Fluent Readers Who Cannot Infer: A Class 7 Comprehension Gap
Your first solo English period is with Class 7 in a government school in Muzaffarpur: 48 students, one shared textbook for every three students. The lesson is a two-paragraph passage about a village fair. While reading you find that 41 of 48 students can read the passage aloud without stumbling. Afterwards, only 11 answer a literal question ('Where was the fair held?'), 4 answer an inference question ('Why did the narrator's father refuse to buy the toy?'), and none can say what the passage suggests about the family's income. Your mentor's advice is to 'do more reading aloud practice'. On the way out a boy says, 'Sir, he don't have money,' and two students laugh.
1. What do the four numbers (41, 11, 4, 0) actually tell you to teach next, and why not more reading aloud?
Read the worked answer
The numbers describe two problems, not one. At 41 of 48, decoding is largely in place; the drop to 11 at literal and 4 at inference places the failure at meaning, not at sounding out. More read-aloud rounds would train a skill already at ceiling and leave the gap untouched. The zero at the evaluative question is a separate signal: the question type is unfamiliar, so nobody has been asked to interpret before.
Two checks worth ten minutes: ask the 7 who stumbled to read three sentences aloud (are they the real decoding group?), and ask five fluent readers to explain two words from the passage in their own words. If vocabulary is the block, pre-teach four or five words before the next reading. If inference is the block, model one think-aloud with the sentence visible. With one book per three students, teach one paragraph well rather than two badly.
2. Design a 15-minute post-reading sequence on the same passage with no photocopies.
Read the worked answer
- 3 min, literal: 'Where was the fair held?' Then the follow-up that does the work: 'Which words in the text told you?' The answer must be quoted.
- 4 min, inference: 'Why did the father refuse the toy?' and 'Which sentence makes you think so?' Accept only evidence-linked answers.
- 4 min, pairs: each pair writes its own question about the family's income; two pairs read theirs out. Pair talk first, because with 48 students a choral 'yes' hides twenty non-answerers.
- 4 min, written: one sentence using the frame 'The family was poor because…' on a slate or in the notebook.
Collect two slates on the way out. Every question is answerable only from the text, so the class practises using the passage rather than guessing from the title. The sentence frame gives the weakest students a way into a sentence they cannot yet start alone, and the four minutes of writing give you evidence that the oral answers did not.
3. The boy says 'he don't have money' and two students laugh. Do you correct him on the spot? Justify your choice.
Read the worked answer
Recast; do not correct publicly. Reply to the meaning first - 'Yes, he doesn't have money' - with light stress on doesn't, then continue the conversation.
Reasoning: the boy's message was clear. Interrupting it to teach grammar tells 48 students that speaking is a risk, and the two who laughed have already made it costly. Public correction fixes the error in one student and the fear in forty. But the error should not be dropped: note it in an error log, and in the next five-minute language slot put he/she/it + doesn't on the board against I/you/we + don't, with three oral examples and one written item. If the same error appears in five notebooks, it becomes a whole-class item rather than one child's mistake. Accuracy is served by planned attention, not by ambush.