Sunday 8 June 2014


Transformations in Scholarly Communication in the Digital World

Universite de Montreal
Ecole de bibliotheconomie et des sciences de l'information




Overview: Digital technologies — easy to update, reuse, access and transmit and require little space — have changed how researchers produce and disseminate scientific knowledge. Based on quantitative studies in the sociology of science, this talk will discuss these transformations, higlighting three aspects: the increase of scientific collaboration, the diversification of publication venues, and the use of social media.

READINGS:
    Wallace, M. L., Lariviere, V., & Gingras, Y. (2012). A small world of citations? The influence of collaboration networks on citation practicesPloS one7(3), e33339.
    Lariviere, V., Gingras, Y., & Archambault, E. (2006). 
Canadian collaboration networks: A comparative analysis of the natural sciences, social sciences and the humanitiesScientometrics,68(3), 519-533. 
    Bollen, J., Van de Sompel, H., Hagberg, A., & Chute, R. (2009). 
A principal component analysis of 39 scientific impact measuresPloS one4(6), e6022.
 http://www.chss.uqam.ca/Portals/0/docs/Canadian_Networks_Final.pdf
 http://ella.slis.indiana.edu/~sugimoto/preprints/OnTheRelationship.pdf
    http://arxiv.org/abs/1304.6460
    http://arxiv.org/abs/1205.4328
    http://arxiv.org/abs/1308.1838
    http://arxiv.org/abs/0809.5250

26 comments:

  1. I'm curious to know what accounts for the spike in humanities citations between 1980 and 1990.

    ReplyDelete
  2. M. Vincent Larivière présentait la proportion d'articles qui ont été cités au moins une fois après deux ans et cinq de publication.

    Comment sait-on le nombre de fois q'un article quelconque a été cité?

    ReplyDelete
    Replies
    1. Web of Science, Google Scholar, SCOPUS...

      Delete
  3. Does all these finding support the hypothesis that science becomes an increasingly collaborative activity?

    About citations in the humanities, perhaps it is possible to differentiate cumulative works from non-cumulative works. It seems the finding regarding the humanities are somehow biased so further differentiation of categories is needed in order to expose the actual trends.

    ReplyDelete
    Replies
    1. The data indeed seems to support the hypothesis that science is increasingly collaborative. However, some of the trends could be due to "looser" authorship attribution practices...

      Interesting hypothesis about the humanities. The challenge, tough, would be to find cumulative branches in the humanities. Or at least, cumulative in the short term in a manner similar to the natural and medical sciences.

      Delete
  4. Very interesting talk. I particularly liked when Dr. Lariviere pointed out that tweets and citations are not correlated. I do believe citations have some value as an indicator of the importance and soundness of an article, but tweets may not. Perhaps, tweets and facebook are effecting scientific literacy among non-scientists.

    ReplyDelete
  5. Fascinating talk. I am quite pleased to learn that scholarly publication in journals in becoming more and more decentralized, heading as it is towards new, smaller journals and open source journals. I am worried, however, to see that the trend is not translating very much into the humanities. I wonder why this is the case. I also wonder whether the fact that humanities research is (typically) less data-driven can account for a part of the observed tendency for this slower-moving trend.
    How can we get the humanities to start moving in the direction of the more general trend? Is this something particular about the humanities, or does it reflect a sluggish shift to new kinds of research workflows and data sharing, or something special about the humanities culture, or something else?

    ReplyDelete
    Replies
    1. I'm not sure we can simply apply the same model. No two disciplines has the same culture of citation, for all sort of reasons. I think it's for the disciplines in humanities to build there own model—each disciplin on their side.

      Maybe what we could do is open the humanities to new possibilities—give them access to new modes of publication, and break the old models. This probably means money.

      Delete
    2. The publication cultures are tightly linked with what researchers value as "important" research products. In the humanities, the book is still the criteria by which reseachers are judged (and are given tenure); these long term traditions are not easy to change. The use of the book, in those disciplines, is also much more adapted to the knowledge that is produced: writing the history of events, etc. is typically much tougher to condense into a short article (let alone a tweet!!!!) than in a book. So, indeed here, we need to take into account the specificities of the disciplines (and of their objects) when thinking about publication models.

      Delete
    3. Thank you for the reply!
      I agree that it has a lot to do with academic culture and adequacy of format to message. I'm still intrigued by Louis' proposal above, to provide the humanities with a greater diversity of publishing options. After all, many sciences that today function almost exclusively through articles, such as physics, were book-based for a very significant amount of time.
      But the fact that tenure is tightly related to book publishing is still a significant obstacle to changing the culture.

      Delete
  6. Something that I think might be interesting to study is the creation of communities within and across different disciplines. By communities I mean groups of scientists that highly cites each others or groups of researchers that publishes often together.

    I'm not sure if anyone already does it, or what kind of analysis could be performed on this data, but I guess that an author that has been cited 100 times by 10 researchers have a smaller impact than if these citations we spread across 50 researchers. It could also provide a different view of collaboration among scientists: instead of number of co-authors by paper, the evolution of the average size of these communities could be an interesting perspective.

    ReplyDelete
    Replies
    1. It's been done. Carolina Ferrer does some of that in UQAM.

