How to Give Spelling Feedback That Teaches, Not Just Marks It Wrong

Every teacher knows the moment. You hand back a story with becuase circled in red, and a week later it comes back in the next story, in the same handwriting, with the same confidence. The circle was accurate. It was also useless, and not because the student ignored it. A circle tells a child that something is wrong without saying what, or why, or what to do about it, so the child does the only reasonable thing with that information, which is nothing.

Better spelling feedback does not mean more marking. In most classrooms it means less: fewer words marked, each one named precisely, and a short request that the student produce the word again from memory. That last step is the one most often skipped. It is also the one that does the teaching.

What follows is a practical account of how to read an error, what to write or say about it, how many errors to take on in a single piece, when to leave spelling alone entirely, and what the evidence says about timing. It is written with classroom teachers in mind, but tutors and homeschooling parents will find it works the same way at a kitchen table.

What the research on feedback actually says

The most cited review in this area is John Hattie and Helen Timperley's "The Power of Feedback," published in 2007. Their headline is usually quoted as "feedback works," but the more useful part is the fine print: feedback has one of the widest ranges of effect of anything a teacher does. Some of it produces large gains. Some produces nothing at all.

An earlier analysis by Avraham Kluger and Angelo DeNisi, in 1996, was blunter. Across the studies they pooled, roughly a third of feedback interventions made performance worse. The failures had something in common: the feedback pulled the learner's attention away from the task and toward themselves. "Careless" is feedback about the self. So, for that matter, is "Great job."

Hattie and Timperley proposed that useful feedback answers three questions for the learner: where am I going, how am I doing, and what should I do next. A red circle answers half of the middle one. It confirms that the student is doing badly, which most of them already suspected, and says nothing about the other two.

There is a second reason the circle fails, and it comes from what we know about how spellings are stored. A spelling is not a picture held in memory. It is a set of connections between the sounds of a word, the letters that represent them, and the word's meaning, a process the reading researcher Linnea Ehri calls orthographic mapping. The student who writes frend has built a perfectly sensible set of connections, since every sound is there. To repair it, they have to notice the specific letter that is missing, attach it to the right place, and then retrieve the corrected version on their own. Only the student can do that work. Feedback succeeds when it sets the work up and fails when it merely records that the work has not been done.

Read the error before you mark it

A misspelling is information about what a student knows. Before deciding what to say, it is worth ten seconds to work out what kind of error you are looking at. Each kind calls for a different response, and a generic correction throws away the one useful thing the error can tell you.

Most spelling errors fall into four broad families.

A sound error misses or misrepresents a sound: stret for street, pin for pen. The student has not yet heard the word accurately, and no amount of looking at the correct spelling will change what the ear is doing. The response is to say the word slowly together and count its sounds.

A pattern error captures every sound but uses a spelling English does not use in that position, as in rane for rain or boks for box. These are intelligent errors. The student found a legitimate spelling of the sound and put it in the wrong place, and the response is to name the pattern that governs that position.

A meaning error spells a word by sound and misses a part that carries meaning. The classic is jumpt for jumped: the past tense ending is spelled the same way whether it sounds like a t, a d or "id." The response is to point at the word part rather than the letters.

A word-specific error involves a part of the word that no rule the student knows can explain, such as the ai in said or the o in women. Here the honest response is that this part simply has to be learned, and that it is the word being unusual, not the student.

The diagnostic question is short enough to ask of every error. Did they mishear it, choose the wrong pattern, miss a word part, or hit a word that has to be learned by heart? For a fuller treatment with dozens of real examples, see what a spelling mistake tells you.

What useful feedback sounds like

Take one error, frend, and three ways of responding to it.

The first is a circle. The second is the correct word written neatly above the error, which is kinder and slightly more useful, since at least the target is now visible. The third takes about fifteen seconds: "You've got every sound in friend. There's a silent i hiding before the e. Cover it up and write it again here." That version tells the student what they did right, narrows the problem to a single letter, and ends with something to do.

For a pattern error, name the pattern rather than the word, because a pattern covers dozens of words and a word covers one. "That's a real way to spell the long a sound, just not in the middle of a short word. There we usually write ai, as in rain, paint and train. Try it again."

