If you’ve got a headache, you might find yourself reaching for some ibuprofen. If you just sprained your ankle, or your arthritis is acting up, ibuprofen can make your day much better. But: these truths do NOT mean that ibuprofen fixes pain. After all, nerve pain — such as sciatica, neuropathy, or neuralgia — doesn’t relent much after an Advil or two. Those little red pills help with some pain, but not all types of pain.
In research world, we call such limitations “boundary conditions” — as in, “this study technique helps under these conditions, but not those conditions.” Good research-based teaching advice makes boundary conditions explicit. After all, no one teaching strategy helps all students learn everything all the time.
Today, I want to focus on “self-explanation”: a teaching technique that does indeed have lots of good research behind it. Because we’ve got so many reasons to use “self-explanation,” we should be aware of the relevant boundary conditions as well. Happily, I’ve found a review article that highlights those essential boundaries…
Core Benefits
Education researchers define self-explanation as “generating explanations for oneself in an attempt to make sense of new information.” One formulation emphasizes that self-explanations are generated “by the learner and for the learner.” That is: “teaching” isn’t “self-explanation” because the person who generates the explanation (the teacher) does so for someone else (the student).
The review article’s authors highlight two kinds of benefit from self-explanation.
- First: self-explanation helps students integrate knowledge. By putting all the pieces together in their own words, students reinforce connections among them.
- Second: self-explanation tends to direct students’ attention to conceptual structures over surface features and details. For this reason, students succeed at transfer tasks more readily after self-explanation.
As you can see, those are powerful reasons for teachers to incorporate self-explanation activities in our classes. Anything that promotes integration of knowledge and transfer of knowledge … well, that sounds like a winner!
Boundaries Beyond Benefits
If all research-based teaching advice has boundaries, we should ask: what are the limits for self-explanation? In their review article, Rittle-Johnson and Loehr highlight a few boundaries that we should keep in mind.
As is so often the case, a strength of self-explanation is also a weakness. As noted above, self-explanation highlights conceptual structures over details. For that reason, it serves well in cases where the teacher wants students to grasp a broad concept instead of a specific example.
In my math class: when I ask my students to express 127 in base 8, I want them to understand the broader principles and processes of thinking in different bases. I don’t really want them to remember the fact that 127 (base 10) is 177 (base 8).
However, sometimes we teach a topic where we DO want students to recall the granular specifics.
- If I’m teaching Elizabeth Bishop’s poem “One Art,” I want my students to focus on and remember the increasingly weighty details — lost door keys, a lost watch, a lost continent — that build up to the poem’s wrenching conclusion. I don’t want them simply to recognize Bishop’s poetic structure (a “vilanelle”) and move on.
- If I’m teaching about Botswana’s independence, specific events and people matter. Students should know the first President Seretse Khama; his focus on constitutional norms and education; the timely discovery of diamonds; the pursuit of moderate economic policies, and so forth. Students should be able to abstract from those details to see broader patterns … but they need to recall those examples as well.
When deciding if I should ask my students to self-explain, I should first decide if I want students to consolidate the deep rules and patterns, or if I want them to remember the piece-by-piece details as well. In the latter case, some other teaching move might be more effective than self-explanation.
Explaining What? In Response to What?
A second boundary that Rittle-Johnson and Loehr emphasize: the content that students are self-explaining.
Unsurprisingly, in most studies that Rittle-Johnson and Loehr review, students benefit most from self-explanations of correct information. If students self-explain beliefs based on prior misconceptions, their self-explanation may consolidate meaningful inaccuracies, and thereby make it harder for students to learn the correct answer, approach, or concept. In this case, the self-explanation technique could very well backfire.
For this reason, teachers should evaluate the timing and content of self-explanation questions. If today’s classroom topic invites common misconceptions, I should explain correct answers first before I invite my students to self-explain. If I start with self-explanations, my lesson may reinforce prior misconceptions rather than correct them.
Third: Rittle-Johnson and Loehr emphasize that self-explanation prompts focus students’ attention on different parts of the cognitive work that they’re doing. In brief:
- a “why” prompt helps students focus on conceptual understanding
- a “how” prompt consolidates procedural fluency
- a “why is this wrong” prompt helps repair misconceptions
Teachers should ask not only “should I have students self-explain right now?” but also “what’s the best question to focus my students’ cognitive work precisely?”
TL;DR
Like all teaching techniques, research-informed strategies work in some cases…but not all cases.
Self-explanation can help students consolidate and transfer knowledge. To use this approach most effectively, teachers should
- use self-explanation to help students consolidate conceptual structures more than granular details,
- correct prior misconceptions before asking self-explanation questions, and
- prompt self-explanation with targeted questions.
When teachers stay within these boundary conditions, our students reap the greatest benefits.
Rittle-Johnson, B., & Loehr, A. M. (2017). Eliciting explanations: Constraints on when self-explanation aids learning. Psychonomic bulletin & review, 24(5), 1501-1510.
