What this page can and cannot tell you
A written page cannot hear you. The habits below are tendencies that researchers have reported in groups of Vietnamese learners of English, not rules; you may have some, all or none.
The studies mostly recorded learners in Ho Chi Minh City, university students in a study published in a Can Tho University journal, or Vietnamese speakers of English in Australia. Several were small, and several recorded learners reading sentences aloud rather than speaking freely, as in an IELTS answer. None was specifically about Hanoi learners, so this page says “often” and “can”, not “always”.
What the Pronunciation criterion rewards
Pronunciation is one of four equally weighted IELTS Speaking criteria. The key assessment criteria published on ielts.org list five key indicators:
- dividing speech into meaningful chunks;
- rhythm and stress timing, plus linking of sounds, giving connected speech;
- stress and intonation used to support meaning;
- individual sounds and word stress, and how much effort they ask of the listener;
- the overall effect of accent on intelligibility.
Accent appears only in the last point, and only through its effect on intelligibility. As far as we can tell, neither that document nor the published band descriptors ask for a native or particular accent.
IDP IELTS, one of the organisations that jointly own IELTS, says in a published article that accent as such gets no score, and that a strong accent can still score well when pronunciation is clear. Band Booster IELTS is a private coaching service, not part of the organisations that own IELTS.
That does not make accent irrelevant. In one study of 16 Vietnamese speakers living in Australia, listeners’ ratings of accent strength and of intelligibility were very strongly correlated, although researchers usually treat the two as related but not the same. A safe reading is that accent itself is not marked, but the features the criteria name, which affect how easy you are to follow, are.
Common habits, and a quick self-check for each
The table lists habits reported in studies of Vietnamese learners, some more firmly supported than others. The examples are approximate.
| Habit | What it can sound like | Quick self-check |
|---|---|---|
| Weak, missing or swapped final consonants | “like” close to “lie”; “next” close to “neck”, “nest” or “net”; “knife” ending in a p sound | Record “like, next, knife” and listen only to the last sound of each. |
| Simplified consonant clusters | “president” without the p at the start, closer to “resident” | Say “president, street, asked” at normal speed and listen for every consonant. |
| Missing -s and -ed endings | “students” heard as “student”; “covered” as “cover” (it can also read as a grammar slip) | Say “I work” and “I worked”, and ask a listener which you said. |
| “th” replaced by another sound | “think” near “tink” or “sink”; “this” near “dis” | Stretch the first sound: “thhhink”, “thhhis”. The air should keep flowing; if the sound stops like a t or d, that is the habit. |
| Vowel pairs blurred | “sit” and “seat”, “full” and “fool”, “live” and “leave” sounding alike | Say “sit, seat, full, fool, live, leave” in mixed order and ask someone to write down what they hear. |
| Small words and unstressed syllables kept long and clear (mixed evidence) | “to”, “the”, “for” and “of” as long and clear as the words around them | Record “I want to go to the shop for a cup of tea” and check that the small words are short and light. |
Many of these habits seem to trace back to Vietnamese itself: standard Vietnamese has no consonant clusters, plural -s or past -ed endings, and allows only a few sounds at the end of a syllable.
Cautions on the evidence
- Endings are not always dropped. In one study of advanced Vietnamese speakers, final consonants were deleted far more often than by Australian speakers, yet final stops were also released more often, with no consistent pattern.
- Clusters are a tendency, not a certainty. In one test with Vietnamese university learners, about 77% of two-consonant and 84% of three-consonant clusters at the start of words were produced as expected. Most errors in two-consonant clusters were changed sounds, not deletions.
- “th”: one team’s judgement. Two researchers rating Vietnamese learners’ recordings judged th produced as a heavily aspirated t-like sound to have little effect on comprehension. That was their own judgement, not a measured test, and it says nothing about th replaced by s.
- Rhythm evidence is mixed. Vietnamese is usually described as syllable-timed, and learners often carry some of that pattern into English, but that split is itself debated. One study suggested more even syllable lengths in learners’ English, yet two other rhythm measures did not separate them from Australian speakers.
How to self-check, and where self-checking stops
- Record a real answer. Pick a theme from the Part 2 cue card themes, prepare for one minute and record one to two minutes, as in the test. Then listen for one thing at a time, starting with the last sound of every word.
- Ask for a second opinion. IDP’s published advice suggests asking a teacher or English-speaking friend for feedback on parts you are unsure about. Ask them to listen without a script and write down what they understood.
Recording yourself is widely recommended, including in IDP’s published advice, but the sources behind this page contain no experimental proof that it raises Speaking scores. A review of 75 studies of pronunciation teaching reports mixed results overall, although many studies showed improvement.
Sources: ielts.org scoring page, Speaking band descriptors and Key Assessment Criteria; two IDP IELTS articles; Tang 2007; Slowik & Doan 2022; Nguyen & Ingram 2004 and 2016; Mixdorff & Ingram 2009; Tran & Nguyen 2022; Wang, Xu & Zhang 2023; Thomson & Derwing 2015.
Where one-to-one live feedback helps
A live listener hears your own speech, which a page cannot. In a 60-minute session at Band Booster IELTS, Mike takes you through Speaking Parts 1, 2 and 3 in the real test format and gives feedback during the session, while the answer is fresh. Pronunciation is one of the four criteria he listens for, so a habit can be pointed out as it happens and you can try the sentence again straight away. The full format is on the IELTS Speaking coaching page.
Mike holds a Cambridge CELTA (July 2017), has taught English and IELTS for more than twelve years in Vietnam, Turkey and Iran, and scored 7.5 overall and 8.5 in Speaking in his own IELTS; see the about Mike page. That is a teaching background, not a promise of any band: progress depends on your starting point and the work you do between sessions.
Price and booking
A regular session is 60 minutes and costs 500,000 ₫ in Vietnam; a 30-minute trial costs 150,000 ₫, once per student, on the first booking only. Students outside Vietnam see their own region's price, in their own currency, on the booking page before paying. There is no subscription, and you can reschedule free up to 12 hours before the slot.
To start, book a trial session from the live calendar, which shows slots in your own timezone, or contact Mike at mkcenter007@gmail.com or +84 383 390 963. A Vietnamese version of this page is lỗi phát âm IELTS Speaking của người Việt.