My views on this have evolved since the beginning of the private beta. But I've come to the conclusion that questions shouldn't be downvoted due to a lack of research. Here's why.
There are a lot of questions that can be answered through a quick Google search. For example, I'm sure that a quick Google search would be enough to find an answer to the question "Why was Bob Dylan awarded the Nobel prize?"
However, just because other people have written about something doesn't mean that we shouldn't write about it as well. "Why was Bob Dylan awarded the Nobel prize?" is an extremely interesting question, and it would be a pity if people couldn't ask and answer that question on this site just because other people have answered it elsewhere.
And of course, this site gets most of its traffic from search engines. If you can't ask questions that other people are Googling, then you're going to miss out on a huge chunk of traffic.
I find that research is usually a shorthand for "effort", e.g. spend some time trying to figure out something before you come to us. But I get the sense that this is just effort for effort's sake, rather than effort that leads to improvements in a question. We want people to ask interesting questions such "Why was Bob Dylan awarded the Nobel prize?"--downvoting these questions means that they won't be asked and that we won't have a chance to answer them.
I think there are far better standards that can be used as markers of question quality. (I'm happy to discuss these in chat or a separate meta post. Of course, I've found myself downvoting less and less questions on the site as I slowly learn what types of questions are interesting/lead to good answers. It turns out that most questions are interesting and lead to good answers.)