      Delete
    2. Interesting idea -- in other words, we would look at numbers of citing authors rather than looking at citing papers. There is related research, but not much: http://onlinelibrary.wiley.com/doi/10.1002/1097-4571(2000)9999:9999%3C::AID-ASI1542%3E3.0.CO;2-T/abstract

      Delete
  7. It was very interesting. We consider the twitter recently for many uses. That's right. I do not know if the impact is great however the impact is big. Thanks VINCENT LARIVIERE.

    ReplyDelete
  8. Vincent presented a lot of informative graphs. I am interested in learning more about why these trends are occurring and what we can achieve with this knowledge. The research presented was very observational and may have increasing impact when applied.

    ReplyDelete
  9. L’explication concernant le fait que les auteurs qui ont des articles très intéressant préfèrent publier plus dans les journaux spécialisés dans domaine de recherche au lieu des journaux très connus (1%), me parait un peu simpliste pour expliquer le déclin de l’engouement pour les grands journaux de publication. est-il possible de connaitre d’autres facteurs qui peuvent influencer positivement ou négativement) le choix des auteur qui ont des résultats de recherche exceptionnelles d’aller ou pas vers une publication reconnue !!!

    ReplyDelete
  10. Thank you for a great talk! If the number of journals have increased exponentially, so has the number of papers. Has the peer-review system changed as well? Do you have any insights into "quality" of the peer-review system now compared to before?
    What do you think a future of the peer-system is going to be?

    ReplyDelete
    Replies
    1. Those are very interesting questions -- Unfortunately, we know little on the evolution of the peer-review system. It seems to me that it did not change much, except the fact that the burden on the shoulders of reviewers is increasingly high. The publication culture, in which researchers are trying to publish as much as possible, is putting a very strong burden on the shoulders of reviewers -- who are actually authors as well -- which could be linked with an increase in "sloppy reviews" and, thus, sloppy research. Peer review is necessary, but despite the fact that there is a one size fits all defition of peer review in papers, I think that different disciplines need different types of peer review. Physicists have been living with arXiv for decades, reading and citing papers that have not been peer reviewed (yet). It is much more difficult to do the same in medical sciences because of its implication for human health.

      Delete
  11. The explanation for the fact that the authors with the very interesting articles prefer to publish in more specialized research area instead of the well-known journal (1%), seems a little simplistic to explain the decline in popularity for major journals.
    Is it possible to know other factors that can influence positively or negatively the choice of those author with outstanding research to not go to a recognized journals !

    ReplyDelete
    Replies
    1. The decline in top journals in mainly due to the creation of new journals (general journals and specialized journals), and to the creation of new subfields which created their own journals. So the main reason for the diversification of top articles' publication venues is actually that there are more journals in which to publish. Hence, historically top journals can only lose market shares.

      We might also add that these top journals are very picky in the articles that they send to reviewers and, ultimately, accept. And some authors have the feeling that the criteria for accepting manuscripts are not always entirely scientific and that these journals will often publish research that is more "trendy". So researchers from these disciplines might get their papers rejected by these general elite journals.

      Delete
  12. Je pense qu’il faudrait rajouter à l’histoire de la communication de l’érudition le fait qu’il y a de moins en moins de collaboration entre les différentes disciplines. Bien que cette école d’été soit multidisciplinaire, il me semble que ce soit le cas. Y a-t-il des données qui confirmeraient cela?
    Par ailleurs, pouvons-nous soutenir l’idée selon laquelle ce phénomène serait dû à la spécialisation, parfois même la sur-spécialisation des chercheurs? Je pense en particulier au siècle des lumières où les auteurs de l’encyclopédie, contrairement à aujourd’hui, n’étaient pas spécialisés, mais touchaient à toutes les disciplines à la fois.

    ReplyDelete
    Replies
    1. Data on interdisciplinary actually shows that it has been increasing over the last decades: http://books.google.ca/books?hl=fr&lr=&id=BVuaAwAAQBAJ&oi=fnd&pg=PA187&ots=tRq4VALfPP&sig=AXKBn2J2vJvFyLAoOXAcZqhfgug

      Actually, science became more and more specialized after the second world war until the seventies, and then after this phase of increased specialization started a phase of interdisciplinarity with these "new" specialities increasingly talking with each other.

      Delete
  13. Dear Vincent, Thank you very much for this great presentation ! I’m particularly interested about the way that scientific knowledge is simplified and/or transmitted to non scientists/experts. Do you have advices to study the impact and coverage of Wikipedia on and from scholarly communications? Do you think this kind of centralization and diffusion of knowledge will modify the dynamic of scientific research?

    ReplyDelete
    Replies
    1. Réponse de Vincent :

      Le Web permet effectivement aux connaissances scientifiques de percoler davantage dans la sphère publique. Je n'ai pas de conseils en particulier, mais il me semble qu'il pourrait être très intéressant de regarder la littérature scientifique citée par les articles de wikipedia. Est-ce qu'ils s'agit de documents qui font également autorité dans la communauté scientifique (et sont très cités) ou bien si ces documents ne sont pas nécessairement plus importants que les autres.

      Pour l'instant, je ne crois pas que Wikipédia change significativement la dynamique de la recherche.

      Delete
  14. This comment has been removed by the author.

    ReplyDelete
  15. I wonder if it would be possible to "engineer" a highly successful scholarly paper by leveraging the trends described in this talk (more authors, more institutions, medical or natural science subject, etc.). You could publish a paper entitled, "This paper will likely be highly cited."

    ReplyDelete