about

REF2014, the latest round of research assessment in the UK, was completed late in 2014.

This micro-site contains the results of an analysis of the REF results for REF Sub-panel 11 – Computer Science and Informatics, comparing the REF scores with citation-based scores.

The main output of REF is an institution-by-institution research quality profile within different disciplines, However, The Computer Science and Informatics sub-panel also published profiles for sub-areas of computing, which are, or have the potential, to influence decision making in computing departments

The citation-based analysis suggests that, despite every effort to ensure fairness, substantial latent bias has emerged between these sub-areas. Not only does this mean that the REF computing sub-area profiles are misleading, but this may also have affected textual feedback to departments, funding to institutions and possibly gender neutrality.

Draft report (v2, 9th May 2015): On REF and citations – how much can we read about sub-areas of computing?

Summary paper (draft v2, 9th May 2015): Citations and Sub-Area Bias in the UK Research Assessment Process

Note: The site author, Alan Dix, was a member of REF 2014 sub-panel 11 “Computer Science and Informatics”, and as such is bound by confidentiality. The analysis is therefore based solely on public domain data and processes, as described in the data section.

2 thoughts on “about

  1. Can you work out how many outputs are co-authored by PHD students? we’re interested in possibly using REF based evidence to argue that the EPSRC should revert to allowing CS PhDs to be funded in projects, not just in the narrow, and sometimes compromised (e.g. gchq captured) CDT programmes.
    I mean every other funding agency in the world (NSF, CNRS, DFG etc etc) funds project phds….but we’d like to have uk specific data to support argument…

  2. I can;t think of an easy way other than spidering the Google scholar pages and then mining individual institution pages – I seem to recall that this kind of thing was a (the?) major demonstrator for the Sem Web IRC, but like so much research, once the project ended … 🙁

    I’d like to get female authorship as well, but have the same problem.

Leave a Comment

Your email address will not be published. Required fields are marked *