For a meaning error, point at the part. "This is jump plus the past tense ending, and that ending is always spelled ed, even when it sounds like a t. Can you find another one in your story?" Very often the student will find walkt or lookt without help. A correction the student finds is worth several you hand them.

Notice what none of these contain: any comment about the student. Not "careless," not "you know this," not "sloppy." This is partly kindness, but mostly it is the evidence. Feedback aimed at the self, praise included, tends to do little for learning. Feedback aimed at the task, and at how to do it better, tends to help.

The retry is where the learning happens

If you change one thing about how you mark spelling, make every correction end with the student writing the word again from memory.

The reason is the testing effect, one of the sturdiest findings in memory research: pulling something out of memory strengthens it far more than looking at it again. In Henry Roediger and Jeffrey Karpicke's well-known 2006 study, students who practiced recalling a passage remembered much more of it a week later than students who spent the same time rereading it. Looking at a corrected spelling is rereading. Covering it and producing it is retrieval.

Feedback and retrieval work as a pair. Andrew Butler and Henry Roediger showed in 2008 that testing without feedback can leave errors in place, while testing followed by feedback corrects them and keeps the benefit of the test. So the loop that builds a durable spelling is simple to describe. Attempt, see the correct form, attempt again.

School psychologists have a named routine for exactly this, called cover, copy, compare. The student looks at the correct word, covers it, writes it from memory, then uncovers it and checks, going round again after a miss. It has been studied for decades, mostly in small classroom trials, with consistently positive results across ages and subjects. It takes about as long as reading this paragraph.

A word about the older alternative. Writing a word out ten times while looking at it feels productive to everyone involved, but the eyes do the work and memory is barely asked to take part. Three genuine attempts from memory are worth more than ten copies.

How many errors to mark

Not all of them. In most pieces of writing, not even most of them.

Think about what the student can actually use. A page with fourteen red marks sends one clear message, and it is not about spelling. Few students will study fourteen words; most will study none, and some will start writing shorter, safer sentences to limit the damage next time. Feedback a student cannot act on is noise, and noise at that volume is discouraging.

A workable rule is two or three words per piece, chosen with some care:

• words using a pattern you have already taught, so the student can repair them with knowledge they have
• very common words the student will write hundreds of times, such as because, said, they and friend
• a word misspelled more than once in the same piece
• errors that change the meaning, such as their for there or quite for quiet

Let the rest go. The student who writes magnifisent in a story about a dragon is taking a risk with vocabulary, and a red mark teaches them to write big next time. If you want to acknowledge the word, a small note at the bottom with the correct spelling and the words "great choice," and no request to fix it, does the job.

Nobody has shown that three is the right number. It is a judgment, and a teacher who knows the class should adjust it. The principle underneath is well supported, though: people learn from feedback they can act on, and they stop reading feedback that swamps them.

When to correct and when to let it go

The purpose of the writing should decide the feedback. In a spelling lesson, spelling is the point, and errors on taught words should be corrected promptly and retried. In a journal, a first draft or a piece of creative writing, the ideas are the point. Heavy spelling correction there tells students that ideas matter less than letters, and many will quietly shrink their writing to the words they are sure of, which costs them far more than a few misspellings ever would.

Age changes things too. A five-year-old who writes kat has heard every sound in the word and chosen a letter that can represent each one, which is exactly the skill early reading depends on. Research on invented spelling, notably by Gene Ouellette and Monique Sénéchal, suggests that encouraging these attempts helps young children learn to read. At that age, praise the sounds, and show the conventional spelling now and then without asking for a repair.

Older students can be held to what they have been taught. A ten-year-old who has studied ai and ay and still writes rane should fix it. The same child writing rithum has met a genuinely hard word and needs teaching, not correction.

With students learning English, watch for errors that come from the sound system of their first language. A Spanish speaker who writes espeak is adding the vowel Spanish places before an s followed by a consonant. Name that pattern once and it will repair dozens of words.

Does timing matter?

Teachers are often told that feedback must be immediate. The research is more interesting than that. Several laboratory studies, including work by Janet Metcalfe and colleagues on vocabulary learning in children and adults, have found that feedback delayed by a day or so can produce retention as good as instant feedback, or slightly better. Other studies, especially with young children learning new skills, favor speed. The fair reading is that timing matters less than what the student does with the feedback once it arrives.

