A learner sees receipt for the first time, hears it once, then receives an instruction to write a sentence. Now the learner must remember the silent p, choose a verb, control the tense and invent a plausible situation. The task looks productive. During this first encounter, it can draw attention away from the word itself.
For a B1 office worker who needs English for invoices and expenses, sentence writing belongs a little later. The first few minutes should secure the word's sound, written form and core meaning. A short production task can follow once those links survive one retrieval.
What the experiments found
Joe Barcroft tested sentence writing with learners of Spanish in two experiments involving 24 new words. In one condition, learners wrote sentences. In the comparison condition, they studied word and picture pairs. Immediate and delayed productive tests favoured the comparison condition. The 2004 study describes the result as an inhibition of word-form learning during initial acquisition. The allocation of time matters because a learner who sees a form four times gets four chances to orient attention, notice a missed letter and restart the sound sequence, while one long encounter can become a single extended struggle. In the first experiment, one group saw each item four times for 6 seconds, while the sentence group had one 48-second encounter. That design limits any simple claim about writing itself. Barcroft ran a second experiment with equal 24-second exposure, and the disadvantage remained.
Timing matters.
A 2006 follow-up used a narrower task. English-speaking learners copied 12 Spanish nouns and did nothing with another 12 while viewing word-picture pairs. Productive recall was higher for the words they had not copied, including a delayed test 2 days later. Copying occupied resources without requiring access to meaning.
These are small, controlled studies of beginners learning concrete Spanish nouns. They do not establish a ban on handwriting, sentence work or output. They show a timing problem.
First contact has a small budget
New vocabulary competes for attention. The learner has to notice a sound sequence, connect it to spelling, identify a meaning and keep the item active long enough to encode it. Sentence writing adds decisions about grammar, collocation and content. A learner who spends 37 seconds deciding whether to write “I lost my receipt” or “I lost the receipt” may rehearse lost and the articles more than receipt. Use the first encounter for a smaller job. Show the word, play or say it, give one plain meaning and ask for a recognition choice. Then remove the word and ask the learner to retrieve it from the meaning or picture. Our article on letting learners hear a new word before seeing it gives one way to protect the sound-spelling connection during this stage.
Attention is finite.
Keep it brief.
For receipt, a teacher might use a shop image, /rɪˈsiːt/, the written form and two choices: “proof of payment” or “a request for payment”. The learner then hears the word again and types it from memory. That sequence still demands effort. The effort stays on the target item.
Move into sentences after retrieval
Sentence writing earns its place when the learner can retrieve the word once without the model in view. At that point, a sentence can develop grammatical and collocational knowledge. The task should constrain the setting so that invention does not consume the whole minute. Give a prompt such as: “You paid NT$1,873 for a client lunch. Write a message asking the restaurant for the document you need.” The learner must recover receipt and place it in a useful request. A second prompt can change the relationship: “Your colleague asks why the expense claim was returned.” Now phrases such as keep the receipt, attach a receipt and ask for a receipt have a reason to appear.
Now write.
Word writing can also behave differently inside a meaningful context. Candry, Elgort, Deconinck and Eyckmans compared writing with meaning inferencing in contextualised vocabulary learning. Their 2017 study treats attention to form and attention to meaning as separate demands that a lesson must manage. Its design differs from Barcroft's isolated word-picture work, which is exactly why broad rules travel badly.
When early sentence writing can work
An advanced learner meeting a familiar word family is a counterexample. Someone who already knows allocate and allocation may benefit from writing a sentence with reallocation immediately because most of the form and meaning are already available. A transparent cognate can also require less encoding than an opaque item such as aisle. The quality of the sentence task changes the load too. Completing “Please keep your ___” asks far less than inventing a sentence from an empty line. Pair discussion may help with meaning while also adding speech, turn-taking and social pressure. A worksheet label cannot tell us how much processing the task requires.
This boundary moves.
I'd use delayed sentence writing for words whose form still slips after one look. For a word that the learner can already pronounce, spell and recognise, waiting may waste time.
A twelve-minute sequence
Choose six words from one useful setting. Spend the first 4 minutes on sound, spelling and core meaning, then run a closed-book retrieval with pictures or short definitions. Review any failed item once. During the next 5 minutes, place the recovered words into short, constrained messages. Use the final 3 minutes for a fresh retrieval in a different order. Check form and meaning separately the next day. Say the meaning and ask for the English word. Show the word and ask for a quick meaning choice. Only then ask for one new sentence. If receipt survives those two routes, the learner can spend attention on the client lunch, the NT$1,873 and the missing attachment rather than rebuilding the word from scratch.