Why it is “D”?

[jutfrank]

I'm pushing this, 5jj, because it's a consistent issue here on the forum, which I feel we've never actually gotten to the bottom of, and as a test designer myself I feel there could be great value if we do.

For the purposes of test design, multiple-choice questions typically have only one answer. In every thread on the forum we discuss questions like this, I use the word 'correct answer' to mean the answer that will get you the point. I think we always agree on whether test instructions can be written more clearly ('choose the best answer' versus 'choose the correct answer'), and we almost agree always on whether a test fails in validity. This, however, is one case where I feel the test is valid. It may be a very (perhaps inappropriately) hard question, but it's not invalid, in my judgement. Lexis is unlike grammar in that two different forms can be equally grammatical, but the meaning of words has no such equality—there's almost always a difference in meaning (/use)—and this is what test designers aim to test.

 

[5jj]

I'm pushing this, 5jj, because it's a consistent issue here on the forum, which I feel we've never actually gotten to the bottom of, and as a test designer myself I feel there could be great value if we do.

Ok, but I am not sure I have anything to add to what I have said here or elsewhere.

This, however, is one case where I feel the test is valid. It may be a very (perhaps inappropriately) hard question, but it's not invalid, in my judgement.

That is the problem we have here. I think it's invalid. As I don't think either of us is going to change the other's mind, we are stuck.

Lexis is unlike grammar in that two different forms can be equally grammatical, but the meaning of words has no such equality—there's almost always a difference in meaning (/use)—and this is what test designers aim to test.

But sometimes, it seems to me, the perception of this difference is very subjective. On more than one occasion in the past you have been adamant that the 'correct' answer is the only one acceptable, and have rejected the alternative that I (and sometimes A N Other) have been happy to accept. My view is that if two or more educated, experienced , native-speaking teachers can't agree on an answer, then it is unreasonable to expect test candidates to find it.

 

[jutfrank]

On more than one occasion in the past you have been adamant that the 'correct' answer is the only one acceptable, and have rejected the alternative that I (and sometimes A N Other) have been happy to accept. My view is that if two or more educated, experienced , native-speaking teachers can't agree on an answer, then it is unreasonable to expect test candidates to find it.

Right, but it's rare that we disagree on any language point per se. It tends to be about what is acceptable within the constraints of the test design.