How to Diagnose Pronunciation Problems

A useful pronunciation assessment tells a teacher what to teach next, why it matters and how to check improvement. A list of sounds that differ from the teacher's accent cannot answer those questions by itself.

In this guide, “diagnose pronunciation problems” means identify language learning needs. It does not mean diagnose a medical condition. ASHA distinguishes accents from communication disorders and describes accent modification as elective. [1] (opens in a new tab) Keep your assessment within your training and explain its educational purpose to the learner.

The practical process below moves from the learner's communication goals to specific evidence and a manageable teaching plan.

Start with the situation the learner wants to handle

Ask for a concrete example of a difficult interaction. “I want better pronunciation” is a starting point; “Customers sometimes misunderstand the dates I give over the phone” is a teachable problem.

Explore who the learner speaks with, what information matters, which situations feel demanding and what change they want. Ask what already works well. A learner's successful conversations can reveal strategies worth retaining.

Agree on how any recording will be used, who can access it and whether it will be retained. Offer a live observation alternative when recording is unsuitable. Explain that unfamiliar vocabulary, task pressure or poor audio can affect performance, so you will look at more than one sample.

Write a provisional outcome such as: “State and confirm delivery dates so a listener records them correctly.” Avoid beginning with an accent score or a promise to sound like another person.

Collect several kinds of evidence

Use a small set of tasks rather than one lengthy reading passage.

For the delivery-date example, use a short customer exchange, a few written dates and a listening task in which the learner chooses between appointment options. Let evidence determine whether numbers, prominence, pace or another feature deserves attention.

Separate an observation from an explanation

Write what happened before deciding why. “Listener wrote 14 when the learner intended 40” is an observation. “The learner has poor rhythm” is a broad interpretation that still needs testing.

Check the production again in a different phrase. Ask the learner to identify the alternatives in a recording. Confirm that both words are familiar. Try a brief model and see whether the distinction becomes clearer.

Similarly, an absent plural ending may reflect pronunciation, grammar, task pressure or a combination. Ask the learner to express the same meaning in another sentence and explain the intended number. A single token cannot settle the cause.

Record patterns provisionally: “Final consonant needs further checking” or “Stress contrast clear with a model, inconsistent independently.” These notes guide follow-up without turning uncertainty into a diagnosis.

Assess the message, sounds and phrasing

The CEFR's revised phonology framework covers overall control, sound articulation and prosody, while giving intelligibility priority over resemblance to a native speaker. [2] (opens in a new tab)

For day-to-day teaching, create a brief task-specific record:

This is a classroom record, not a validated test or an automatic CEFR level assignment. Keep notes specific enough that another teacher could understand what was observed.

A practical first-session workflow

Begin with a short conversation about goals and a low-pressure sample. Then use one realistic task, such as arranging an appointment. Let the learner finish before analysing individual features unless meaning genuinely breaks down.

Choose one uncertain feature for a closer check. Compare listening, production in a phrase and a supported attempt. Avoid testing the entire sound system simply because a checklist is available.

Next, give a brief teaching intervention. For example, model a clear contrast between “fourteen” and “forty” within the learner's appointment sentence. Ask them to communicate a different date to a listener who cannot see it.

End by reviewing the evidence together. Say, “The listener understood the place and month. We will work on distinguishing these numbers in new appointment messages.” Agree one or two initial targets, a practice task and the next review point.

Turn the findings into priorities

Choose priorities using three questions: Does the feature affect this learner's communication? Does it occur often enough to matter? Can we define a useful practice task and observe change?

A noticeable accent feature with no demonstrated effect on the agreed task may be a lower priority than unclear grouping of essential information. A learner can still choose to explore an additional accent model for personal reasons; record that as a separate preference.

Illustrative case: Nila, a fictional learner, reads a prepared product description clearly but gives long, crowded explanations during a customer role play. A listener misses the two available options. In a second attempt, Nila pauses between the options and highlights the contrasting quantities; the listener records them correctly.

An appropriate initial plan is to practise grouping options in new customer explanations. This example does not establish that Nila needs a complete vowel course or that one successful retry proves lasting improvement.

Check progress without overclaiming

Keep the original sample where consent permits. At review, include a comparable task with new content. You may also repeat an earlier task, but distinguish improvement on familiar material from transfer to something new.

Use the same instructions and similar audio conditions. Where practical, ask an additional listener to record the key information without seeing the intended answer. Note listener familiarity instead of assuming every change comes from the speaker.

Collect learner feedback separately: “How manageable did that feel?” A confidence rating can inform teaching, but it is not a substitute for evidence of what a listener understood.

Finish the review with a decision: continue the target, reduce support, test it in another setting or move to the next priority.

Frequently asked questions

Can one reading passage diagnose pronunciation needs?

It can provide useful evidence, especially for comparing selected features. Combine it with relevant spontaneous speaking and focused follow-up before deciding what to teach.

Should an app score determine the teaching plan?

Use it as an additional observation if its purpose and limitations are clear. Check the actual recording and human listener evidence. Do not treat a single score as proof of a disorder or an objective measure of accent quality.

What if the concern falls outside language teaching?

Explain your professional limits and encourage the learner to consult an appropriately qualified professional. A voice and accent trainer should not label or treat a suspected clinical condition without the necessary qualifications.

Build your assessment skills

Keep a portfolio of anonymised task designs, observation notes and teaching decisions. If you want to explore further preparation for this work, review the CVAT® Voice and Accent Trainer Certification and assess its relevance to your development as a trainer.

Sources

1. ASHA Practice Portal: Accent Modification (opens in a new tab)

2. Council of Europe: Phonological competence (opens in a new tab)

Related reading

Article 07 What Is Mother Tongue Influence (MTI) in English Pronunciation?

Article 08 How to Teach English Pronunciation Effectively

Article 17 How to Correct Pronunciation Without Reducing Learner Confidence

Article 19 How to Design a Voice and Accent Training Session