Page 285 - Becoming Metric Wise
P. 285

277
                                                            Research Evaluation

              Table 8.1 Publication scores in the Norwegian model
                                              Level 1                 Level 2

              Articles in journals            1                       3
              Articles in books               0.7                     1
              Books                           5                       8


                 In each field at most 20% of the channels accepted as being of scien-
              tific value (levels 1 and 2) belong to level 2.
                 The bibliometric indicator makes a distinction between three times
              two types of publications. Publications themselves are subdivided into
              journal articles, contributions in books and books themselves. Next, each
              category has two levels. Level 1 is the ordinary level, while level 2 con-
              tains the more selective journals and book publishers. These receive an
              extra weight. Different Norwegian research councils determine, each for
              their own field, the level of journals and publishers. A journal which is
              considered to be of level 2 by one research council is also of level 2 if
              used by a researcher of another research field. Finally, publication scores
              are distributed as shown in Table 8.1.
                 Scores are divided fractionally according to the institutes to which
              authors belong (or the author belongs). Fractionalization is considered to
              be essential in order to avoid an inexplicable increase in the number of
              authors. Since this model has been in use the number of Norwegian pub-
              lications in the WoS has increased, but their impact has stayed the same.
              In other words: the model works as it is supposed to.
                 Although the Norwegian model is not especially made for the social
              sciences and humanities, it is constructed in such a way that all fields are trea-
              ted equally. We further note that CRIStin is not unique as a CRIS: more
              and more countries (and fields) design and implement current research
              information systems (Gime ´nez-Toledo et al., 2016; Sales & Sayao, 2015).

              8.9.2 The Flemish Model

              This section is based on Verleysen et al. (2014) and Ossenblok (2016)
              where more detailed information can be found. The Flemish model is
              inspired by the Norwegian one. In 2008 a decree of the Flemish
              Community in Belgium made it legally possible to construct the Flemish
              Academic Bibliographic Database of the Social Sciences and Humanities.
              Further on in the chapter we will use the official (Dutch) abbreviation
              VABB-SHW. Through this database it became possible for non-WoS
   280   281   282   283   284   285   286   287   288   289   290