Thelwall, M. "Weak benchmarking indicators for formative and semi-evaluative assessment of research (Article, English) " Research Eval. 13(1):63-68, April 2004

[Eugene Garfield]

Mike Thelwall : m.thelwall at wlv.ac.uk

TITLE:          Weak benchmarking indicators for formative and semi-
                evaluative assessment of research (Article, English)

AUTHOR:         Thelwall, M

SOURCE:         RESEARCH EVALUATION 13 (1). APR 2004. p.63-68 BEECH TREE
                PUBLISHING, GUILDFORD


ABSTRACT:       Scientometric indicators are traditionally split into two
types: evaluative and relational. The purpose of the former is to evaluate
research-related activities whereas the latter are used to describe or
explore research relationships. In practice, however, there are many
indicators that are of an assessing nature but do not satisfy the
reliability or validity criteria required for use in an evaluative role. The
phrase 'weak bench-marking' is introduced for this kind, applicable to
indicators for which yardstick values can be calculated. Applications
include flagging potential causes for concern when benchmark figures are not
met and identifying areas of apparent good practice. This can be either in
the context of a monitoring role, to assess whether remedial action is
needed, or in the context of providing partial evidence as part of a larger
formal evaluation.



AUTHOR ADDRESS: M Thelwall, Wolverhampton Univ, Sch Comp & Informat
Technol, Wulfruna St, Wolverhampton WV1 1SB, England