In short
- A definition question checks whether a pupil can match a word to an explanation; a context question checks whether they understand or can use the word in a sentence.
- Before you write a question, decide whether it tests the word's meaning, its form or its use.
- Three options are usually enough, as long as every wrong option is plausible and only one answer is right.
- Most weak questions fail in the same few ways: the sentence gives the answer away, two options are correct, or the wrong options are a different part of speech.
Most vocabulary quizzes ask pupils what a word means. That is useful, but it is not the same as knowing the word when you meet it in a story. This article shows you how to test the word where it lives: inside a sentence from the text your class has just read. It gives a five-step method, three question forms, rules for wrong options and three worked examples at A2.
What is a vocabulary-in-context question?
A vocabulary-in-context question shows the pupil a real sentence and asks them to recognise, choose or use a word inside it. The sentence comes from the text they read. The question checks whether they understand the word in that sentence, not whether they can repeat a definition from a list.
Compare two questions about the word "arrived". A definition question asks: "What does 'arrive' mean?" A context question asks: "Mia ____ at school at nine o'clock. a) arrived b) went c) came back." The second one makes the pupil read, think about the sentence and choose. It is closer to what reading asks of them.
Read (2000), in his book on vocabulary assessment, describes this as a difference between context-independent and context-dependent testing. In a context-dependent item, the pupil has to use the surrounding text to give the right answer. Read (2000) also separates tests of vocabulary on its own from tests where vocabulary is embedded in a wider reading task. A quiz after a story sits naturally at the embedded, context-dependent end.
Why test a word inside a sentence?
Knowing a word is more than knowing its meaning. Nation (2013) describes word knowledge in three broad parts: form, meaning and use. Form is how the word is spelt and said. Use covers grammar, typical word partners and where the word fits. A definition question touches only meaning. A context question can reach all three.
The sentence you choose also matters. Webb (2008) gave 50 Japanese university students ten unknown words in short sentences taken from graded readers. One group met the words in sentences with more clues to meaning; the other met them in sentences with fewer clues. The group with the richer sentences scored significantly higher on both tests of meaning. There was no significant difference on tests of spelling and word form.
Webb’s learners were adults and the words were disguised forms, so treat this as a guide rather than a rule for younger pupils. The practical point holds: a sentence with no clues to meaning makes a poor test sentence. Pupils also need several meetings with a word, as How Many Times Must a Learner Meet a Word? explains.
A five-step method for writing the question
Work from the text outwards. Each step takes a minute once you are used to it.
- Take a sentence from the text. Choose one the pupils have just read. It should make sense on its own and contain at least one clue to the target word's meaning.
- Choose the target word. Pick a word that matters for understanding the story and that pupils at this level should be learning. Avoid names and very rare words.
- Decide what you are testing. Meaning (does the pupil know what the word refers to here?), form (can they recognise or produce the right spelling or word form?) or use (do they know how the word behaves in a sentence?). One question should test one thing.
- Write the stem. Keep it short. Use the sentence as it appears in the text, or a close version with one gap. Give a clear instruction such as "Choose the word that completes the sentence."
- Write the wrong options. Make each one plausible to a pupil who does not know the word. Base them on mistakes pupils really make. Check that only one answer is right.
Step three is the one most often skipped. Without it, you tend to write a question that tests two things at once, and a pupil can fail it for a reason you did not intend.
Three question forms that work
Gap-fill
Remove the target word from the sentence and give three options. This tests recognition of meaning and, if the options differ in form, recognition of form. It is the quickest form to write and to answer.
Choose the meaning in this sentence
Show the whole sentence with the target word in bold. Ask what the word means here. This is useful for words with more than one meaning, because the sentence decides which meaning is right.
Choose the sentence where the word is used correctly
Give three sentences that all contain the target word. Only one uses it correctly. This tests use: grammar, word partners and fit. It suits pupils from A2 upwards, because they must read three sentences instead of one.
| Form | What it mainly tests | Best for | Watch out for |
|---|---|---|---|
| Gap-fill | Meaning, sometimes form | Quick checks after reading, all levels | A sentence that gives the answer away |
| Meaning in this sentence | Meaning in context | Words with several meanings | Options that are all near-synonyms |
| Correct use | Use: grammar and word partners | A2 and above | Wrong sentences that are only slightly odd rather than wrong |
Other forms exist, and comprehension questions about the plot work differently. Reading Comprehension Question Types, With Examples covers those.
Rules for good wrong options
Wrong options, often called distractors, do most of the work in a multiple-choice question. Haladyna, Downing and Rodriguez (2002) reviewed 27 textbooks on testing and 27 research studies. They produced 31 guidelines for classroom multiple-choice items. Several apply directly to vocabulary questions.
- Only one option is right. Read each option in the sentence. If two work, rewrite one of them.
- Make every wrong option plausible. An option nobody would choose tells you nothing. Haladyna and colleagues list this among their guidelines.
- Use typical pupil errors. Their guidelines also recommend basing distractors on the mistakes students really make. For vocabulary, that means words pupils confuse, look-alike spellings and common grammar slips.
