Query about journal (not author) self-citation rates

[Stevan Harnad]

On Tue, 25 Mar 2003, Small, Henry (Institute for Scientific Information)
wrote:

> Not aware of any systematic studies of these issues, although there are tons
> of journal citation studies based on JCR.  My guess is that authors alter
> their citation patterns for specific journals because they think readers
> will be more familiar with the prior literature in that journal and they a)
> don't want to be seen as not citing something they should cite, or b) want
> to cite what's relevant to readers. i.e. it's author not editor driven.

Many thanks for your reply. I guess the first step is to confirm (1)
*whether* there are disproportionately elevated same-journal citations in
some journals, relative to other (comparable) journals and then, if so,
(2) to try to infer what their causes might be, from their correlates. It
should be possible, for example, to determine from cross-citation patterns
(i.e., citations to other journals and vice versa) how "inbred" a journal
usership is likely to be. If it is indeed inbred, then that supports
the inference that authors (rightly) judge that their users are likely
to rely mainly on that journal, and that they are unlikely to be users
of other journals. If not, then we would have to look elsewhere for an
explanation of disproportionate same-journal citation (if there is any).

As a first pass, the following approximate (but only suggestive, because
the underlying distributions don't meet its assumptions) analysis might
work (for the JCR journals):

As a measure of "degree of egocentricity" relative to overall
in/out/self citation patterns I (who am not a statistician!)
would at first be inclined to look at the mean and standard error
for the ratio:  S = self/(in + out)
                T = in/(in + out)
and then for:   1/(T-S) as a rough measure of egocentricity.

(But statisticians will no doubt have better suggestions!)

Cheers, Stevan