Since the advent of No Child Left Behind, public education has been throwing the term "research based" into nearly everything. I'm all for schools having a reason behind what they're doing...but I don't think that the term has been fully defined. Or, perhaps I should say there is no definition that is commonly shared/used.
I have my own ideas about what constitutes a research base. It's been shaped by my recent doctoral work. When I think about the term now, I envision reading the actual research. Studies published within the last five years are preferable, but sometimes it helps to go back and see the staying power of various ideas. Looking at the original work also gives some insight on the relevance of current work.
Now, here is where I start to diverge from most of the educators I run with. A book by an educational researcher? This does not make something research based. By "something," I mean whatever project you're working on: a piece of legislation, a grant proposal, review tools for curriculum, budget decisions, and so forth. The book itself might have some synthesis of work, but it is not the original research. It's filtered and shaped. It often does not contain the opposing citations. You are trusting that the author has correctly interpreted things (and has the background to do so) and picked the very best information available. The book may well make particular ideas accessible to a larger audience...but it, in and of itself, is not a research base. At best, you can say that you've gained some background knowledge in reading it.
I understand that when selecting pieces of actual research to use that we can get into a whole other quagmire about the rigor applied to the design. But I think most of us can tell whether or not something passes the sniff test. We're not writing dissertations. We just need to look as see the kinds of situations where a particular strategy works before jumping to claim "It's research based!"