How to Correct Grammar Without Killing Student Confidence

Every English teacher has had this experience: you spend hours marking a stack of essays in red pen. You hand them back. Students glance at the grade, ignore the comments, and shove the papers in their backpacks.

What if I told you the research suggests the red pen approach not only fails to improve writing — it actively suppresses risk-taking and stunts long-term language growth?

This post is the alternative. A research-backed framework for correcting grammar that produces measurable writing growth and preserves student confidence. Drawn from Lyster and Ranta's seminal 1997 corrective feedback typology, Hattie and Timperley's Power of Feedback (2007), and Ferris's Treatment of Error in Second Language Student Writing (2nd ed., 2011).

The Truscott vs. Ferris Debate

In 1996, John Truscott published The Case Against Grammar Correction in L2 Writing Classes (Language Learning), arguing that grammar correction doesn't improve student writing — and may even harm it.

Dana Ferris responded with two decades of research showing that some kinds of grammar correction do produce gains. The current consensus, from the IES Practice Guide on writing (2017) and subsequent reviews:

• Comprehensive error correction (marking everything) doesn't work and can demotivate students.
• Focused error correction (one or two specific patterns) produces measurable, transferable gains.

The Five Types of Corrective Feedback

From Lyster and Ranta's classic 1997 study:

1. Recast. Teacher restates the correct form. Student: Yesterday I go to store. Teacher: Oh, yesterday you went to the store?
2. Explicit correction. Teacher names the error and provides the correct form. That's wrong. It should be 'went.'
3. Clarification request. Teacher asks for clarification, implying something was wrong. I'm sorry, what?
4. Metalinguistic feedback. Teacher provides a comment about the error. That's a tense problem.
5. Elicitation. Teacher prompts student to self-correct. Yesterday I ___?

Lyster and Ranta's finding: Recasts produced the lowest student uptake and self-correction. Elicitations and clarification requests produced the highest. In other words, the more cognitive work the student does, the more they learn.

What This Means for Grading Stacks of Essays

Stop marking every error. Instead:

1. Identify each student's most common error pattern from this draft. (Takes 30 seconds per essay.)
2. Mark only that one pattern.
3. Use elicitation, not correction. Underline the error and write the category in the margin: VT for verb tense, SVA for subject-verb agreement, RO for run-on. Don't write the correction. Make the student fix it.
4. Require revisions for credit. The grammar mark means nothing if the student doesn't engage with it.

Total time: about the same as comprehensive marking. Learning gain: significantly higher.

The Five-Minute Error Conference

When a student has a deeply fossilized error, written feedback isn't enough. Conference with them. Five minutes per student, once a week, on rotation:

Min 1: Sit next to the student with their draft.
Min 2: I noticed this pattern. Point to 3 examples.
Min 3: Tell me what's happening here. Wait. Let the student do the cognitive work.
Min 4: Co-construct a fix. What could you change?
Min 5: Now find the other 4 examples and fix them. I'll check tomorrow.

This conference produces more learning than 30 minutes of written marking. See our grammar conferencing strategies post.

The Confidence Side of the Equation

Grammar correction doesn't have to crush confidence — but it often does, because of how it's delivered. Three moves that preserve confidence:

1. Name what's working first. Your topic sentence is strong. The evidence is well-chosen. Let's focus on one pattern that's holding you back.
*2. Frame errors as patterns, not failures. Lots of students do this. Here's the pattern. Here's how to fix it.*
3. Make corrections low-stakes. Don't grade first drafts on grammar. Grade revisions on the specific feature you asked them to revise for.

What Research Says About Praise

Hattie and Timperley's 2007 review of feedback is clear: praise has near-zero impact on learning. Specific, task-focused feedback has high impact.

So instead of Good job!, write You used three complex sentences in this paragraph — exactly what we've been practicing. The student knows what they did well and what to keep doing.

Praise the work, not the writer. Specific, task-focused.

The Color-Code System

A useful tool for written feedback at scale:

Green: This sentence shows what we've been working on. Keep doing this.
Yellow: This is the target error pattern for revision. Fix it.
Underline only: Other errors. Don't fix now; we'll get to them later.

Students learn to pay attention to yellow. Other errors don't disappear — they get addressed in other weeks.

Common Mistakes in Correction

Mistake 1: Recasting in class discussion. Student says yesterday I go to store. Teacher says Oh, you went to the store? Research is clear: recasts barely register. Use elicitation: Yesterday I ___?

Mistake 2: Correcting in front of peers. Public correction triggers shame, which inhibits future risk-taking. Save individual corrections for individual conferences.

Mistake 3: Correcting beginners' production. Newcomers in the silent period or early production benefit from more input, less correction. Save heavy correction for intermediate and advanced students.

Mistake 4: Inconsistency. If you correct an error one week and ignore it the next, students learn the rule is arbitrary. Pick patterns; stick with them.

A Sample Feedback Routine That Works

Step 1: Read the draft for content first. Note 2 strengths.
Step 2: Identify the most common grammatical error pattern.
Step 3: Write 1 strength + 1 pattern note at the top.
Step 4: Mark 5–7 instances of the pattern using elicitation (underline + margin code).
Step 5: Require a revision focused on that pattern.
Step 6: Grade the revision, not the first draft.

Time: 4–6 minutes per essay. Learning gain: substantially higher than comprehensive marking.

Feedback for Newcomers

Newcomers need:

• More feedback on content and effort.
• Less feedback on grammar.
• Specific, model-based feedback when given.

Sample: I love that you wrote about your sister! Here's a model sentence you could try: My sister is 12 years old. Try writing one more sentence using this pattern.

Build, don't tear down.

What Research Says About Self-Editing

Ferris (2011) shows that students who are trained to self-edit produce more accurate final drafts than students who receive teacher correction alone.

Self-editing training is concrete:

• Show students a checklist of common errors.
• Walk them through self-editing one of their own drafts.
• Require self-editing before any teacher review.
• Over time, they internalize the checks.

Our Common Errors Detective Sheet is built for exactly this.

Frequently Asked Questions

Should I ever give comprehensive correction? Rarely. Maybe on a final draft when the student requested it. Otherwise, focused correction wins.

*What about students who want every error marked?* Talk about it. Explain that focused correction produces more growth. Most students adjust once they see results.

Doesn't this lower standards? No. Focused correction raises standards by ensuring students master one skill before adding another.

What if my school requires me to mark every error? Push back with the research. The Ferris and Bitchener literature is substantial.

How do I track which student needs feedback on which pattern? Class error log. See Formative Assessment for Grammar for the format.

Where to Go Next

Three concrete next steps:

1. Try focused correction on your next stack of essays. One pattern per student.
2. Conference with one student per day for a week. Five minutes each.
3. Automate the system. Grammar Spy Membership tracks each student's error patterns automatically, generates targeted practice, and lets you focus your feedback time where it matters most.

Grammar correction is supposed to make writers better, not break them. Focus the feedback, conference the patterns, and watch both confidence and writing grow.