Google.com has tabbed navigation in their main search result template (although they don't style the tabs as such). However, my UX colleagues believe these tabs are not in much use, and they believe this is common for all search UIs: People are not making use of tabbed navigation on result templates.
My question: Is this backed up by actual research?
I have tried to google the issue. I found a quite recent post from Norman Nielsen on best practices for tabs in general and also a meta-analysis indicating that tabbed navigation yielded issues for some library search UIs. But this study does not answer whether there were problems in the individual implementations or if the strategy of tabbed navigation in result templates as such is flawed. Therefore, references to more research would be appreciated.
Needless to say, it would be great if the default relevancy was so good that we wouldn't need filters or tabs. Also, in our ongoing project, we are employing group results to lessen our dependency on tabs. Nevertheless they remain a key concept in our current templates and we would like to identify whether usability issues are of a tactical (implementation) or strategic (conceptual) nature.
I'm including the above screenshot (slightly redacted) as a bonus. We user tested this, and not all users were able to make productive use of the tabs. At that point we didn't have the result counts, which we hope will yield some "scent of information". Also we are considering adding borders and background colors to the tabs, as encouraged by Jacob Nielsen.