A Guide to SDSS Publication Procedures

by David Weinberg, SDSS Scientific Publications Coordinator

This is an informal guide to SDSS publication procedures. It is intended to supplement the formal description of the SDSS Scientific and Technical Publication Policy. If you have never read the publication policy, you should do so.

If you are wondering specifically about a paper that uses only on publicly released SDSS data, please read the advice of the CoCo on this topic here.

As the SDSS Scientific Publications Coordinator (SPC), it is my job to ensure that SDSS scientific papers follow the publication policy. The purpose of this guide is partly to save you having to re-read the publication policy before every paper, but mostly to highlight the most significant points and to give some additional details --- in some cases these are ``implementation customs'' that have developed subsequent to the policy itself. You should read this guide every time you are ready to submit a paper, in part to see whether anything has changed.

I'll first address journal article submissions, then conference proceedings submissions.

Steps in submitting an SDSS scientific publication to a journal

Decide on the ``analysis group'' author list. The publication policy, Section 7, defines the analysis group to be those who were heavily involved in the specific analysis described in the paper and in the writing of the paper.
Check that your manuscript includes the up-to-date version of the SDSS acknowledgment, which you should get from the policies page of www.sdss.org. Do not copy the acknowledgment from some previous paper, since it changes from time to time as new institutions join the SDSS; always get the new version.
Titles should not have "based on data from the SDSS" footnotes; the only standard acknowledgment is the one above, at the end of the paper.
As stated in Section 3 of the publication policy, "All SDSS scientific papers should reference the appropriate technical papers." You can effectively read "should" as "must" --- the purpose of this clause is partly to ensure that readers get pointers to the right technical information, but equally to ensure that those who have developed the technical infrastructure of the SDSS are appropriately recognized and rewarded by having highly cited papers. You can find a list of the required technical references here. One of my jobs as Scientific Publications Coordinator is to tell you to include any technical references that you missed, but my life will be easier if you include them all the first time. The list is presently short because many of these papers have not yet been submitted for publication, but it will grow a lot in the next few months, so check this list before every paper you submit.
Post your manuscript to the publication index. Posting the paper will automatically generate an e-mail to sdss-general announcing the posting and create a link from the index page to the file that you specify, which people can download to get the manuscript. (Note that this procedure supersedes the description in 8.1.ii of the publication policy, which says that the contact author forwards the paper to the SPC.)
A manuscript is eligible for submission to the journal 3 weeks after it has been posted to the publication index.
Very important:Posting a manuscript to the publication index implies that the analysis group authors all consider the paper ready for submission to the journal. Do not post rough drafts of papers in order to "start the clock" --- it is not fair to other members of the collaboration to ask them to decide whether to request co-authorship on a paper that isn't complete, and it is a waste of everyone's time to send in comments and corrections on a paper that hasn't been polished by the primary authors. If all that remains is a specific change of very limited scope, you may post the paper with a comment about what remains --- e.g., "we are calculating results for one additional model, which will be added to Figure 5."
Of course it can be very valuable to get feedback from SDSS collaborators on initial or intermediate drafts of papers; the best way to do this is through the mailing list of the most relevant working group. Most SDSS manuscripts do get revised and improved, often substantially, after they are posted to the publication index, because the authors get useful comments from a broader spectrum of the collaboration. That should continue to be the case. However, the criterion for posting to the publication index should be that the analysis group authors consider the manuscript complete and ready to go. If I judge that a posted manuscript falls far short of this standard, I will inform the authors that the "3-week clock" will not start until a new version of the manuscript is posted.
Author lists should not conclude "for the SDSS Collaboration" or "and the SDSS Collaboration," nor should the begin "The SDSS Collaboration:". The only exceptions are Key Project papers (publication policy, Section 6) and "Conference proceedings which are primarily overviews of the SDSS project status or scientific results" (publication policy, Section 7.3). These should have "for the SDSS Collaboration" at the end of the author list.
Other members of the collaboration can request co-authorship on the paper via the publication index web page and/or by e-mailing the contact author. The publication index does not automatically generate an e-mail to the contact author; it is that person's responsibility to check all authorship requests on the index before submission and incorporate them. If you request authorship after the paper has been submitted, you should be sure to send e-mail to the contact author, not rely on them checking the index after the paper has been submitted.
If there are people who you think would be particularly appropriate co-authors but are not in the analysis group list, you should invite them to request co-authorship. In the event of disagreements about co-authorship, contact the Scientific Publications Coordinator (me).
Technical papers, those whose purpose is to "describe the SDSS instrumentation, calibration, software, strategy, and targeting algorithms," are not open to general co-authorship requests. The publication index listing indicates whether a paper is a Scientific Paper or a Technical Paper. If you believe you have been incorrectly omitted from the author list of a Technical Paper, communicate directly with the contact author, and if necessary with the Technical Publications Coordinator (currently Jill Knapp).
This is new. Although it is not required by the publication policy, I would like to strongly encourage all authors to repost their papers one week prior to the intended submission date. "Repost" means: update the link from the publication index so that it points to an up-to-date version of the revised manuscript (with the current author list), briefly describe changes that have been made in the comments section on the publication index (or say that there have been no changes), send a message to sdss-general saying that the paper will be submitted in a week and briefly describing changes that have been made.
The purpose of this new procedure is to let people know whether and how manuscripts have changed and to let them know that the paper is about to be submitted (some papers are revised extensively and are therefore submitted long after they are posted to the publication index).
When you submit the paper, update the comments on the index page to indicate that the paper has been submitted, though you do not need to send an additional sdss-general e-mail.
There is no general policy on whether papers should be posted to astro-ph upon submission or after acceptance (or at all, for that matter). This decision should be made by the analysis group authors.

Steps in submitting an SDSS scientific paper to a conference proceedings

The procedure for submitting a conference paper based on SDSS data is the same as that described for journal articles above, with the following exceptions:

Because of their length constraints, conference papers may be exempted from the authorship policy (i.e., from allowing co-authorship requests). However, quoting from Section 7 of the publication policy: "Any conference paper that does not follow this authorship policy may only include quantitative results (tables and figures, for example) based on papers that have been published or submitted for publication, and it must reference those papers as the primary source for the results. The rationale for this requirement is that conference papers authored by a single author or a small group should not become the primary reference for results that are properly products of the SDSS Collaboration."
If you post a conference paper under this clause, for which you do not intend to accept co-authorship requests, you should note this in the comments section of the index posting and in the sdss-general e-mail message. If you are unsure whether you paper qualifies for this treatment, you can contact me before posting. Also contact me if you think that someone has erroneously posted a paper under this clause. Based on the rationale given above, I usually make a judgment in such cases by asking myself whether the conference paper is likely to be referenced as the source of an SDSS scientific result --- if so, it should follow the usual authorship policy.
Although the requirements about technical references and acknowledgments apply to conference papers as well as journal articles, I am in practice somewhat more lenient in enforcing them for conference papers with tight length constraints (2 or 4 pages).
New: The CoCo has agreed to the following policy on a trial basis. The abstract and initial author list for a conference paper may be posted in advance of the draft manuscript, and this will start the three week "clock." The full draft manuscript must be available on the publication index for ten days.
This modified policy is intended to give authors a bit more breathing space to meet conference proceedings submission deadlines. In order for it to work, it is essential that authors not trim its corners.
Note that the abstract must still be posted three weeks ahead of submission. You should send an additional sdss-general message when the manuscript itself is available, noting that submission is intended in ten days (or three weeks from the abstract posting, whichever is later).

Go to David Weinberg's Home Page
Updated: 2001, June 4[dhw]