Spell-Check and Autocorrect: Are They Making HK Students Worse at Spelling?

Last term, I set my P5 class a written task on paper — no devices, no autocorrect. The results surprised even me. Children whose digital writing seemed perfectly spelled were suddenly producing errors I had not seen since P2. Freind for friend. Becuase for because. Recieve for receive.

Their fingers had been typing correctly because autocorrect was silently fixing errors without their knowledge. Their brains had not necessarily learned the spellings at all.

This is the autocorrect problem, and it is more significant in Hong Kong than in many English-speaking places, for a specific reason: for Cantonese-first learners, there are fewer natural pathways to spelling reinforcement. A native English speaker encounters correct spellings in ambient reading constantly — signs, packaging, screens, books. Cantonese-first children in HK encounter correct English spellings largely through their English schoolwork. If that schoolwork is mediated through autocorrect, the learning loop is broken.

What Spell-Check and Autocorrect Actually Do

It is worth being precise about what these tools are doing.

Spell-check (the red squiggly underline) flags potential spelling errors and offers suggestions. It requires the user to notice the flag, consider the suggestion, and choose to apply it. This involves some engagement with the spelling, even if it is minimal.

Autocorrect is more automatic: it changes what you type without requiring any decision. You type freind, it silently becomes friend. You do not see the correction; you do not engage with the spelling. It is as if the error never happened.

Predictive text goes further: it may complete words before you have even attempted to spell them, removing the spelling challenge entirely.

The spelling learning that happens through correction requires the learner to notice they were wrong, consider the correct form, and apply it. Autocorrect eliminates this loop entirely.

The Evidence on Spell-Check and Spelling Development

Research on this is mixed and nuanced. The short version:

• For children who already have strong spelling foundations, spell-check is a useful safety net that does not significantly harm further development.
• For children still developing spelling skills, heavy reliance on autocorrect appears to impede the consolidation of correct spelling patterns.
• The most concerning pattern is over-dependence: children who cannot reliably spell common words without technological assistance because they have never needed to.

The problem is not the existence of spell-check. It is the absence of unassisted spelling practice.

What I Actually Observe in My Classroom

I see three distinct student profiles in terms of technology and spelling:

Profile 1: The Unaffected Strong Speller. This child reads widely, has always been interested in words, and has a natural orthographic memory. They might autocorrect freely without harm because their underlying spelling knowledge is solid.

Profile 2: The Spell-Check-Dependent Competent Speller. This child can spell correctly when prompted to check, but their automatic spelling (in handwritten work, or typed quickly without review) is shaky. Autocorrect has created surface competence without underlying mastery.

Profile 3: The Spell-Check-Masked Poor Speller. This child's digital work looks fine. Their handwritten work reveals consistent errors in even high-frequency words. Without assistance, they are significantly below their grade level in spelling. Neither they nor their parents have realised this because everything typed looks correct.

Profile 3 is the one that concerns me most, and in my experience it is not uncommon in P4–P6.

A Healthy Relationship with Spell-Check

My view is not that spell-check is bad and should be removed. Spell-check is a genuine productivity tool that skilled adult writers use appropriately. The goal is developing a relationship with it that does not undermine underlying spelling competence.

Do: use spell-check as a final check, not a substitute for trying to spell correctly first. The habit should be: attempt the spelling, write the draft, then review with spell-check at the end. This preserves the active spelling attempt.

Do: when spell-check flags a word, actually look at the correct form rather than mindlessly accepting the suggestion. Take two seconds to notice what was wrong and what is right.

Don't: allow autocorrect to silently fix everything without the child ever seeing the correction. On school assignments, consider whether autocorrect should be disabled.

Do: maintain regular handwritten writing practice. A piece of writing done by hand each week is not a punishment — it is essential diagnostic practice that reveals what the child can actually spell unaided.

What Parents Can Do

Weekly handwritten writing: Even ten minutes of handwritten English per week — a diary entry, a letter to a relative, a short story — maintains contact with unaided spelling. Review it together and correct gently.

Post-dictation analysis: Your child's school English dictations are already handwritten tests of unaided spelling. Take the marked dictation papers and look at them for patterns. Repeated errors in the same words or patterns tell you exactly where to focus practice.

The "hide autocorrect" experiment: Occasionally ask your child to type something with autocorrect disabled (this can be done in device settings). See what the underlying spelling is like. You may be surprised.

Praise the attempt, correct the error: When a child writes becuase and then says "Oh, autocorrect usually fixes that" — this is information. It tells you there is a spelling gap to address. The response is not alarm, but targeted practice of that specific word.

Spell-check is a useful tool in the right hands. The right hands are ones that already have the underlying skills. Building those skills remains the work of education — not just in school, but in the habits we build at home.