What the Research Actually Says About Writing Instruction

Writing instruction is one of the most important things secondary teachers do and one of the least reliably grounded in research. Every English teacher has a theory of writing instruction — usually derived from how they were taught, what worked for them personally, or what feels pedagogically coherent. These theories vary enormously, and some of them are contradicted by substantial evidence.

The research on writing instruction is more developed than many teachers know. Graham and Perin's meta-analysis of writing instruction (Writing Next, 2007) identified 11 instructional elements with strong evidence, and subsequent research has deepened that evidence base. Understanding what it shows doesn't mean abandoning professional judgment — it means grounding judgment in evidence.

The Strongest Evidence

Strategy instruction: Teaching students explicit strategies for planning, drafting, and revising — and practicing those strategies with guidance before applying them independently — is the most well-supported approach to improving writing quality. Self-Regulated Strategy Development (SRSD) is the most researched model and has strong effects across grade levels and populations.

Strategies that show consistent benefit:

• Planning before writing (any structured approach: outlines, maps, freewriting, brainstorming)
• Thinking aloud during planning (modeling the decision-making process)
• Explicit revision strategies (category-based checklists, revision protocols)

Collaborative writing: Students who plan, draft, or revise with peers produce better writing than students who work alone throughout. The benefit appears across grade levels and writing types. Peer feedback, when structured, improves both the writing being feedback on and the writing of the student providing feedback.

Increased writing time: Students who write more write better, across grade levels. Schools and classrooms that treat writing as a low-frequency special event produce weaker writers than those where writing is a regular, frequent activity across subjects.

Sentence combining: As noted in grammar research, sentence combining — combining short, simple sentences into complex ones — improves syntactic maturity. It's more effective than grammar instruction in improving sentence-level writing quality.

Process writing: The process writing movement's core insight — that writing is recursive, that revision is not cleaning up a finished product but rethinking the substance — is strongly supported. Students who treat first drafts as exploratory and revision as developmental produce better work than students who treat their first draft as the draft.

Specific feedback: Feedback on specific features of the writing (this paragraph lacks a topic sentence; this claim needs evidence; this transition is confusing) produces better revision than general evaluation (good work; this needs improvement). The more specifically feedback names what to address and where, the more useful it is for revision.

What Has Weaker Evidence

The five-paragraph essay: The five-paragraph essay is the most widely taught writing structure in secondary school and one of the least defensible. There is no evidence that it improves writing quality. It is easy to grade and produces a predictable surface structure, which explains its persistence. It trains students in a form that is not used outside school and actively discourages the organizational flexibility that complex arguments require.

Grammar instruction disconnected from writing: As the grammar research shows, this is reliably ineffective. Students who improve on grammar exercises don't necessarily improve in their writing.

Extensive teacher error-marking: Research by John Hattie and others finds that extensive teacher correction of surface errors on student writing does not improve writing quality. Students often don't read the corrections carefully, don't understand how to act on them, and make the same errors in subsequent writing. The time teachers spend marking errors would be better spent on higher-level feedback or conferencing.

Reading-writing connection assumptions: It is commonly assumed that wide reading automatically improves writing. Wide reading does improve vocabulary and familiarity with syntactic patterns, but the reading-writing transfer is not automatic. Students who read widely but don't write extensively don't necessarily write well. Both activities need to be explicitly taught and practiced.

High-Leverage Practices

Given what the research shows, the highest-leverage practices for writing instruction:

Regular, frequent writing: Multiple short writing tasks per week, across subjects and genres, produce more growth than occasional extended essays.

Explicit strategy instruction with models: Don't just assign writing — teach students how to approach it. Model the planning process. Show what revision looks like. Make the cognitive work visible.

Structured peer feedback: Teach students to give specific, useful feedback. Provide protocols that structure what to look for and how to say it. Practice the process before using it on high-stakes writing.

Conferencing on work in progress: Even brief (5 minute) one-on-one conferences on drafts produce significant improvement. The question "what are you trying to do in this section?" is often more useful than any written feedback.

Mentor texts: Students who read examples of the kind of writing they're learning to produce develop stronger understanding of what the form requires than students who only receive instructions about it.

LessonDraft can help you design writing units, strategy instruction sequences, and feedback protocols grounded in the research on what actually works.

Writing instruction grounded in evidence looks different from writing instruction grounded in habit and intuition. The difference is not always dramatic in any single lesson — but over a school year, systematic application of what the research shows produces substantially stronger writers than approaches that have weak or no evidence behind them.