providing conceptual instruction prior to problem solving was the more effective sequencing of activities than the reverse. We compare these results with previous, contrasting findings to outline a potential framework for understanding when instruction should or should not be delayed.