Towards a Computer Science Publication Model in the Social Sciences

In the next series of posts I argue how the publication system of my research area – Internet research with a background in media and communication as well as management – could profit from becoming more like the publication system in computer science, in a first step, and then physics, in a second step. However, as the social sciences of the Internet (which I consider myself being part of) are a very broad and interdisciplinary research area, I need to specify that the following explications only apply to the domains that I know: sociology of the Internet, political science of the Internet, (social) psychology of the Internet, and – partly –philosophy of the Internet as well as information systems. Thus, I refer to the major social science Internet outlets, such as Journal of Computer-Mediated Communication, New Media & Society, Information, Communication & Society, Computers in Human Behavior, Social Science Computer Review, Journal of Broadcasting & Electronic Media, Journal of Information Technology & Politics etc. I also include the open access journals, such as First Monday and International Journal of Communication, although in some ways they are different/better. But how is the publication system in social science Internet research (and probably also in other social science fields) flawed?

  • It takes too much time from submission to publication
  • The review system isn’t efficient and often quite arbitrary
  • The system places too much weight on the publication brand (journal reputation, impact factor, rankings, h-index etc.) and not enough weight on the content, i.e., the research results and the novelty and ingenuity of findings

Ad Point 1): Research productivity and progress is severely inhibited by the long time it takes from submission to publication. Often, research in this domain substantially lags current social developments, so that the data are frequently several years old when a publication goes online. This is very problematic, since the Internet is a rapidly evolving field of study, where it is especially important to stay up-to-date (so as to inform current debates in policy and media, for example). Normally, it takes at least several months and in some cases – with all revisions – even years from submission to publication. Of course, good research takes time, but this is time that was already invested at the time of submission and shouldn’t be unnecessarily extended once the results are written up.

The long process from submission to publication includes several problematic sub-points: stupid formal requirements, retaining journal issues in the digital age, no early view and long time from acceptance to publication. Many journals require their authors to fulfill completely stupid and arbitrary formal requirements, such as exotic citation formats, the use of endnotes (instead of footnotes or no such things at all) and very special table formatting. Every time an author has to resubmit a rejected article to another journal, it takes forever to update the formalities. It would be a great first step to reach agreements among the publishers and journals to have one – and only one – set of formal requirements. Better so it would be to do away with such requirements completely and let the authors write as they want – publishing the unformatted raw text (much like a post-print but maybe with subsequent page numeration) as the final version. This leads to the next issue: the issue issue. Journal issues slow down the publication process immensely and should be disposed of entirely. Once an article is published online (nowadays referred to as “early view” or “early cite”) and has a DOI, this should be the final, citable version. Journals without early view should introduce it as soon as possible. What’s the point in holding articles back for several months, when they could be published right away? Finally, and this is somehow similar to the last sub-point,  journals should do their best to make the time spans between acceptance, copyright agreements, sending out the formatted version to the authors, implementation of the proofs and (early view) publication as small as possible. The whole process from acceptance to publication should take a couple of weeks, but not entire months.

In the computer science publication model the submission-publication cycle is shorter, because the conferences provide clear deadlines in terms of reviewing and presenting. The standardized format and the clear communication of relevant information, such as acceptance rates, the reviewing process, the conference fees, the conference structure and the nature of the proceedings as well as their accessibility allows for more predictability for the authors. Of course, the computer science publication system is far from being homogenous and perfect – and there are many criticisms within the field (http://yann.lecun.com/ex/pamphlets/publishing-models.html) – but I think it would be a first step of advancement for the social sciences to go into that direction.

More on the second and third points soon.  

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s