Citations and arXiv.org

A couple of years ago, a paper came out saying that astronomy papers posted on astro-ph get about twice as many citations as those which do not. Plenty of people think that if you are going to post a paper to arXiv, then you should time it so that it, for example, shows up near the top of the daily astro-ph listings. The second paper on today's astro-ph list takes this a step further, showing that the papers appearing near the "top" of the daily astro-ph listings receive essentially twice as many citations as those further down on the list. The interesting question here—one which it's fairly difficult to disentangle—is whether this difference is really purely due to a positional bias (people glance at the first few papers on astro-ph in the morning and then stop, and thus are more likely to remember and cite those first few papers), or if it is a selection effect (people who care enough to make sure their paper is near the top of the list might simply have better science to share, or be better at self-promotion in other contexts, like conferences).