Four Scripts, One Period, Five Days to the Unit Test

Script A needs sentence boundaries, not vocabulary. The ideas are already the hardest part; the run-on means the child does not yet control the full stop as a unit of meaning. Correcting every missing punctuation mark in red produces a page of ink and no learning. Instead: ask the child to read the paragraph aloud and pause for breath - the pauses are the sentence boundaries - then insert stops only there. Follow up with a five-minute oral retelling to a partner before rewriting, so the structure is fixed in speech first.

Script B needs a reason to be personal, and gentle honesty about copying. The mechanics are fine; the problem is that transcription has been rewarded as composition. Do not accuse the child of cheating, but do not award full marks either: state the criterion in advance ('at least three details that only you could write'). Give a prompt: 'What is in your village that is not in the textbook chapter?'

Deliberately not corrected: spelling and style in A until sentence boundaries are secure; grammar in B until the child has written one original sentence. Correcting everything at once produces no usable next step.

No - not on the basis of one script. Matra placement is a late-mastered feature of Hindi orthography; vowel-sign errors are common in Class 5 and are usually orthographic development, not dysgraphia. Isolated spelling patterns do not meet the criteria for a specific learning disability, which requires persistent difficulty across a range of tasks despite adequate instruction, not one error class in one piece of writing.

Collect first, cheaply and within the week:

• A five-minute dictation of ten familiar words read aloud, to see whether the same matra errors appear when spelling is not competing with composing.
• A short oral reading sample, to check whether the child reads matras correctly - if reading is accurate and writing is not, this points to production/orthographic mapping rather than a language disorder.
• The same child's scripts from three earlier months, to see whether the pattern is improving.

If the errors persist across dictation and free writing and the child also shows poor phonological manipulation (blending, counting syllables), a referral is justified. Until then, targeted word-sorting of की/कि and र/ड़ pairs is the appropriate response, and it should be recorded so the evidence exists if a referral is later needed.

Separate the two functions of assessment. A mark serves reporting; feedback serves learning. They do not need to be delivered in the same act.

Practical plan within one 40-minute period:

• Use a four-criterion rubric decided before marking (content and detail, organisation, sentence control, mechanics), each out of 2.5, then collapse to 10. This takes about 20 minutes for 44 scripts if the teacher marks criterion by criterion rather than script by script, and it makes the mark defensible to the head teacher.
• Return scripts with the mark plus one written next step only - not a list. A gets 'end each sentence with a full stop'; B gets 'add three details only you know'.
• Ten minutes of the period goes to a whole-class mini-lesson on the most common error across all 44, which converts individual feedback into instruction and saves marking time.

For Script D specifically: mark what exists against the same rubric, but hold a 90-second individual conference first - one question about the village that the child can answer ('who fetches water at your house?') and an instruction to write those three sentences. Confidence, not rubrics, is the blocker; a zero written in red confirms the child's own prediction and teaches nothing.