Speaking Practice

Never answer Part 1 with a bare "yes," "no," or one word. Give a direct answer, then add ONE reason, example, or extra detail.

A string of one-word answers gives the examiner almost nothing to assess — Fluency & Coherence specifically rewards natural extension, regardless of how accurate your grammar is.

Part 1 questions ask a specific ANGLE on a topic — what you LIKE, HOW OFTEN, WHY, WHO WITH. Catch that exact word before you answer, or you'll answer a different question than the one actually asked.

A fluent, grammatically perfect answer to the wrong angle still sounds off-topic to the examiner — they're listening for whether you genuinely answered what was asked, not just whether you talked about the right subject.

"Do you usually...," "What do you do..." asks about a HABIT — present simple. "Did you enjoy...," "Was there a time..." asks about ONE specific past event — past simple. Spot which type before you open your mouth.

Grammatical Range & Accuracy is a full quarter of your Speaking score — mixing up a habitual question with a one-off past event is one of the most common, most avoidable slips in Part 1.

Examiners are trained to notice over-rehearsed, essay-like answers. Part 1 should sound like a normal conversation — contractions, everyday vocabulary, a relaxed tone — not a memorized paragraph.

A memorized-sounding answer can actually be marked DOWN — natural, spontaneous production is exactly what Fluency & Coherence assesses, and sounding rehearsed suggests the answer isn't genuinely spontaneous.

A short natural filler ("That's a good question," "Hmm, let me think...") while you organise your thoughts sounds far more natural than dead silence — but repeating the SAME filler on every question sounds robotic, so vary it.

Several seconds of total silence before every answer damages Fluency & Coherence far more than a brief, natural filler — examiners expect a little real-time thinking, not instant scripted replies.

You cannot write full sentences in 60 seconds. Jot ONE keyword or short phrase per bullet point only — enough to prompt your memory, not a script to read aloud.

Trying to write full sentences means you run out of prep time before covering all 4 bullets, and reading written sentences aloud sounds noticeably less natural than speaking freely around keyword prompts.

The card's bullets ARE your structure — address each one in the order given. Never skip the last bullet, which is almost always "explain how you felt" and the most commonly forgotten.

Coherence & Cohesion is scored partly on logical organisation — jumping between bullets or leaving one out entirely makes a talk feel incomplete, exactly what following the card's own order prevents.

For any "describe a time..." card, structure chronologically: the situation BEFORE, what happened DURING, how you felt or what changed AFTER. This is a ready-made map for almost any cue card of this type.

Chronological order is the easiest structure for an examiner to follow in real time, and it naturally produces a genuine range of past tenses (past simple, past continuous, past perfect) — directly boosting Grammatical Range & Accuracy.

Don't just state facts — add how something looked, sounded, or felt, and your genuine emotional reaction. The final "how you felt" bullet deserves real detail, not one flat sentence.

Lexical Resource is scored on your range of vocabulary — emotional and descriptive words ("relieved," "proud," "frustrated," "over the moon") are exactly the natural range examiners listen for, and a bare factual retelling never gives you access to this vocabulary at all.

If you run out of content before time's up, extend naturally with a linking phrase ("what's more," "looking back now," "the interesting thing is") rather than stopping abruptly.

Stopping 20–30 seconds early is one of the most common Part 2 mistakes, and it directly caps Fluency & Coherence — the task explicitly asks for 1–2 minutes of continuous, sustained speech.

Part 3 always wants your view AND your reasoning — never a bare fact or a one-line opinion with no support behind it.

An unsupported opinion gives the examiner nothing to assess beyond a single sentence — the justification is where you actually demonstrate range of language and reasoning, which is what Part 3 is scoring.

Part 3 questions ask about people or society in general ("Do people...," "How do families...") — not about you personally. Shift your language from "I/my" to "people," "many families," "tend to."

Answering a general question with only a personal anecdote misses what's actually being asked — Part 3 is specifically testing your ability to generalise and discuss ideas beyond your own direct experience.

Many abstract questions have no single right answer. Use hedging phrases — "it really depends on," "to some extent," "it could be argued that" — instead of forcing an absolute, one-sided claim.

Hedging language is itself a marker of higher-level Lexical Resource and shows mature reasoning — an overly absolute answer to a genuinely nuanced question can sound naive rather than confident.

Many Part 3 questions implicitly ask you to weigh two sides ("Which is more important...," "Are X better than Y..."). Briefly acknowledge both before landing on your own opinion — don't jump straight to one side.

Acknowledging the other side first demonstrates a wider range of comparative and concessive language ("while," "on the other hand," "that said") — landing on an opinion too quickly loses this opportunity entirely.

Ground an abstract opinion with a concrete example, OR extend it by predicting how something might change in the future. Either move adds real substance beyond a general claim.

A pure abstract opinion with no grounding can sound thin — a specific example or a genuine future prediction shows you can develop an idea fully, which is exactly what Part 3's longer, more analytical answers are meant to demonstrate.