[ Previous ] [ Home ] [ Next ]
Written by: Ian Hickman
What Came First...
We know that there is a correlation between downloads and citations and that highly cited papers get downloaded more. What we do not know is whether the papers are being downloaded because of the citations to them, or whether papers are being cited because they have been downloaded. Are papers cited then downloaded or downloaded then cited?
To answer the question of what is driving what, the time difference between when a paper is downloaded and when another paper that cites it is deposited needs to be calculated. If this time difference is negative the download was before the citation else the citation was before the download. As there is no way of knowing if the same user downloaded and cited a paper, all the citations to a paper and downloads of a paper need to be considered. Taking the time of all the downloads of a paper minus the time of deposit of every paper that cited it will produce the graph below.
Splitting the paper set into high, medium and low impact papers produces the following graphs.
[ Previous ] [ Top ] [ Home ] [ Next ]