228
views
0
recommends
+1 Recommend
1 collections
    0
    shares

      If you have found this article useful and you think it is important that researchers across the world have access, please consider donating, to ensure that this valuable collection remains Open Access.

      Prometheus is published by Pluto Journals, an Open Access publisher. This means that everyone has free and unlimited access to the full-text of all articles from our international collection of social science journalsFurthermore Pluto Journals authors don’t pay article processing charges (APCs).

      scite_
       
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      The publish or perish book

      Published
      book-review
      a , * ,
      Prometheus
      Pluto Journals
      Bookmark

            Main article text

            by Anne-Wil Harzing, Melbourne, 2010, 246 pp., ISBN 978-0980848519

            Don’t be misled by the title. This is not a tome about the pressure on academics to publish, or current issues in the scholarly communication process. It is a book, published by the author herself, about a piece of software she has developed called Publish or Perish. This software, which is free of charge to download and use, is intended to be used to carry out citation analyses on scholars based upon records held in Google Scholar of publications by that scholar. And this represents the fundamental problem with the book; despite the author’s protestations, Google Scholar is an unreliable way of assessing scholarly outputs – but more of that later.

            The book provides a well-written step-by-step description of how to use the Publish or Perish software on Google Scholar to obtain sophisticated citation statistics and analyses about individuals, and provides worked examples of the uses to which such statistics can be put. The author does warn about exercising caution in calculating and interpreting the results, which is wise.

            The book considers the question of the best databases to use for citation analysis. She is dismissive of ISI’s Web of Science, quoting a number of valid reasons why the service is not totally reliable, and some invalid ones. She hardly considers Web of Science’s major competitor (Scopus) at all, but is full of praise for Google Scholar, which is far more comprehensive than either of the other services. This raises a fundamental point. It is true that both Web of Science and Scopus are selective in the number of journals they take as the basis of their citation counting, but this does not matter at all as long as the order of results is the same as with a more comprehensive service, such as Google Scholar. Thus, for example, if Professor A got 100 citations and Professor B got 50 citations according to Web of Science, but Professor A got 200 citations and Professor B got 100 citations according to Google Scholar, Professor A is consistently scoring better than Professor B. The order is the important matter, not the absolute number. So there is nothing wrong with using Web of Science or Scopus for citation analysis rather than Google Scholar, assuming the services are equally reliable. Google Scholar is free (as is the author’s software), whereas Scopus and Web of Science are both expensive. So, if the services were equally reliable, then there would be good argument for using Google Scholar, with its free software provided by the author, rather than either Web of Science or Scopus (which provide citation analysis tools bundled into their services).

            So, just how reliable is Google Scholar? It relies on the automatic parsing of records, whilst both Scopus and Web of Science use human editors to check materials. Web of Science and Scopus are both highly structured databases with preset fields and consistent indexing, in contrast to Google Scholar. The author does refer from time to time to parsing errors in Google Scholar and the fact that Google Scholar covers only journals whose full text is available in open access format somewhere. She also notes that it apparently records only the first 1000 hits it finds, that it ‘occasionally’ reports duplicates, and that, unlike the competitor services, the recording of journal titles in Google Scholar is inconsistent. She also correctly notes that the lack of an ‘affiliation’ field in Google Scholar is a problem. This is all true, but overlooks the frequent duplication of records (because the same item is available in open access form from more than one repository), and the lack of transparency of selection criteria used by Google (in contrast to the complete transparency of the competing services). She also does not consider that if you do a search on ‘Fred Smith’ in Google Scholar, you will retrieve articles jointly authored by Fred Bloggs and John Smith – something that does not occur with its competitors.

            To test the reliability of Google Scholar, I carried out a search for my publications in the period 2009–2010. For 2010, Google Scholar claimed I had published 39 items, but in fact only 10 of them were by me; for 2009 it claimed I had published 78 items when in fact only 22 were by me. Furthermore, the majority of items found by Google Scholar correctly attributed to me were book reviews rather than research articles, a weakness Harzing does not mention in her book. Many of my research outputs in both years were not picked up. To make matters worse, one of the 2009 publications it claimed was mine was in fact published in 1994, and another was published in 1993, and in both cases, it was claimed the items were co-authored with other named people, when in fact they were not! Finally, of the 32 items it correctly attributed to me, five were duplicated, so the true figure was just 27 correctly attributed items. Quite simply, Google Scholar is simply not yet robust enough to be considered a reliable service.

            The author makes a number of errors in her text. She claims Google Scholar ‘covers citations in all academic journals’ though it covers citations in only those academic journals available in open access electronic format, a quite different thing. She claims that calculating citations per year is much easier with her software than with Web of Science, which is not true as Web of Science (and Scopus) have tools to give such results instantly. She also fails to mention InCites, the Thomson Reuters tailored service for in-depth citation analysis. The argument that Google Scholar provides a ‘more accurate measure of citation impact for junior scholars’ is dubious. At one point, she confuses average with median. She claims journal editors use her software to evaluate the publication record of potential editorial board members; any good editor will know this already. She claims social science journals tend to publish articles published in many countries, whereas in fact they tend to be focused on single countries; it is in the sciences where journal articles come from multiple countries. At another point, the author mixes up editorial board members with journal article referees. They are not always synonymous. She claims that Ghoshal’s book, Managing Across Borders, is not to be found in a Web of Science citation search, when it is. She claims that most conference papers and working papers ‘eventually find their way to published articles’, which is debatable. Finally, Harzing does not mention that others have worked on improving her software to clean up some of the problems that Google Scholar raises; for example, duplication. The service is called POP Clean, and is available at http://cleanpop.ifris.org. The book has some typos. ‘Thomson’ (as in Thomson Reuters, the company that owns Web of Science), ‘fare’ and ‘advise’ are mis-spelt in places, and a stray question mark appeared in one chapter title. There is also some repetition of material scattered about the book.

            Overall, can the book be recommended? If you wish to use Google Scholar, for all its faults, in bibliometric analyses, then this book is a good guide to software that can be used to help you; but bearing in mind the adage ‘garbage in, garbage out’, I would not yet use Google Scholar for such purposes.

            © 2011 Charles Oppenheim

            Author and article information

            Contributors
            Journal
            cpro20
            CPRO
            Prometheus
            Critical Studies in Innovation
            Pluto Journals
            0810-9028
            1470-1030
            June 2011
            : 29
            : 2
            : 181-183
            Affiliations
            a Market Harborough, Leicestershire , UK
            Author notes
            Article
            567849 Prometheus, Vol. 29, No. 2, June 2011, 181–183
            10.1080/08109028.2011.567849
            d2ea7ce9-f18a-4f3b-bb57-00d22fefe3e8
            Copyright Taylor & Francis Group, LLC

            All content is freely available without charge to users or their institutions. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles in this journal without asking prior permission of the publisher or the author. Articles published in the journal are distributed under a http://creativecommons.org/licenses/by/4.0/.

            History
            Page count
            Figures: 0, Tables: 0, References: 0, Pages: 3
            Categories
            Book Reviews

            Computer science,Arts,Social & Behavioral Sciences,Law,History,Economics

            Comments

            Comment on this article