Key takeaways
- Dictation is listen-and-write. It is the only common drill that forces listening, segmentation and orthography to succeed at the same time, which is why it exposes gaps the other drills leave hidden.
- A dictation error is diagnostic: it tells you whether you failed to hear the sound, failed to find the word boundary, or heard the word correctly and could not spell it. Those are three different problems with three different fixes.
- Dictogloss is the structured classroom version, described as four stages — preparation, dictation, reconstruction, and analysis with correction — and its rationale is noticing the gap between what you produced and the target.
- The evidence is real but not overwhelming. One controlled comparison found an input-based instruction method outperformed dictogloss on both comprehension and production of a specific grammar target, so dictation is a good tool rather than a superior one.
- Handwriting shows an encoding advantage over typing in brain-activity measures, though in the same study behavioural test scores did not differ significantly — a reason to favour the pen for a new script without overstating the case.
Dictation is listen-and-write. You hear a short stretch of the language and you produce it in writing. That is the whole drill, and it is unglamorous enough that most self-study plans skip it entirely — which is a mistake, because it is the only common exercise that requires listening, word segmentation and orthography to all succeed at once.
Every other drill lets one of those three hide. You can shadow audio fluently without knowing how a single word is written. You can review flashcards for months without ever having to find a word boundary in continuous speech. Dictation removes both escape routes in the same sentence.
What a dictation error actually tells you
The reason to run dictation is not the transcription. It is the error, because a dictation error is unusually diagnostic. When you get a word wrong, it failed at one of three distinct points, and each points to different practice.
You did not hear it. The sound did not register, or registered as a different sound. This is a perception problem, and more writing will not fix it — it needs focused listening at the level of the individual contrast.
You heard it but could not find the boundary. You caught a stream of sound and cut it in the wrong place. This is segmentation, and it is the single most under-practised listening skill. It improves with volume of connected speech, not with vocabulary.
You heard it, segmented it, and could not write it. This is an orthography problem, and it is the one that only dictation catches. In a language with an unfamiliar script it is extremely common to recognise a spoken word instantly and still have no idea which characters it takes.
Most learners, on finding a page of dictation errors, conclude "my listening is bad" and do more listening. Sorting the errors into those three buckets first is a better use of five minutes than another hour of audio.
Dictogloss: the structured version
The classroom form of this drill is dictogloss, and it is worth knowing because it turns a test into a learning cycle. As described in the applied-linguistics literature, it runs in four stages: preparation, dictation, reconstruction, and analysis with correction.
The mechanism is not transcription accuracy. It is the reconstruction step, where learners rebuild a text from partial notes and in doing so are pushed to attend to the forms they do not yet control. The rationale usually cited is noticing — the claim that learners acquire a form partly by registering the gap between what they produced and what the target actually was. Dictogloss makes that gap impossible to ignore, because it is sitting on the page in your own handwriting.
For solo study you can compress this. Prepare by skimming the topic vocabulary, take the dictation, reconstruct from what you caught, then compare against the source and mark the differences. The comparison step is the lesson; skipping it reduces dictogloss to a quiz.
What the evidence does and does not support
It would be easy to oversell this, so here is the honest picture.
Language-testing research treats dictation and partial dictation as constructed-response formats — formats where you have to produce the answer rather than pick it. A 2020 review in Frontiers in Psychology notes that partial dictation can effectively evaluate test-takers' listening and their ability to reconstruct what they have heard, and also reports that test takers generally perform better on multiple-choice items than on constructed-response formats. That is a useful pair of findings: dictation measures something real, and it measures it at a higher difficulty than recognition tasks do.
On the teaching side, the picture is genuinely mixed. A 2019 study in the Eurasian Journal of Applied Linguistics compared three methods for teaching the English passive voice to fifty-one pre-intermediate learners — processing instruction, dictogloss, and meaning-based output instruction — and found the processing-instruction group outperformed both of the others on comprehension and production tasks.
