How should I prepare e-prints (pre-prints and post-prints) of already published papers?

Question

I am about to embark on the publication of pre-prints or post-prints of about a dozen papers that have already been published in peer-reviewed journals. I plan to use the SHERPA/RoMEO database to decide whether to publish a pre-print or post-print, and when it is appropriate. I have backed up the published papers at both the pre-peer review and post-peer review stages.

These documents will be lodged primarily with my institutions repository, but also on a personal website and perhaps academia.edu. Due to my field of research, arvix is not appropriate.

I am aware that I may need to make some modifications to the manuscripts before lodging them as a pre-post print. For example, it seems appropriate to link to the canonical, published version of the paper on the cover sheet. But I am unaware of what other changes I may need to make to the manuscript. A checklist of modifications to make would ease this process and help me to avoid missing things.

What steps should I go through to prepare a manuscript for publication as a pre-print or post-print?

UPDATE: I've made an example post-print here. Are there any concrete improvements I could make on this, or important things I'm missing?

Answer 1

Based on my experience, this is my checklist

1. See the relevant journal entry on http://www.sherpa.ac.uk/romeo/ (or http://rchive.it). I always read of the publisher's copyright transfer agreement, as well. You never know.
2. Decide whether I can self-archive only the preprint or the postprint or both. The second one is usually defined as the one after peer-review but before the publisher's typesetting and proofreading (if any is performed). Of course, self-archiving the postprint is more desirable, as its content is almost the same one as the published paper.
3. Take the appropriate version of the paper from my files. That is, I take the .doc, .tex or whatever the source file is. I either create a cover page or a footnote after the title/last author name. In the footnote, I put the following text:

This is an author generated preprint (or postprint) of the article:

//full citation here

Copyright XXXX The Publisher (if required).

The final publication is available on http://dx.doi.org/DOI-HERE (or the publisher's digital library entry)

Never had any problem doing like that. Just make sure to clearly state what the publisher asks you to state.

Your example looks fine to me. I would rewrite the "This is an Accepted Manuscript of an article published" to clearly state either preprint (before peer-review) or postprint (after peer review).

As stated in another answer, never self-archive the publisher's PDF anywhere (unless it is an open access journal, but you do not really need to self-archive then). I am re-writing it because authors typically do not understand this.

Answer 2

One thing I suggest is do not put up the paper in submitted-manuscript form. You are free to change from double-spaced (if that is the case) back to single spaced (or as appropriate); you can put figures and tables where they belong; you can use a nice typeface; you can even change the referencing style to a one you like better.

Use the version that incorporates changes suggested by the referee if at all possible. My guess is you didn't sign away rights until after you made said changes and sent it back to the journal -- or this may have happened, and it may be that the original submission came with agreeing to terms that include all rights in various versions of the paper in the case of acceptance - I hate these types of journals. In any case, knowing your field would help regarding advice here.

This website can help you find out if you are allowed the post-referee-comments version: http://rchive.it/ (it uses Sherpa/Romeo data), but the ultimate answer will come from the journal website.

This will make your paper more pleasant to read, and help dispel the myth that preprints are ugly.