The Puzzling Gap In Research On Writing Instruction
There's little or no data on whether it makes sense to start at the sentence level.
There isn’t much good research on writing instruction. And the research that does exist has overlooked a key question: Should we teach students to write sentences before asking them to write at length?
Writing may be the hardest thing we ask students to do in school. Studies have shown that writing places a huge burden on “working memory,” a term used to describe what happens in the mind when it processes information.
Inexperienced writers may be trying to juggle everything from forming letters of the alphabet (if they’re young) to choosing words to use, spelling them, structuring sentences, and generally organizing their thoughts. Working memory can hold only about four items for perhaps 10 to 15 seconds before becoming overwhelmed. The best way to overcome the limitations of working memory is to have more information stored in long-term memory, which is virtually infinite. Once you know the mechanics of writing—and you’re familiar with the content you’re writing about—the task becomes easier.
Writing is much harder than reading—difficult as that is. On national writing tests, only about a quarter of students score proficient or above. That’s a lower proportion than the one-third or so who score proficient on reading tests. And yet researchers have paid far more attention to reading than writing.
Large-scale analyses of writing instruction studies have shown that certain approaches can have positive effects: teaching the “process” of writing—including planning, drafting, and revising—or having students collaborate. There are many studies of an approach called SRSD (which stands for Self-Regulated Strategy Development) showing that it’s effective, although there’s debate over how many qualify as rigorous.
A hugely popular curriculum developed by literacy guru Lucy Calkins, often referred to as Writers’ Workshop, is the subject of a recently released rigorous evaluation—the first ever conducted of the approach, despite the fact that it’s been around for decades. Researchers found significant gains beginning in the second year of implementation, although they were smaller for students from lower-income families and other vulnerable groups.
But, as education writer Jill Barshay has pointed out, the same approaches that prove successful in some studies prove unsuccessful in others, and no one approach is clearly superior. It can also be hard to tell why a multi-component approach works. The peer editing? The focus on drafting? Both? In the case of Writers’ Workshop, the study lumped it together with Calkins’ other popular product, Readers’ Workshop, so it’s impossible to discern the effect of the writing program alone.
To be fair, it’s hard to conduct a study that points to a clear causal relationship between a particular method of instruction and a better outcome. The kind of research allowing for that—a randomized controlled trial—requires identifying two large and similar groups of students, only one of which is exposed to the method, and following them while trying to control for confounding factors (like teachers who deviate from the method). And especially with writing, it can be difficult to determine which outcome is “better.” Fortunately, a new approach to evaluation called comparative judgment holds the promise of greater reliability.
Still, there’s a puzzling gap in writing research: examining whether it’s best to start instruction at the sentence level. There are reasons to suspect that approach would help:
· Writing imposes a heavy burden on working memory, especially for inexperienced writers.
· Writing at length increases that burden. (I’m not aware of much research to support that, but it seems like common sense.)
· Having inexperienced writers stick to writing sentences should therefore free up space in working memory that could be used to learn writing conventions and analyze what’s being written about.
So why is there no research looking into this? Perhaps because almost all educators and researchers have assumed that children need to write at length from the get-go, in part to develop writing fluency, “voice,” and stamina. But it’s hard to become fluent or develop a voice if you’re overwhelmed by the mechanics of writing.
Some approaches spend time on sentence-level skills but still expect novice writers to compose multiple paragraphs. Advocates for that practice may cite the many studies showing that teaching grammar—and sentence construction skills—in the abstract doesn’t improve student writing. That’s true, but one kind of sentence-level activity, sentence combining, has been shown to produce good results. Instead of just learning rules, students take several short sentences and combine them into one longer, more complex sentence.
Still, some may object, if students are just combining sentences, they aren’t leveraging the power of writing to build and deepen their knowledge. To be sure, there’s evidence that when students write about the content they’re learning, their comprehension improves; that’s the basis of an approach called “writing to learn.” And it’s true that sentence combining—and some other sentence-level activities—probably won’t have that effect. But some sentence-level activities can provide powerful boosts to learning if they’re grounded in the content of the curriculum.
Consider something as basic as helping students understand the concept of a complete sentence as opposed to a fragment—a concept that eludes even some college students and adults. Providing the definition of a sentence as “a group of words that includes a subject and a predicate and expresses a complete thought” is too abstract for many. Instead, a teacher can give students lists that include both fragments and complete sentences and have them practice determining which is which. Then—and here’s where the knowledge-building comes in—students need to add information to convert the fragments into sentences.
At the same time, students are learning to use conventions like capitalization and punctuation. Studies have found that teaching grammar in the abstract isn’t effective, but teaching grammar in the context of students’ own writing is—and it’s a lot easier to do that at the sentence level than in the context of pages of error-filled writing.
This activity can be used to build knowledge in any subject. For example, a math teacher could include the fragment can be expressed as a fraction or a ratio. Students who have been learning about rational numbers would be able to turn that into the sentence, A rational number is a number that can be expressed as a fraction or a ratio. In the process, they’ll be retrieving information they may have slightly forgotten and putting it into their own words, thereby boosting their comprehension and retention of information. In fact, sentence-level activities like this should be more effective in building knowledge for inexperienced writers than writing at length, because they free up precious space in working memory.
Completing sentence fragments is one of many knowledge-building sentence activities that are part of a method called The Writing Revolution. There may be other approaches that are designed to be grounded in the content of the curriculum and begin at the sentence level—while holding off on writing at length—but I’m not aware of any. I’m the co-author, with Judith C. Hochman, of a book explaining the method and I’m on the advisory board of the organization that disseminates it, so I’m not entirely objective. But I have no financial interest in the book or anything relating to the method.
The Writing Revolution doesn’t stop at the sentence level—it goes all the way through the argumentative essay. But the groundwork for that kind of writing is laid through well-designed practice with sentences. And while students may not write at length from the beginning, they’re not limited to working with sentences. Even very young children can use the method to learn how to do things like plan and outline paragraphs orally, under a teacher’s guidance. I once saw a kindergarten class collaboratively outlining a paragraph (it happened to be about Play-Doh, but it could have been grounded in curriculum content), learning how to think logically—and completely engaged in the effort.
It’s true there are studies supporting other approaches, especially SRSD, but it’s important to consider what they’ve been evaluated against. We don’t yet have studies of the impact of an approach that, like The Writing Revolution, starts with knowledge-building sentence-level activities. Given our general lack of progress in improving students’ writing, maybe it’s time we did.
This post originally appeared on Forbes.com.