Swedish Political Science: a bibliometric analysis. Citations, productivity measures and rankingshave become reality in modern university life. Many of the bibliometric reportspresented by ranking institutes and non-professional bibliometricians are flawed dueto methodologically unsound procedures. This article discusses three important methodologicalproblems involved in bibliometric studies: 1) number of personnel at universitydepartments; 2) counting of articles from these departments; and 3) countingof citations to these articles. Relating to earlier research (Hix, 2004) it is shown thatthe counting of personnel - a very important reference value - can be conducted inseveral different ways. Following Dale & Goldfinch (2005) we discuss the limitationto political science journals proposed by Hix. There is a large influx of non politicalscientist to the area and a similar outflow of political scientists to other journal categories(e.g. environmental studies). Therefore, the proposed limitation is questioned.Implementing advanced methods for field normalized citation scores (van Raan,2004) we conclude the article with an analysis of Swedish university departments inpolitical science during the period 1998-2005. The result is a promising 33 per centbetter citation scores than the world average, but the downside is a low number of articlesper researcher.
2007. no 4, 339-357 p.