Measuring Science (II) Morten Brendstrup-Hansen • No science without scientific publications • Scientific publications are direct and tangible products of scientific activity • Therefore, the idea of a measure of scientific performance based on publications is sound and straightforward Peer review vs. bibliometric analysis • Peer review may be accurate, but is time consuming and may easily be (suspect of being) biased • The accuracy of a scholarly founded formal bibliometric analysis may be examined and its validity may be discussed • A bibliometric analysis is based on publicly available (and easily collected) data In bibliometrics, what should we count/calculate? Where do we find the data for bibliometric analyses? Total number of papers • Measures quantity, but does not take quality into account; does not give due weight to influence Number of 'quality papers' e.g. defined as papers in ISI-journals • Relies on the inclusion in a particular journal as a measure of quality instead of trying to assess the actual quality of the paper Total number of citations • Measures influence, but may be inflated by a small number of unrepresentative big hits Number of citations per paper • Punishes productivity Number of papers with >x citations • Combines publication data with citation data • Thus rewards quality as well as quantity if a fair value of x is chosen, • but different values of x need to be decided upon for different fields of research h-index • A scientist has the index h if h of his or her papers have at least h citations each - Hirsch JE (2005) PNAS 102(46): 16569-16572 Nc h h Np h-index • A scientist has the index h if h of his or her papers have at least h citations each - Hirsch JE (2005) PNAS 102(46): 16569-16572 Nc h h Np g-index • A set of papers has a g-index g if g is the highest rank such that the top g papers have, together, at least g2 citations Egghe L (2006) Scientometrics 69(1):131152 This is only the beginning • More indices will probably be coined • Indices should be validated e.g. by testing their predicative power Software link • Publish or Perish is a piece of software that calculates several bibliometric indices from Google Scholar data. It is provided free of charge at http://www.harzing.com/