Unseen passage: when a wrong answer, a near-miss and a copy all need different verdicts
An unseen-passage test gives 18 first-year students a short Telugu passage and four questions worth 10 marks (3+3+2+2). The passage reads:
'మా ఊరి గ్రంథాలయంలో సాయంత్రం ఐదు గంటలకు తలుపులు తెరుస్తారు. గ్రంథపాలకుడు రామయ్య ప్రతి పుస్తకం ఇచ్చే ముందు ఒకే ప్రశ్న వేస్తాడు: ఈ పుస్తకం ఎందుకు చదవాలనుకుంటున్నావ్? పిల్లలు మొదట ఇబ్బంది పడతారు; కొన్ని నెలల తర్వాత తమ కారణం చెప్పి పుస్తకం అడగడం అలవాటు అవుతుంది. రామయ్య దగ్గర పుస్తకాల కంటే ప్రశ్నలే ఎక్కువ.'
Question 3 (3 marks) asks the intention behind the librarian's question; the key expects 'పిల్లలు చదవడానికి కారణం ఆలోచించేలా చేయడం'. Nine of the 18 wrote 'పుస్తకాలు ఇవ్వకుండా ఆపడం'; two wrote 'పిల్లలను పరీక్షించడం'. Question 4 (2 marks) asks for the last sentence to be explained in the student's own words; one student copies it verbatim and adds nothing. The teacher has the scripts and one revision period before the papers are returned.
1. Score the two 'పిల్లలను పరీక్షించడం' answers and the nine 'పుస్తకాలు ఇవ్వకుండా ఆపడం' answers out of 3, distinguishing them rather than lumping them together.
Read the worked answer
'పుస్తకాలు ఇవ్వకుండా ఆపడం' - 0/3. The passage contradicts it in the same sentence: he asks the question before giving the book, and the books are given. This is a misreading of the surface, and it is the same answer nine students produced independently - a signal that the item, not only the students, needs looking at (see next task).
'పిల్లలను పరీక్షించడం' - 1/3. Partially defensible: a question does test the child, and the passage shows the children are initially uncomfortable, which is what being tested feels like. But the key asks for the librarian's intention, and the passage supplies the outcome - 'కారణం చెప్పడం అలవాటు అవుతుంది'. Testing is the mechanism; forming the habit of justifying a choice is the purpose. Means-only answers get partial credit, not full.
Marking rule to state aloud: if the question asks 'why did he do it', an answer that describes what happens to the children earns 1; an answer that names the change the librarian wanted earns 3. Full marks require a word from the passage that shows outcome.
Moderation point: distinguishing 0 from 1 here matters more than the 1 mark itself - collapsing them teaches students that any plausible guess is equivalent to a supported reading.
2. Nine of 18 students chose an answer the passage explicitly rules out. Decide whether this is a reading failure or an item-design failure, and revise the item without altering the passage.
Read the worked answer
Verdict: both, but the design share is large enough that the item should change.
Evidence for item-design failure. A distractor that 50% of a class selects is functioning as a second key. The passage never states the intention directly; it states a practice and an outcome. Students who read 'he asks before giving the book' and reason quickly will produce 'he is delaying the book'. Nine independent readers converging on the same wrong inference means the item was not discriminating between readings - it was discriminating between students who guessed which inference the setter wanted and students who did not.
Evidence for reading failure (real but secondary). The last clause - 'పుస్తకాల కంటే ప్రశ్నలే ఎక్కువ' - is the author's evaluation and settles the intention. Students who skipped the final sentence lost the decisive evidence.
Revised item: 'గ్రంథపాలకుడి ప్రశ్న వల్ల పిల్లలలో ఏ మార్పు వస్తుంది? సమాధానానికి ఆధారంగా ఒక పదబంధాన్ని రాయండి.' - now the item demands the outcome and forces an anchor phrase. The 'stop giving books' distractor dies on contact with the phrase.
Principle: when a distractor captures half the class, fix the stem's evidence requirement before you re-teach the passage. Re-teaching a passage the students read correctly does not remove the ambiguity.
3. The Question 4 answer copies the final sentence word for word. Award it, and write the marking note you would put on the script.
Read the worked answer
Award 1/2, not 0 and not 2.
Why not 2: the item says 'in your own words'. A verbatim copy demonstrates location, not comprehension; if copying scored full, the 2-mark item would be testing scanning alone.
Why not 0: the student has found the sentence that the question is about, which is the first step of the task and is worth something in a 2-mark item. Zero would tell a student who located the right evidence that locating evidence is worthless - the opposite of what an unseen-passage section should teach.
Marking note on the script: 'Right sentence, no explanation. Half the marks are for finding it; the other half are for adding a clause beginning with ఎందుకంటే - e.g. రామయ్య పుస్తకం కంటే ప్రశ్నకే విలువ ఇస్తాడు ఎందుకంటే దాని వల్ల పిల్లలు ఆలోచిస్తారు.'
Class rule to announce (30 seconds, before returning papers): 'A quotation earns credit only when it is followed by a because-clause. Copy plus reason is full marks; copy alone is half; reason alone with a wrong quotation is half.' Stating the rule before the scripts go back prevents the appeal that will otherwise arrive from the one student who copied well and writes neatly.