How to Measure Your English Progress Yourself
Somewhere around week four, most adult students on an intensive course have the same private thought: *am I actually getting better, or am I just getting comfortable?*
It is a reasonable worry, and the school's level test is a poor cure for it. Placement assessments are designed to sort students into classes, not to detect the improvement that happens between them. They run infrequently, they compress a complicated ability into one number, and that number is stubbornly stable during exactly the phase when the most useful changes are happening — which is why the week 3–4 plateau feels so much worse than it is.
The fix is to measure things the test does not. All five measurements below can be run with a phone and a notebook, take under fifteen minutes a week, and — crucially — move on a timescale short enough to keep you honest.
Do all of them once in your first three days. A baseline you did not take is a comparison you cannot make.
1. The recording comparison — the one that convinces you
The method. On day one, record yourself speaking for two minutes on a fixed prompt: *describe your job and what you find difficult about it*. Do not script it, do not re-record, do not delete the bad one. Save it. Repeat with the same prompt every two weeks.
What to listen for when you compare, in this order:
- Length of your longest unbroken stretch. Early recordings are studded with stops. Later ones run in longer arcs.
- Filler and hesitation sounds. Count them, roughly. The count usually falls before the grammar improves.
- Self-correction. In week one, students often plough past their own mistakes. Later they catch and repair them mid-sentence — this is progress, not regression, even though it can sound messier.
- Rhythm. Whether the sentence rises and falls like speech, or arrives in equal-weighted blocks like a translation.
Why it works. Improvement in speaking is invisible from the inside because your own baseline resets daily — you only ever hear today's voice. The day-one recording is the only unbiased witness you will ever have, and playing it back at week six is the single most motivating thing most students do on a course.
2. Response latency — the gap you actually want to shrink
The method. In conversation, notice the pause between the end of someone's question and the start of your answer. You cannot time it precisely without being weird about it, so use a three-band scale after each class: *instant*, *a beat*, *a visible silence*. Log which band was typical today.
What to watch for. The band improves in an uneven order — it drops first on familiar topics (your work, your country, your weekend), and much later on unfamiliar ones. That is normal and worth knowing, because it means you should test the band on new topics before concluding you have plateaued.
Why it works. Most adults who describe themselves as "not fluent" have adequate grammar and vocabulary. What they lack is retrieval speed. Latency is the closest thing to a direct measurement of it, and it is the variable that changes how conversations feel — to you, and to the person waiting for your answer.
3. Vocabulary recycling rate — the harshest and most useful number
The method. Once a week, in a single conversation or class, keep a small tally of new words or phrases you have learned *and then used yourself unprompted*. Not words you recognized. Not words you wrote in a notebook. Words that came out of your mouth in a real sentence, without a teacher prompting them.
What good looks like. The absolute number matters less than the trend. Most students discover something uncomfortable in week two: they are collecting vocabulary at ten times the rate they are deploying it. That gap is the real bottleneck, and it is fixable — the correction is simply to force usage within twenty-four hours of learning a word, ideally the same day.
Why it works. Recognition vocabulary grows almost automatically through exposure. Production vocabulary only grows when you make it. This is the measurement that tells you which one is happening.
4. The listening ceiling — measured against real speech
The method. Choose one podcast or interview clearly above your comfort level — native-speed, unscripted, not made for learners. Listen to the same three minutes at baseline and every two weeks. Rate on a five-point scale: *lost / gist only / most of it / all but details / all of it*, and write down one thing you missed.
What to watch for. Comprehension does not climb steadily; it lurches. And it improves in a specific order — first you stop losing the thread entirely, then you catch specific words that used to be mud, then you follow speakers who overlap and interrupt each other. Content spoken by non-native speakers in a group is its own ceiling, which is precisely what group classes train.
Why it works. Listening is the ability that improves most during intensive study and gets the least credit, because there is no obvious moment where it becomes visible to you.
5. The unplanned-topic test — the honest one
The method. Once a week, have someone give you a topic you have never discussed in English and did not prepare — *the tax system in your country*, *whether cities should ban cars*, *how your grandparents met*. Talk for ninety seconds. Note how it went in one line.
What to watch for. Whether you can build an argument, not just retrieve sentences. Whether you can say something you have never said before. Whether you can handle a topic where your specialist vocabulary does not exist.
Why it works. Almost everything else you do on a course is rehearsed to some degree — the same self-introduction, the same descriptions of your job, the same familiar opinions. Rehearsed material improves first and flatters your sense of progress. The unplanned topic is where you find out what is actually available to you under pressure, which is exactly the condition of a real meeting or interview.
How to run the whole thing without it becoming a chore
- Fifteen minutes on Sunday. One recording every two weeks, one listening check every two weeks, a weekly vocabulary tally, a weekly unplanned topic, and a daily one-word latency note. That is the whole system.
- Never compare week two with week three. The signal is too small and the noise — sleep, mood, who you spoke to — is too large. Compare in four-week blocks and the picture is clear.
- Expect the plateau to appear in the numbers, and to be wrong. Around weeks three and four, subjective progress stalls while latency and listening usually keep improving. That mismatch is the entire reason to measure rather than to feel.
- Bring the results to your teacher. "My latency is fine on work topics and terrible on anything abstract" is a far more useful thing to hand a teacher than "I want to improve my speaking." It changes what they do with your next twenty lessons.
What none of this measures
Two honest limits. First, these measurements track fluency behaviors, not exam performance. If your goal is a specific test band, they will not predict it — that requires timed practice under exam conditions, and our IELTS preparation guide covers that path. No school, method or measurement guarantees a score.
Second, they measure the trend, not the destination. What a given number of weeks realistically produces is a separate question, covered in our 4, 8 and 12-week progress guide. The point of self-measurement is narrower and more practical: knowing whether *this week* is working, while you can still change what you are doing about it.
FAQ
How often should I record myself?
Every two weeks is the sweet spot. Weekly recordings show too little change and become discouraging; monthly ones let a bad month run too long before you notice.
Is it normal to sound worse in week three than week one?
Fairly often, yes — because you have started attempting harder structures and catching your own errors, both of which make speech messier before they make it better. Judge the trend across four-week blocks.
Should I keep tracking after I go home?
Yes, and this is when it matters most. Language decays quietly, and a monthly two-minute recording is the cheapest early warning system there is. Our guide to keeping your English after returning home covers the rest.
Will my school do this for me?
Schools run their own placement and progress assessments, and formats vary between institutions — ask what yours does and how often. These five measurements are meant to sit alongside that, not replace it, because they move on a weekly timescale that formal testing cannot.
Every CAESA member school is government-licensed, and enrollment runs through a partner agency. See the 8 CAESA schools to compare programs, or get in touch to talk about what your goal actually requires.