Most designers believe in the validity of research. They seriously intend to plan and execute said research every time they start a new program, but the effort is often derailed by money, timing, and follow-through.
I came across an excellent summary of research initiatives and applications on Core77 that could help you select the right research effort, so that you at least have a good chance it'll get done. Often you're aiming too high when research programs fall apart. What you need to do depends on the hypothesis you're testing - and less is often more. Overshooting won't really get you better actionable results, just more work and more money down the tubes.
Summary table of research initiatives and app's (though I'm less fond of focus groups for ANY application than this author):