Future UK RAEs to be Metrics-Based

[Stevan Harnad]

On Fri, 11 May 2007, J.F.Rowland at lboro.ac.uk wrote:

> Sadly, in their RAE (research assessment exercise) assessments of
> individual academic staff, many UK university administrations still look
> solely at impact factor.  This week's Times Higher Educational
> Supplement contains one very distressing report of the kind of effect
> this one-dimensional measurement can have on a researcher's career.  In
> addition to its other flaws, dependence on journal impact factor alone
> gives undue influence to a commercial organisation (Thomson ISI) which
> consistently over-represents journals published in North America, and
> under-represents ones published in languages other than English.

This is all true, but stay tuned! It is all poised to change, radically.

With OA, OA scientomentrics, and the new metric RAE, a rich and diverse array
of new metrics will be available, of which the Journal Impact Factor will merely
be one among many. Each of the metrics will be weighted by its predictive power, 
with different profiles for different fields.

    "Future UK RAEs to be Metrics-Based"
    http://users.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/5250.html
    
    "Let 1000 RAE Metric Flowers Bloom: Avoid Matthew Effect as
    Self-Fulfilling Prophecy"
    http://users.ecs.soton.ac.uk/~harnad/Hypermail/Amsci/5417.html

    Harnad, S. (2007) Open Access Scientometrics and the UK Research
    Assessment Exercise. In: Proceedings of 11th Annual Meeting of the
    International Society for Scientometrics and Informetrics.
    Madrid, Spain.
    http://eprints.ecs.soton.ac.uk/13804/

(Note that the past versions of RAE did not look at citation counts at all, whether
individual counts or journal impact factor; however, many departments, in preparing
their RAE returns, did, and used journal impact factor as one of their criteria for
deciding which 4 publications to return. In future, the OA metrics themselves will be
picked up directly, and there will be no point in restricting them to 4 publications
-- or, alternatively, there could be a "best 4 metric," with its own weight, along with
alongside all the other many metrics.)

Stevan Harnad