Fluency Without Meaning in Class 5 Hindi
In a Class 5 Hindi classroom of 44 children, the teacher's fortnightly records show a pattern worth examining. A 20-word dictation test averaged 17.2 correct spellings; 33 children read a Grade 4 passage aloud at more than 45 correct words per minute; and 39 could recite the set poem with correct rhythm and pauses. Yet on a four-item test asking why a character in the same passage behaved as she did, only 7 children scored 2 or more out of 4, and 12 answers were word-for-word copies from a commercially printed guide. Separately, Sunita, who stammers, was given 0 out of 10 in oral recitation because she took about 40 seconds to produce her first line; she scored 16 out of 20 on the written paper. The half-yearly examination is five weeks away and the oral component carries 10 of the 100 marks.
1. What does the pattern in the data show about the children's reading? Explain the distinction at work and why more loud reading will not fix the inference scores.
Read the worked answer
The class has decoding and fluency but not comprehension. The simple view of reading separates word recognition from linguistic comprehension; reading comprehension is the product of the two. Dictation and 45 wpm show the first component is largely in place. Only 7 of 44 reaching 2/4 on inference questions shows the second is not.
Loud reading measures pronunciation and pace, not meaning. A child can read a sentence perfectly and still not know what 'why' is asking. The 12 guidebook copies are a symptom, not the disease: they suggest children do not know what a text-based answer looks like and have not been asked to build one.
What will move the scores: vocabulary and background knowledge before reading; a purpose question set before the first read; teacher think-alouds modelling 'how do I know?'; oral discussion in pairs before writing; and requiring one line of text as evidence. Note also that this is a class-level instructional issue, not 37 individual deficits, so changing the task is more efficient than 37 separate remediations.
2. Was Sunita's oral score valid? Redesign the 10-mark oral component for the whole class so that it measures oral expression without penalising a stammer.
Read the worked answer
The score was not valid. It measured speed of onset, which is not the construct being assessed. Stammering is a speech and language disability under the RPwD Act, 2016, so reasonable accommodation is required, not exemption.
Redesigned 10-mark oral component:
- Content and sequence of the retelling: 4 marks
- Vocabulary and expression: 3 marks
- Response to two follow-up questions: 2 marks
- Listening and participation while others speak: 1 mark
Accommodations for all: two minutes' preparation time, permission to begin with a prepared line, option of a small-group or recorded presentation, and an explicit rule that dysfluency is not scored. The 40-second onset is ignored.
Sunita should still present, because oral expression is a genuine learning outcome. The rubric makes that possible without punishing her for a disability.
3. Twelve children copied answers from a guide. What is your response, and what will you change in the next unit?
Read the worked answer
Treat copying as information, not as a crime. First, ask each of the twelve to explain their written answer orally. Most will not be able to, which tells you they copied because they could not construct an answer, not because they were trying to deceive.
Immediate response: no mass punishment; a short individual conversation, return the work with the instruction to add one sentence in their own words, and inform parents of the expectation rather than the accusation.
Changes for the next unit:
- Use an unseen short passage instead of the textbook lesson.
- Add 'which line tells you that?' so answers must be anchored in text.
- Include open questions with more than one defensible answer, marked on the reason given.
- Build in pair talk and a written draft before the fair copy.
- Teach how to use a guide properly, to check after attempting.
Re-test the same skill in three weeks with a new passage to see whether the change worked.