For spelling, two things seem to matter. The student should meet the correct form before the wrong one is rehearsed many more times, because a misspelling written thirty times in a month becomes genuinely harder to shift. And the student should retrieve the correct form soon after seeing it, and then again a few days later.

So there is no need to hover over every desk. Marking a piece overnight is fine. What undoes the feedback is handing the work back and moving on. Build five minutes into the lesson for students to read your notes and do their retries, then bring the same words back later in the week, and again the week after, in a quick oral or written check.

That return visit is where most feedback is lost, and it is where the spacing effect does its work. A word retrieved today, again in three days and again in a week is far more likely to last than a word practiced three times in one sitting. The routine in spaced retrieval in the classroom shows one way to fit this into a crowded week.

Feedback that does not discourage

Many students who spell badly have been told so for years. They expect the red before they see it, and some have concluded they are simply "not a speller" and stopped trying. Feedback that works has to get past that belief. The way past it is accuracy.

Start with what is right, because something nearly always is. Becuase contains every letter of because; two of them have swapped places. Saying so is not flattery. It describes the problem at its true size, which is small.

Keep corrections private where you can. Make them about the word rather than the person: "this word has a tricky part" points at something the student can change, while "you always get this wrong" points at something they are. And give students a way to see progress. A short personal list of words they used to miss, with a tick each time one appears correctly in real writing, turns feedback into a record of improvement. Students who watch that list shrink begin to believe spelling can be learned. That belief is what keeps them trying. For older students who carry real shame about spelling, the piece on spelling and self-esteem is worth your time.

The one habit to change this term

End every spelling correction with a retry from memory. Choose two or three words from a piece. Name what went wrong in a phrase, show the correct spelling, have the student cover it and write it again, and bring the word back a few days later. It costs a few minutes per student. It turns marking from a verdict into a lesson.

If you want those return visits to happen without tracking them by hand, a personal word list in a notebook works well, and so does a practice tool that schedules the reviews. Spelling.School is one option: a student hears the word, types it, sees exactly which letters went wrong, and meets the word again for review on a spaced-repetition schedule. Every account starts with 14 days of everything, then a free plan with every review and three new words a day, and you can begin from the home page.

Questions people ask

How should teachers correct spelling mistakes?

The most effective correction names the kind of error, shows the correct spelling, and then asks the student to write the word again from memory without looking. A circle alone tells a student that something is wrong but not what or why. Pointing to the specific letters, and to the pattern behind them when there is one, gives the student something to learn. The retry from memory is the step that actually builds the correct spelling.

Should you mark every spelling error in a piece of writing?

Usually not. A page covered in corrections overwhelms most students, and few will study more than a handful of words at once. Choosing two or three errors per piece, such as taught patterns, very common words or errors that change the meaning, gives the student something they can act on. The remaining errors can be picked up another day.

What kind of feedback helps spelling most?

Feedback that is specific, aimed at the word rather than the student, and followed by an action. A teacher might note that every sound is present, point to the one missing letter, and ask the student to cover the word and write it again. Research on feedback, including Hattie and Timperley's 2007 review, finds that comments a learner can act on help far more than grades, circles or general praise.

Is immediate feedback better than delayed feedback for spelling?

Not necessarily. Some studies find that feedback given a day later works as well as instant feedback, or slightly better, for long-term memory. What matters more is that the student sees the correct form before rehearsing the wrong one many more times, then writes it from memory soon after and again a few days later. Marking overnight is fine as long as students get time in class to act on the corrections.

How do I correct spelling without discouraging a student?

Begin with what is right, since most misspellings get most of the word correct, and say so plainly. Keep corrections private where possible, describe the word rather than the student, and limit corrections to a few words at a time. A personal list of words the student has mastered lets them see their own improvement, which is what keeps a reluctant speller willing to try.

Does writing a word out ten times help fix a spelling mistake?

Copying a word repeatedly while looking at it feels productive but builds surprisingly little memory, because the eyes do the work and recall is never tested. Covering the word, writing it from memory, checking it and repeating after a miss is far more effective. Three honest attempts from memory are worth more than ten copies, and a return to the word a few days later helps even more.

Fix the words you actually get wrong

A few minutes a day, spaced repetition, built only for spelling. Try five words without signing up.