What would ALT-REF look like?

So in the UK the Research Assessment Exercise is being replaced by the REF (Research Exercise Framework - see it's got a different name?). This will evaluate the research standing of individuals and departments and allocate money accordingly. Here's what they say:

"The REF will consist of a single unified framework for the funding and assessment of research across all subjects. It will make greater use of quantitative indicators in the assessment of research quality than the RAE, while taking account of key differences between the different disciplines."

In a twitter exchange with Joss Winn and Brian Kelly this morning I talked about creating an alternative to the REF:

I have in mind something that takes in the sort of distributed identity we have online, so measures activity in blogging, delicious, slideshare, YouTube, twitter, etc. It would need to measure not just activity but influence, impact, etc in some data driven manner.

A key principle must be that it's easy to do and doesn't involve expert panels, peer review etc. What I would be interested then is seeing how an ALT-REF would compare with REF? Would there be any overlap in the top researchers? Would the fields be the same? Would online reputation have any correlation to research funding?

I don't have the right knowledge for this though - it needs data and metrics experts, but it'd be an interesting experiment wouldn't it?

Comments

[Jo Badge]

given the way that google is moving, perhaps they will soon be doing this type of data analysis for us? Maybe the metric could be where you fall on the wonder wheel. Only half joking, I seriously think that this sort of data mashing is coming. Google already ranks by popularity, which is a sort of impact isn't it?