Grammar Keys in Doubt Before Placement
During piloting of a 25-item grammar and vocabulary placement test, 90 incoming students take the test. Item 7 reads: 'If I ___ more time, I would have finished the report.' Options: A had, B would have, C had had, D have. The intended key is C. Overall, 31% choose A, including several students who later placed in the advanced class. Item 11 tests 'used to' versus 'would' for past habits, but its context permits both answers. Item 18 reads: 'The committee ___ divided on the issue.' Options: is, are, has, have; the key is 'is', but 28% choose 'are'. Teachers disagree about whether item 18 tests grammar or arbitrary prescriptive usage. The test must be finalized in three days; changing many items would require re-piloting, which is not possible.
1. Which items can be retained with a defensible key, and which must be dropped or revised without re-piloting?
Read the worked answer
Item 7 should be dropped from scoring unless the syllabus explicitly tests careful written third-conditionals. The intended key C is standard: 'If I had had more time, I would have finished.' But option A is a common spoken reduction, and advanced students choosing A may be processing meaning rather than missing the grammar. A distractor that attracts strong candidates is not automatically wrong, but here it creates a disputed key.
Item 11 must be dropped because both 'used to' and 'would' are defensible in its context. Item 18 is also not a clean key: 'is divided' and 'are divided' vary by standard variety and emphasis, so it should be dropped or rewritten with a clearly singular subject, such as 'The committee has issued its report.'
2. How should the placement decision be reported given the three problematic items?
Read the worked answer
Use the remaining 22 items and report the placement decision on that reduced set, with a note that three items were excluded for key defensibility. Do not prorate by guessing or assign credit to both keys, because that changes the construct.
For future forms, write items with one clearly correct answer in the target standard, avoid collective nouns whose agreement varies, and include a spoken-versus-written usage note in the item-writing guide. If the program needs a total out of 25, use equating from the pilot rather than inventing scores. The immediate constraint is transparency: candidates and teachers should know which items were excluded and why.
3. What does the pattern of advanced students choosing A on item 7 imply about the test's construct?
Read the worked answer
It suggests the item may be testing metalinguistic carefulness under written conditions rather than general grammatical competence. Advanced students who choose A may be using a natural spoken form or reading 'If I had more time' as a meaning unit and overlooking the perfect auxiliary that 'would have finished' requires in formal written English.
That matters for placement: if the construct is communicative grammar, the item may penalize a normal spoken reduction. If the construct is academic written accuracy, C is defensible, but the item still needs a clearer context. In either case, the ambiguous distractor should not decide placement on its own.