Katherine McEldoon and her colleagues argue in their new paper, most of these studies are flawed because they failed to control for the extra time spent on self-explanation. So a typical study has compared, say, 30 minutes practice against 30 minutes practice plus time spent on self-explanation. This means any apparent benefit of self-explanation could just be due to extra time spent on studying.
McEldoon's team attempted to avoid this shortcoming. Sixty-nine children, average age 8.8 years, were split into three groups. All had previously struggled with the focus of the study - mathematical equivalence. One baseline group received 50 minutes instruction and practice on solving mathematical equivalence problems (e.g. 6 + 3 + 4 = 6 + _). Another group received the 50 minutes instruction and practice, but they were also prompted to explain why answers to the questions were right or wrong. A final "additional practice" group acted as controls - they received the 50 minutes instruction and practice, and they spent extra time on solving more equations to control for the time taken by the second group on self-explanation. Right after this, and again two weeks later, all the children completed a test of their conceptual understanding and skill at mathematical equivalence problems.
The children in the self-explanation condition showed superior conceptual knowledge compared with the other children, in terms of their knowledge of equation structures (tested with questions like "“Is 8=3+5 true or false?”) but not their understanding of the equals symbol. Their advantage over the additional practice group didn't actually reach statistical significance, though power calculations suggested this could be due to the small sample sizes.
In terms of actual problem solving skill on mathematical equivalence items, the self-explanation group did not differ significantly from the other two conditions. The highest scores were actually achieved by the additional practice group.
Lastly the researchers looked at what's known as "procedural transfer" - the ability of the children to apply themselves to new versions of the mathematical equivalence problems that involved subtraction and the blank being in different position. Here the researchers said the self-explanation group "performed somewhat better" than the other two groups. That is, their scores were higher, but the difference did not reach statistical significance - again possibly due to the samples being too small.
Unfortunately, these results just aren't clear cut enough to provide any solid take-out messages for teachers or parents. More research with larger samples is needed. McEldoon and her colleagues concluded that their findings suggest "self-explanation prompts have some small unique learning benefits, but that greater attention needs to be paid to how much self-explanation offers advantages over other uses of time."
McEldoon KL, Durkin KL, and Rittle-Johnson B (2013). Is self-explanation worth the time? A comparison to additional practice. The British journal of educational psychology, 83 (4), 615-32 PMID: 24175685
Post written by Christian Jarrett (@psych_writer) for the BPS Research Digest.