Every sentence makes sense on its own
These are one pupil's answers to a Flyers Reading and Writing Part 5 text, where you choose one of three words for each gap.
Sam wanted a present for his mother. He looked in three shops, (1) ___ he could not find anything he liked. [but / so / because] Finally he chose a blue scarf. His mother loved blue, (2) ___ he knew she would like it. [but / so / because] At home, his sister and brother saw the scarf. (3) ___ all wanted to try it on. [It / He / They]
She chose: (1) so, (2) so, (3) It.
In class she reads aloud accurately and can explain the meaning of any single sentence from the text when it is pointed at. She scored full marks on the earlier Part 4, a multiple-choice vocabulary cloze. She has no difficulty with 'but', 'so' or 'because' as words, and uses them correctly in speech.
1. What one weakness explains all three answers? Use the evidence in the setup, not just the answers themselves.
Read the worked answer
All three answers are defensible if you read only the sentence containing the gap, and all three are wrong once you read the sentence before it.
- (1) 'so' signals a result. The second clause is a failed result - he looked in three shops and found nothing - so the relation is contrast: 'but'.
- (2) 'so' is correct here. This one is right, but it is right by the same default, not by a different decision.
- (3) 'It' grabs the nearest singular noun, 'scarf', and ignores that the subject must be 'his sister and brother'.
So the pattern is not a vocabulary gap and not a narrow grammar gap. She is processing one sentence at a time and treating each gap as a local choice. The support for that reading is that she explains each sentence correctly in isolation, and she is fine on Part 4, which by design can be answered within a single sentence. Nothing in her performance suggests she cannot understand the relations; she is not looking for them.
2. Her teacher proposes a lesson defining 'but', 'so' and 'because' with examples. Why will that not fix these answers?
Read the worked answer
Because she already knows the words. The lesson would produce a child who can recite that 'so shows a result' and will still write 'so' at (1), because the failure happens before the word choice: she has not compared the two clauses to see whether what follows is a result or a contradiction.
Reteaching the meaning of the connector trains the step that is already working. The missing step is the look-back: read the previous sentence, decide the relation, then choose.
A cheap diagnostic: cover gap (1) entirely and ask her to say what happened in her own words ('he looked everywhere and found nothing'). If she can say 'and found nothing' but still picks 'so', the problem is confirmed as procedural, not lexical.
3. Give one concrete routine that targets the missing step, and say how you would know it had worked.
Read the worked answer
Relation-first gap filling.
Before any word is chosen, the pupil reads the two sentences with the gap covered and marks the relation with an arrow on the page: → (same direction / result) or ⇄ (opposite / surprise). Only then do they look at the three options.
Do it first with pairs that are unambiguous, e.g. 'It was raining. ___ we stayed inside.' (→) versus 'It was raining. ___ we still went out.' (⇄), then transfer to the Sam text.
For (3), a separate thirty-second move: underline every noun in the two sentences, then ask 'how many people are we talking about?' before looking at the options.
Evidence it has worked: on a new unseen text she marks the relations correctly before choosing, and - the real test - she gets a pronoun item right where the nearest noun is the wrong referent. If she is still selecting by the word that sits closest to the gap, the routine has been done as an exercise, not learned.