Nice to see a question about a new feature in sql 2012.

Would have been a good question except for the ambiguities introduced by hving the result column that should have been labelled salesgoal labelled sales, and asking about a sales value in the result when we couldn't tell whether the mistake in the code was just one column label, or a missing column. As it is, I'm not so sure.

I took a guess as to what the code was intended to be, and got it right. But there are ambiguities here which we have to resolve by pure guesswork, and maybe the low success rate is the result of this. The current success rate is 23%, and the expected result of chosing one at random from each mutually contradictory pair is 25%, which indicates either no knowledge at all in the respondents (which I don't believe) or enough ambiguity in the question to have much the same effect as if answers had been random.

edit: the explanation is a bit substandard too; it's correct (it's a direct quotation from BoL) but there's no comparison going on here so it's irrelevant to this question. A quotation from further on down the BoL page (where it talks about NULL showing up when the third parameter of LEAD is omitted) might have been more relevant, and less misleading - the reference to comparing with next row does somewhat suggest that the sales column and salesgoal column should both be present, maybe that's why only 40% got that half of the question right.