Common Dictation Mistakes and What They Reveal

You finish a dictation, compare it with the transcript, and see a page full of red. The natural reaction is to count the errors and feel bad. The better reaction is to read them. Dictation mistakes are not random: each one points to a specific gap between how you think English sounds and how it actually sounds. This article shows you how to sort your errors into six types, what each type reveals about your listening, and a shadowing drill that fixes it. If you have not built a routine yet, start with our English dictation practice method, and if you are still choosing clips, the level-by-level dictation plan for beginners tells you which material fits your level.

Why Your Dictation Mistakes Are Useful Data

A score such as "32 out of 40 words" tells you how you did today. It does not tell you what to do tomorrow. The pattern behind the mistakes does. Ten missed words can come from ten different problems, or from one problem repeated ten times. In practice it is usually the second: most learners have two or three dominant error types that cause most of their misses.

So after every session, do one extra step. Next to each mistake, write a short label: WF (weak form), LK (linking), END (ending), NUM (number or name), HOM (homophone), VOC (unknown word). After two weeks, count the labels. The biggest group is your next month of practice.

Common Dictation Mistakes: The Six Types

1. Missing small words (weak forms)

What it looks like: you write "I want go shop" for I want to go to the shop, or "cup tea" for a cup of tea.

What it reveals: you are listening for the dictionary pronunciation of grammar words. In real speech, to, of, for, and, can, have, was lose their stress and shrink to a quick schwa: to sounds like "tuh", of like "uh", can like "c'n". Your ear skips them because they do not match the sound you expect. This is closely tied to rhythm, which is why the guide to English intonation and sentence stress helps here too: stressed words stay clear, everything in between gets squeezed.

The fix: take three sentences you got wrong and shadow them ten times each, deliberately making the small words as short as the speaker does. This lesson on reductions and contractions is full of good examples to practise with.

2. Wrong word boundaries (linking)

What it looks like: you write one strange word where there were two or three normal ones. "Turn it off" becomes "turnitoff" or "tur need off". "An apple" becomes "a napple".

What it reveals: you are searching for gaps between words that do not exist. English links a final consonant to the next vowel, so the boundary moves. Your brain hears the syllables correctly but cuts them in the wrong place.

The fix: when you check the transcript, draw a small arc between linked words. Then shadow the phrase slowly, pushing the consonant onto the next word, and speed up until you match the speaker. A set of daily phrases for connected speech gives you short, repeatable material for exactly this drill.

3. Dropped endings (-s, -ed, contractions)

What it looks like: "She work there" for she works there, "we walk home" for we walked home, "he gone" for he's gone.

What it reveals: this is the most common error in dictation, and it has two causes. First, the endings are acoustically tiny, especially when the next word starts with a consonant: in walked to, the /t/ of -ed blends straight into the /t/ of to. Second, if your first language does not mark grammar at the end of words, your brain does not expect the information to be there.

The fix: use grammar as a second pair of ears. Every time you write a verb, ask: who is the subject, and when did this happen? Then replay and listen specifically for the ending. Finally, shadow the sentence while slightly exaggerating the final sound, then say it again at normal speed. Your mouth teaches your ear what to expect.

4. Numbers, names and spelling

What it looks like: "fifty" for fifteen, "thirteen" for thirty, a name written three different ways.

What it reveals: numbers and names carry no context. You cannot guess them from the rest of the sentence, so they test pure sound recognition. The -teen / -ty pair is the classic trap: the difference is mostly stress (fifTEEN versus FIFty) and the length of the final vowel.

The fix: practise numbers on their own. This short lesson on pronouncing English numbers is perfect for dictation: pause after each number, write it, and check. For names, dictate letter by letter, then practise saying the English alphabet sounds, since the vowels in A, E, I are a frequent source of spelling errors.

5. Homophones and near-homophones

What it looks like: "there" for their, "weather" for whether, "right" for write, "accept" for except.

What it reveals: good news, actually. Your ear heard the sound perfectly; the mistake is in choosing the spelling. This is a reading and grammar problem, not a listening problem.

The fix: keep a short homophone list and write one example sentence for each. Do not spend shadowing time on these; spend it on types 1 to 3, where the real listening gaps are.

6. Unknown words

What it looks like: a blank, or a word written phonetically ("en-trer-pre-nur").

What it reveals: a vocabulary gap. If this type makes up more than a quarter of your errors, your clips are too hard. Go back one level.

The fix: add the word to your vocabulary list with the sentence it came from, and shadow that sentence. Learning words from sound first makes them much easier to catch next time.

What Your Error Profile Says About You

Once you have counted your labels, you will usually fit one of these profiles:

• Mostly weak forms and linking: your vocabulary is fine, but you expect "textbook English". Spend most of your time shadowing natural, fast conversation. This is typical at B1–B2.
• Mostly endings: your ear needs grammar support. Slow the audio slightly, predict the grammar before you replay, and do more careful shadowing of final sounds.
• Mostly unknown words: your clips are too difficult. Move down a level and build vocabulary through easier material.
• Mostly homophones and spelling: congratulations, your ear is doing its job. Move to harder material, such as the C1 practice videos.

A 15-Minute Error-Fixing Session

Use this after any dictation where you made more than five mistakes:

1. Label (3 minutes). Mark each mistake with its type.
2. Pick the top type (1 minute). Choose only the biggest group. Ignore the rest for today.
3. Replay with the transcript (3 minutes). Listen to each sentence with that error until you clearly hear what you missed.
4. Shadow (6 minutes). Repeat each of those sentences with the speaker five to ten times, copying the exact rhythm and the reduced sounds.
5. Re-dictate (2 minutes). Close the transcript and write the same sentences again. They should now be easy.

The shadowing step is what makes the fix last. When your own mouth produces "wanna", "c'n" and "walk-tuh", your ear recognises them immediately in other clips. Our overview of pronunciation practice exercises has more drills for the individual sounds behind these errors, and the page on accurate English pronunciation explains how clear production and clear listening support each other.

Track Your Progress Month by Month

Every four weeks, redo one clip you dictated at the start of the month and compare the labels. You are not only looking for fewer mistakes. You are looking for a change in type. Moving from "mostly weak forms" to "mostly spelling" is a big step forward, even if the total number of errors looks similar. It means your ear has started hearing real English, and the remaining work is on the page, not in your listening.

Find and Fix Your Dictation Mistakes on ShadowingEnglish

On ShadowingEnglish, Dictation mode checks your answer word by word, so every missed ending, dropped weak form and wrong boundary is highlighted immediately. Then open Shadowing mode on the same clip and repeat the sentences you missed: AI scores your pronunciation down to individual sounds, so you can see whether you are really producing the reduced forms you failed to hear. Label your errors, fix one type at a time, and your dictation page will turn from red to clean faster than you expect.