Words in context questions ask your child to choose the single word that best completes a sentence, and they appear in almost every GL Assessment-style verbal-reasoning paper. Like every VR type, they stop being hard the moment your child has a fixed method they trust under time pressure. This guide gives you that method, a fully worked example, the classic traps that cost marks, and a plan for practising the type until it becomes automatic.
What "words in context" actually tests
This type is one of the 21 GL verbal-reasoning question types (Bond labels them Type A to U). It is sometimes called sentence completion or cloze. On the page your child sees a sentence with a gap and five answer options labelled A to E on the separate OMR answer sheet. Only one option makes the sentence read correctly, sensibly and in the right register. GL does not tell you in advance which of the 21 types will appear in a given paper, so sentence completion has to be drilled alongside synonyms, antonyms, odd-one-out, analogies and the letter-and-number code types. Papers typically carry three to five of these questions, each worth the same single mark as every other question — which is exactly why a fast, reliable method matters.
The five-step method
The single biggest improvement comes from predicting before you peek. Teach your child to run these steps in order, every time:
- Read the whole sentence to the end, not just up to the gap. The clue often sits after the blank.
- Cover the options and predict a word that would fit. Even a rough guess primes the brain to spot the right answer.
- Match to the closest option by meaning — the correct answer is usually a synonym of your prediction.
- Check the register. Formal sentences want formal words; a jokey word will not fit a serious sentence and vice versa.
- Read it back in full with your chosen word in place. If it sounds natural and keeps the sentence's meaning, commit and move on.
A worked example
Take the sentence: "The judge delivered a ___ verdict, and every member of the jury had agreed." The options are A happy, B unanimous, C loud, D late, E fair. Predict first: the second half tells us everyone agreed, so we want a word meaning "agreed by all". That prediction points straight to unanimous (B). Notice the traps built in: "happy" is positive and tempting but says nothing about agreement; "fair" is plausible and sounds sensible, yet the sentence gives no evidence about fairness, only about agreement. The clue after the gap is what decides it. This is why reading to the end before choosing is non-negotiable.
The traps that cost marks
Sentence completion is designed with attractive wrong answers. The commonest reasons children lose the mark are:
- Choosing the fanciest word instead of the fitting one. A long, unusual word feels "clever" but is often a decoy.
- Ignoring collocations — the natural word partnerships in English. You deliver a verdict, reach a decision, draw a conclusion. The wrong verb can be grammatical yet still wrong.
- Stopping at the gap and guessing from the first half only, missing the clue that follows.
- Confusing register — slotting a casual word into a formal sentence, or an emotional word where a neutral one belongs.
- Grammar mismatch — picking a word that means the right thing but is the wrong tense or part of speech.
A useful home-marking habit: when your child gets one wrong, ask them to say why the correct option is better than the one they chose. Naming the reason ("it collocates with verdict" or "the clue was after the gap") fixes the lesson far more firmly than simply showing the answer.
Timing and pacing
A GL verbal-reasoning paper runs to roughly 50 minutes and packs in a lot of questions, so the working budget is only about 35 to 50 seconds each. Sentence completion should sit at the faster end once the method is grooved, because the answer is either obvious or narrowed to two by elimination. Teach the skip-and-return habit: if two options still look equal after a full re-read, pencil a mark, choose the more likely one, and come back only if time allows. A blank scores zero; a considered guess between two words scores half the time.
How the score works
GL reports a Standardised Age Score (SAS), not a raw mark. Raw scores are converted to an age-adjusted scale of roughly 69 to 141 where 100 is average, so a child who is young for their year is not penalised for their exact age in months. Many grammar schools look for an SAS around 111 or higher, though the exact qualifying score varies by area and by year and is never fixed in advance. The practical message for parents: getting the trickier, clue-dependent sentences right is what lifts the SAS, so it pays to practise the harder examples rather than only the easy ones.
Building the underlying vocabulary
Sentence completion rests on vocabulary. The winning approach is little and often: five to ten new words a week, each used in a sentence your child writes themselves, plus quick flashcard reviews. Reading widely across fiction and non-fiction does the heavy lifting, because it teaches words in context — exactly the skill the question tests. Pair new words with their synonyms and antonyms so the same effort feeds the synonym and antonym question types too.
How to practise this type
Little and often wins here as well: five questions a day for a week beats fifty in one sitting. Our free verbal-reasoning worksheets are organised by exact question type, and the Arena's VR mode gives instant feedback with XP so mistakes turn into quick corrections. When the type feels automatic, fold it back into full timed mock papers so pacing develops alongside accuracy, and use the revision hub to rotate through the other GL types. For the full landscape of what a paper can contain, see the 21 VR types overview, and check the GL papers guide for format details.
Frequently asked questions
How many verbal-reasoning question types are there?
GL draws from 21 named types (Bond labels them Type A to U). Words in context, or sentence completion, is one of them, sitting alongside synonyms, antonyms, odd-one-out, analogies, hidden words, letter codes and number sequences.
Does GL tell you which types will appear?
No. GL does not disclose the mix in advance, so your child needs a reliable method for every type. Practising all of them, then rehearsing under timed conditions, is the safest preparation.
What score does my child need?
Scores are reported as a Standardised Age Score, age-adjusted on a scale of about 69 to 141 with 100 as average. Many grammar schools look for roughly 111 or above, but the exact pass mark varies by region and year and is never guaranteed.