Hands Up in Class, Blanks in the Unseen Passage
A Class 6 English teacher has 38 students; 31 speak Santhali or Kurukh at home and Hindi in the playground. Her lesson on the 90-word passage 'The Postman' always runs the same way: she reads it aloud twice, translates each sentence into Hindi, writes seven word-meanings on the board, and then asks the five questions printed at the end. In class more than thirty hands go up and the answers are correct. In the monthly test, using an unseen 80-word passage with five wh-questions, only 9 of 38 scored above 5 out of 10. The scripts show 'Why was Ravi late?' answered as 'Ravi was late.', 'How many letters did he carry?' answered by copying an unrelated sentence, and three students leaving all five blank. Asked to explain, the teacher says the children 'cannot read English' and plans to add a full Hindi translation of every new passage plus twenty more word-meanings per lesson.
1. Explain the gap between the classroom performance and the test performance.
Read the worked answer
The classroom was measuring something different from what the teacher thought.
- In the lesson, the teacher's translation supplied the meaning. Once the passage has been translated sentence by sentence, no student needs to construct meaning from English; the comprehension work has already been done by the teacher.
- The printed questions follow the order of the text, so a student can match a question word to a nearby sentence and copy it. That is a word-matching strategy, and it succeeds in class and on the textbook exercise.
- The unseen passage removes every support: no translation, no teacher intonation, no predictable question order. What is left is genuine comprehension, which has not been taught.
- The error patterns are diagnostic, not random. 'Why was Ravi late?' answered 'Ravi was late.' shows that the student located the right sentence but did not process the causal relation - the question word was decoded and discarded. Copying an unrelated sentence for 'How many' shows keyword hunting without syntax. Five blanks indicate the passage is at frustration level for those three students.
- 9 of 38 above 5/10 confirms the unseen text was pitched above independent reading level for most of the class.
'They cannot read English' is too coarse a conclusion. They can decode and locate; they cannot yet build a situation model, and they have no strategy for an unknown word. Those are teachable.
2. Redesign one 40-minute lesson on an unseen 80-word passage, in three stages, with reasons.
Read the worked answer
Pre-reading, 6 minutes. One picture from the passage, three prediction questions answered in single words, and only four key words taught - chosen because they block meaning, not because they are new. Reason: pre-teaching twenty items overloads working memory; four usable items plus prediction give the reader a frame to hang meaning on.
During reading, 18 minutes. First read for gist: 'Who is in the story? Where are they?' - one-word or one-line answers. Second read in pairs with three inference questions, and the rule that every answer must be supported by underlining the sentence that proves it. Reason: this teaches evidence-based answering and attacks the copy-a-sentence habit directly by asking students to justify rather than find.
After reading, 12 minutes. Pair retelling in two sentences each, then one written question of the same type that will appear in the test. Reason: the test format should be practised in the lesson, otherwise the assessment measures format familiarity rather than reading.
Close, 4 minutes. Exit ticket with one inference question, collected. Reason: gives the teacher same-day evidence of who still cannot infer, so the next lesson is aimed rather than repeated.
Translation is used only after students have attempted the item, and only for the word or sentence that blocks the whole answer.
3. Critique the plan to translate every new passage and add twenty word-meanings per lesson. How would you assess comprehension instead?
Read the worked answer
The plan treats a strategy problem as a vocabulary problem and will not transfer.
- Translating every passage keeps the teacher as the meaning-maker. The student practises reading Hindi, not English, and the skill disappears the moment the translation stops - which is what the test already showed.
- Twenty meanings per lesson exceed what can be retained; most will be gone in a week, and the lesson time comes out of actual reading. Five high-frequency words recycled across a week, plus the strategy of guessing from context and word parts, is more durable.
- Word lists teach items; comprehension needs inference, reference (what does 'he' refer to), sequencing and main idea.
Assessment that does not depend on translation:
- short constructed responses with a 'which sentence tells you that?' requirement;
- true/false with correction ('Ravi was late because the bus broke down' - correct the false part);
- sequencing four events from the passage;
- matching a pronoun to its referent;
- a two-sentence oral or written retelling scored on who/what/where;
- cloze on connectives (because, so, but), which tests whether relations were understood.
Finally, test in the format you taught: unseen passage, five questions, at least two inferential. Otherwise the score reflects format shock rather than reading ability.