I am assuming that there is a correlation between a paper's publication month and the number of citations it attracts, based on the following premises (that may or may not be wrong, that's also open for discussion):
- When people do literature research they tend to focus more on papers that have already attracted many citations, rather than uncited papers.
- As a corollary of 1., people will be more willing to read and cite an uncited paper if it has been recently published (compared to an uncited older paper), assuming that it may have attracted few (or no) citations because there has not been enough time to do so since publication.
- When looking at how recently a paper has been published -and therefore establishing its "age"-, people only look at the year, rather than the month, of publication.
As a consequence of 1., 2. and 3., a paper that has been published towards the end of the year becomes "old" (and less likely to attract citations, c.f. 2.) faster than a paper that has been published towards the beginning of the year. For instance, consider a paper published in December 2014, which becomes "one year old" (according to 3.) already in January 2015, after one month of existence. A paper published in January 2014 will stay "newborn" until January 2015, therefore it enjoys 11 extra months of "freshness" compared to the other paper, and will therefore be more affective at attracting citations.
My questions are: is my argumentation above flawed (and how)? And, if not, is there any data supporting a correlation between publication month and number of citations attracted by a paper?
Note that an effective data analysis would likely need to exclude very highly cited papers that might introduce a bias.