How to interpret quality of computer science journals that charge publication fees?

Question

I usually publish my research work at conferences in the field of Computer Science, I do not have so many papers and the question that I am doing here it might sound foolish for some (or maybe a lot) of researchers in different fields. A situation has arosen when a colleague of me wants to publish a co-join work that we have done in a CS journal. For my experience conferences charge a fee for authors when their work is accepted, but I have some doubts about the fees that I found for some journals in the field:

• Some journals I have seen are free and have an open access so that everybody could read or access a research work
• Other journals charge some fees for accepted works that range from 200 USD to 600 USD, they are indexed in Scimago
• Other journals that are indexed in Scimago in quartiles Q1 and Q2 can charge like almost 2000 dollars for accepted article, I have seen those in an Elsevier table, or maybe am I wrong?

I mean according to some PhD students that I have talked, they told me that usually good journals do not charge publication fees, but why for example the aforementioned editorial does that? I mean sometimes a journal could be accused of predatory because it charges for publishing, but it is not almost the same with some respectable journals in the field that do the same?

Bottom line, should one aim to publish only on free fee for publishing journals? Or one should better look out for the quality of the articles accepted within a journal instead of the fees charged for publishing?

Answer 1

From a comment of yours, you seem to be specifically interested in the Journal of the Brazilian Computer Society.

From the website it appears that this journal is part of a set of Springer journals with an open access policy (Springer Open). This is also clearly stated in the fees and funding page:

Open access publishing is not without costs. Journal of the Brazilian Computer Society therefore levies an article-processing charge of £695/$1085/€885 for each article accepted for publication.

Among the reputable journals there are usually three different publication policies:

1. Closed-access journals. Authors don't pay a dime, or they pay only if their paper is longer than a certain number of pages. Here the publication fees are covered by those -- typically libraries -- who subscribe to the journal. Only the subscribers can access the journals' content. Actually, many publishers require institutional subscribers to buy the subscription to a large number of journals, with very high subscription costs (can be in the range of 30 k$/yr).
2. Journal with mixed access. Here the authors choose what type of access they want. If they opt for closed access, they don't have to pay, but their paper will be accessible only by subscribers; if, instead, they opt for open access, their paper will be accessible by anybody but they have to pay a fee, which can be as high as 2-3 k$. The fees should, in principle, pay the article-processing costs (something that is questioned by many people).
3. Open-access journals. Here authors have always to pay a fee which covers the article-processing costs. The articles are however free to access to anybody.

I would highlight two points from the fees and funding page that may be of interest to you. First:

If the corresponding author's institution is a Member, the cost of the article-processing charge is covered by the membership, and no further charge is payable.

And the final paragraph:

It is the wish of the JBCS team that all quality articles will be published in the journal independent of the funding capacity of the authors. Thus, if the authors are unable to pay the APC charge, we recommend that they contact the editors. The JBCS team will provide support to find alternative ways of funding. In particular, a grant from the Brazilian Internet Steering Committee helps sponsoring the publication of many JBCS articles.

Answer 2

This is less a response to the main question than to one of the answers, which says, "[At full or non-hybrid OA journals] authors have always to pay a fee which covers the article-processing costs."

It's not true that all OA journals charge publication fees. In fact, it's not even true that most charge publication fees. On the contrary, most (about 70%) peer-reviewed OA journals charge no fees at all, and about half of all articles published in OA journals are published in the no-fee variety.

To find no-fee OA journals in CS, go to the Directory of Open Access Journals < http://www.doaj.org/ > and look for journals in CS < https://goo.gl/imUOGt >. To learn whether or not a given journal charges a publication fee, just click through to the DOAJ record on that journal.

BTW, even when OA journals charge publication fees, the fees are seldom paid by authors. They're usually paid by the author's funder (59% of the time) or the author's employer (24%), and only 12% of the time by the author out of pocket. See the SOAP study < http://arxiv.org/abs/1101.5260 >, Table 4.