LinguaGrade Blog

Exam prep, AI grading and language-center operations.

Admins manage articles in Content Studio. Only published articles appear publicly.

Trung tâm

Measuring learner progress with exam data instead of impressions

A 0.5 band rise over three months tells you nothing until you know where it came from. These four indicators are already sitting in your exam data.

View detail →

Trung tâm

Student data privacy when your centre uses an AI marking platform

Voice recordings and written work are personal data. Before signing, there are six questions worth asking any vendor — and one that is hardest to undo if answered badly.

View detail →

Trung tâm

Building an exam bank for a language centre: components first, papers second

Centres run out of papers faster than they expect. The durable approach is to manage components — passages, tasks, listening sections — and assemble papers from them.

View detail →

Trung tâm

Online exam integrity: what browser controls can and cannot do

No online exam room prevents cheating outright. What it can do is set visible rules, count violations, log them, and tell candidates the rules first.

View detail →

Trung tâm

How far you can trust AI marking for Writing and Speaking

The useful question is not whether AI marks accurately. It is which responses may be released automatically and which must stop for a human.

View detail →

Trung tâm

Running IELTS mock exams: an operations checklist for language centres

A mock exam rarely fails because of the papers. It fails because of the audio, the timing, or because results arrive two weeks late. This checklist covers all three.

View detail →

Chứng chỉ

IELTS and Vietnam's six-level foreign language framework

The framework itself is well defined and aligns with CEFR. The IELTS-to-level mapping is not universal — it depends on the regulation of the receiving institution.

View detail →

Luyện thi

How many IELTS practice tests are enough, and why repeats mislead you

Fifty papers without error analysis lose to fifteen papers with an error log. And re-sitting a paper you have already done gives you a number that measures memory.

View detail →

Luyện thi

Self-assessing IELTS Writing: how far it takes you

Self-marking catches everything countable: length, repetition, thin structure. What it misses is the well-written answer to the wrong question — the costliest error of all.

View detail →

IELTS

Computer-delivered vs paper IELTS: what actually differs

The content and the scale are identical. What changes is the mechanics: transferring Listening answers, counting words, and marking questions to revisit.

View detail →

IELTS

IELTS Listening structure: four parts, 40 questions, and 10 easily lost minutes

Forty questions, four parts, rising difficulty. The paper test grants ten minutes to transfer answers; the computer test does not — and that gap catches people out.

View detail →

IELTS

The four IELTS Speaking criteria, and the one everybody misreads

Four criteria, equal weight, averaged into your Speaking band. Pronunciation scores intelligibility, not accent — and that distinction changes how you practise.

View detail →

IELTS

The four IELTS Writing criteria, and why Task 2 weighs more

The four criteria are equally weighted within a task. Across the paper, Task 2 counts for more — so a weak Task 2 costs roughly double.

View detail →

IELTS

Academic vs General Training: same scale, different thresholds

Both versions share the nine-band scale, but Reading uses two different conversion tables: General Training demands more correct answers for the same band.

View detail →

IELTS

How IELTS scores are calculated: from 40 correct answers to your overall band

Thirty correct answers out of 40 give band 7 in Listening and in Academic Reading — but General Training Reading needs 35 for the same band. Here is the full path from raw score to overall.

View detail →