Nice simple question. Thanks for a gentle start to the week.
Uses an "undocumented" (maybe Paul Randall's blog entries is generally documentation as reliable as BoL, so maybe dbcc ind is not quite undocumented) dbcc function, it probably should have had a pointer to something about that in the explanation.
Given that it was only possible to select 1 answer, and rebuilding an index certainly isn't going to build any page splits, it ought to be be very difficult to get it wrong. But 36% of responders so far have achieved that.
Until I noticed that 36% figure I though tat adding a "both" option would make it only marginally harder, since it would still be a select one only question and it makes no sense to select any two of the three possibilities (if both were selected, both the others would logically have to be; and if it wasn't, only one of the others could be) so the SQLServerCentral QoTD code would make it a single choice question, thus givibg the game away. Making it a multiple choice question could make it a bit harder if enough answers plausible at first sight could be dreamt up, but I thought that would be rather difficult to achieve. So perhaps it should have just been "will ALTER INDEX...REORGANISE fix it or not", to make it difficult enough to be worth a whole point. The 36% wrong answers figure suggests that my thoughts were not in line with reality!