That result does not mean dictation is worthless. It means dictation is one tool among several, and that for a specific grammatical target an input-processing approach did better in that trial. The reasonable conclusion is to use dictation for what it uniquely tests — the listening-to-writing link — rather than as a general-purpose teaching method.
Pen or keyboard, when the script is new
If you are learning a script you cannot yet write from memory, there is a reason to prefer handwriting.
A 2021 study in Frontiers in Human Neuroscience compared ink pen on paper, digital pen on tablet, and keyboard typing while learning new words, using an N400 event-related potential measure. The repetition priming effect was greater in the two pen conditions than in the keyboard condition, and for participants already comfortable with a stylus, the tablet performed comparably to paper.
The caveat matters and is part of the same paper: behavioural measures — the actual test scores — showed no significant differences between the methods. So the fair statement is that handwriting showed a neural encoding advantage that had not yet surfaced as a measurable performance gap in that study.
For a learner, that is still enough to tip the decision. Writing an unfamiliar character by hand forces you to commit to stroke order, proportion and component identity in a way that tapping a keyboard does not, and those are exactly the things a new script demands. Use a keyboard once the script is automatic, if you prefer the speed.
Running dictation on a phone
Learn Lao Language Fast & Easy — our own app, free with ads and no in-app purchases — lists dictation among its practice formats alongside listening, matching, vocabulary review, sentence building, quizzes and pronunciation activities, with mistake review built in. Learn Cantonese for Beginners, also ours, carries a Dictation Mode within a review set that includes Daily Review, Mistake Rescue, Listening Blitz, Sentence Builder and Match Blitz.
Two features of that arrangement matter more than the drill itself. The first is that the dictation material is drawn from what you have already covered, so you are being tested on your own weak points rather than on arbitrary sentences. The second is the mistake list: dictation without a record of what you got wrong is a quiz, and the record is where the learning is.
A ten-minute routine
Keep the session short and the analysis long.
Pick three or four sentences you have met before. Play the first, write what you hear, play it once more, and stop. Do not replay it five times — the fifth listen is memory, not comprehension, and it contaminates the diagnosis.
Do the same for the rest, then compare everything against the source at once. Mark each error into one of the three buckets: not heard, not segmented, not spelled.
Then act on the pattern rather than the individual mistakes. A page full of spelling errors means script practice. A page full of boundary errors means more connected listening at natural speed. A page full of sound confusions means targeted pronunciation contrast work.
Finally, put the missed items into whatever review system you use, so they come back. Dictation pairs unusually well with spaced repetition because it generates a clean, honest list of things you cannot yet do, which is exactly the input a scheduling algorithm needs.
Three sessions a week is enough for the diagnostic value to show. The wider language learning library covers the other habits that go around it — but if you are learning an unfamiliar script and have never run a dictation, that is the gap most worth closing first.
Good to know
Frequently asked questions
What is the difference between dictation and shadowing?
Does dictation actually work?
Should I handwrite or type my dictation?
How long should a dictation session be?
Sources
- Assessing Second Language Listening Over the Past Twenty Years: A Review Within the Socio-Cognitive Framework (opens in a new tab)Frontiers in Psychology (He & Jiang, 2020) — accessed
- Processing Instruction Revisited: Does it Lead to Superior Performance in Interpretation and Production? (opens in a new tab)Eurasian Journal of Applied Linguistics (Farahian & Avarzamani, 2019) — accessed
- Advantage of Handwriting Over Typing on Learning Words: Evidence From an N400 Event-Related Potential Index (opens in a new tab)Frontiers in Human Neuroscience (Ihara et al., 2021) — accessed
About the author
Reign Creative LLC — The independent app studio that builds the apps covered on this site and publishes every article here. On Google Play it publishes as Reign Collective Apps.
We use AI tools to help research, draft and edit our articles. Each article lists the sources its factual claims rely on, and details about our apps come from their live Google Play listings. No article on this site has been reviewed by an independent expert unless the reviewer is named on it.
Spotted an error? Tell us — how corrections are handled is set out in our editorial policy.
