I totally agree that it is kinda weird. I try to vote on any post (both questions and answers) that I read, especially the ones that I answer. Or if I see a question that I haven't voted on and it has generated quite a few interesting answers, I will vote then. Because we have had some quite "uninteresting" questions that went on to produce many answers that proved that the question itself was more interesting and needed to be thought about, more that we initially thought.
Generally, you can see this phenomenon quite often across the various SE sites,i.e. users are more likely to answer a question and/or vote on the answers, rather than vote on the question itself.
There is quite a long discussion on meta.SE:
There are 56 answers to the post, so there are many valid points being made, but here is one that stood out for me:
As I see it, the 'workflow' of reading a question/answer goes like
this:
User opens page
User reads question
- Unless the question is abnormally good or bad, or otherwise provocative, this isn't likely to elicit any
emotional response. It's just a question, carry on.
- User scrolls down and begins reading answers
As there are many answers, and good answers are rewarded by being 'accepted' and also
with increased reputation, this puts the user in the mindset of 'make
the answers better'
The emotional response behind having your answer accepted or
upvoted is "I know stuff, I'm smart, I feel good." Likewise,
conferring that reward on someone else is quite a powerful thing too.
This provides a very strong motivation to rank and provide answers.
- Because of this motivation, people will put a lot of effort into writing answers (like me with this diatribe) and ranking them.
This works very well for providing and filtering good answers, but
there's no such motivation behind voting for questions. For most
questions, the strongest response they are likely to elicit is "I have
that problem too", which while it's strong, is only going to apply to
a small portion of the viewers/answerers.
anon