- Keep options the same kind of thing. All options should be the same part of speech and a similar length, so grammar does not give the answer away.
- Avoid clues in the stem. A word in the stem that echoes the right option makes the question easier than you meant.
How many options? Rodriguez (2005) combined 80 years of research on this question in a meta-analysis. He concluded that three options are optimal: more three-option items fit into the same testing time, without harming the quality of the scores. Haladyna, Downing and Rodriguez (2002) say the same: writing a fourth good option is rarely worth the effort. For a short quiz after reading, one right answer and two strong wrong ones is enough.
Three example questions at A2
Here is a short text written for this article, at about A2 level, followed by one question in each form.
Classroom example
Text: Mia's bike was broken, so she walked to school in the rain. She arrived late and her shoes were wet. Her teacher smiled and gave her a towel.
Question 1 (gap-fill, tests meaning): Her shoes were ____ because it was raining. a) wet b) dry c) warm. Answer: a. "Dry" is the opposite that a guessing pupil may pick; "warm" fits the grammar but not the sentence.
Question 2 (meaning in this sentence, tests meaning): Her teacher gave her a towel. What is a towel here? a) a cloth to make something dry b) a warm drink c) a school book. Answer: a. All three are nouns, and all three are things a teacher could give.
Question 3 (correct use, tests use): In which sentence is "arrived" used correctly? a) Mia arrived at school at nine o'clock. b) Mia arrived her bike to school. c) Mia arrived the towel to her teacher. Answer: a. The wrong options use "arrive" with an object, a mistake learners often make.
None of these asks for a definition from memory. In each, the wrong options are the same part of speech as the answer, and the question tests one thing only.
Common mistakes and how to fix them
Check each question against this list before you use it.
- The context gives the answer away. If the stem says "She was so hungry that she ate everything. She was very ____", pupils need only the first sentence. Fix it by keeping a clue but not a restatement.
- Two options are correct. "She arrived ____ ." with options "late", "early" and "home" has three grammatical answers. Fix it by using the full sentence from the text so the context decides.
- Wrong options are a different part of speech. If the answer is a verb and the other options are nouns, grammar alone gives it away. Fix it by making all options the same word class.
- The question tests two things. A gap-fill where the options differ in both meaning and spelling makes it hard to know why a pupil failed. Fix it by varying one thing at a time.
Grammar questions follow the same logic but need their own care. Teaching Grammar Through Stories looks at those. For a sense of which words to target at each level, see How Many Words Do You Need at Each CEFR Level?
How iRead handles this
In iRead, quiz questions are generated from the story the class is reading. They use real sentences from the story, the grammar that appears in it, the plot and the moral. Every quiz is levelled to the CEFR, from A1 to C2. Each quiz answer also feeds a progress record for that word, and teachers see retention per word and per class.
The core principle: Test the word where the pupil met it. Take the sentence from the text, decide what you are testing, and make every wrong option one a real pupil might choose.
See it in iRead: iRead quizzes are generated from each story, using its real sentences, its grammar, its plot and its moral, levelled from A1 to C2.
See how quizzes workKey takeaway
A good vocabulary-in-context question uses a real sentence, tests one thing, and has one right answer and two plausible wrong ones. Check every question for giveaway clues, double answers and mixed parts of speech before you use it.
Frequently asked questions
What is the difference between a definition question and a context question?
A definition question asks what a word means, out of any sentence. A context question shows the word inside a real sentence and asks the pupil to recognise, choose or use it there. The second is closer to real reading, because the pupil has to use the sentence to decide.
How many options should a vocabulary question have?
Three is usually enough: one right answer and two plausible wrong ones. A meta-analysis by Rodriguez (2005) found that three-option items give scores of similar quality to four- or five-option items, and you can ask more of them in the same time.
Where should the wrong options come from?
From the mistakes your pupils really make. Good sources are words they confuse, words with a similar spelling, the opposite meaning, and common grammar errors with the target word. Every wrong option should be the same part of speech as the right answer.
Are vocabulary-in-context questions suitable for A1 pupils?
Yes, in simple forms. At A1, use short gap-fills with a picture-able word and three short options. Save the "choose the correct sentence" form for A2 and above, because it asks pupils to read and compare three whole sentences.
Sources
- Read, J. (2000). Assessing Vocabulary. Cambridge University Press (Cambridge Language Assessment series). doi.org/10.1017/CBO9780511732942
- Nation, I. S. P. (2013). Learning Vocabulary in Another Language (2nd ed.). Cambridge University Press. doi.org/10.1017/CBO9781139858656
- Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign Language, 20(2), 232-245. doi.org/10.64152/10125/66826
- Haladyna, T. M., Downing, S. M., & Rodriguez, M. C. (2002). A review of multiple-choice item-writing guidelines for classroom assessment. Applied Measurement in Education, 15(3), 309-334. doi.org/10.1207/S15324818AME1503_5
- Rodriguez, M. C. (2005). Three options are optimal for multiple-choice items: A meta-analysis of 80 years of research. Educational Measurement: Issues and Practice, 24(2), 3-13. doi.org/10.1111/j.1745-3992.2005.00006.x