Uncitedness and self-citations: assumption of guilt?

[eackerma]

I must say that, my reputation for crankiness notwithstanding ;), I found
Quentin Burrell's notion that there may be a linkage
  between the uncitedness of a published paper(s) and the amount of
self-citation quite interesting. I certainly have nothing
  against speculation in general, and think that this notion of a
uncitedness-self-citedness linkage certainly bears further
  empirical investigation. I would also hasten to add that I am not suggesting
that self-citations are *never* used to avoid
  uncitedness or low-citedness. After all, researchers are people, and exhibit
the same range of behaviors both desireable and
  undesireable as any other human beings. However...

  ...at the risk of becoming the SIGMETRIC list's in-house curmudgeon (an
honor which may be too late to avoid! 8-D), I
  must disagree that one can assume that "one of the possible *prime* reasons
for self-citations is to avoid uncitedeness, or at
  least low-citedness" (empahsis mine). Even if self-citations "are the sole =
  citations or at least predominate a paper's citations" as Burrell points
out, how does that fact alone justify the treatment of all
  self-citations with suspicion and automatically discounted value? What of
the researcher building on a previous body of
  reputable research who is also working in a field not currently fashionable
or in vogue? This researcher is likely to have few
  options but to draw from their own work. Should their self-citations be
*automatically* discounted?

  In some repects this difference of opinion between those that automatically
view self-citations with suspicion and those that
  don't is analogous to the classical notion of how one view's a glass
partially filled with water. Is the glass half-full or
  half-empty? Are self-citations automatically assumed to be primarly
self-serving entities for unnecessarily inflating citedness
  unless proven otherwise (the glass of water is half-empty), or are they
assumed to serve the same legitimate purposes as
  other citations until proven otherwise (the glass of water is half-full)?

  However, once having established that in a particular case (or group of
cases) that the self-citations encountered are indeed
  primarily self-serving, then it certainly would be legitimate for devising
some method for discounting or eliminating them from
  a study's data. In which case, I think Burrell might be on to something:

  >One simple way to build this in would be by discounting =
  >self-citations according to the degree to which they dominate the =
  >citations as follows:
  >
  >If a paper receives N citations, of which a proportion p are =
  >self-citations, then its discounted citation score (DCS) is (1-p)N + =
  >(1-p)pN, so that each of the (1-p)N non-self-citations is given full =
  >weight but each of the pN self-citations is discounted by a factor =
  >(1-p).
  >In the examples above, 20 non-self-citations + 80 self-citations gives, =
  >since p=3D0.8 here, a DCS of 0.2x100 + 0.2x0.8x100 =3D 20 + 16 =3D 36.
  >Similarly in the converse mix, 80 non-self-citations + 20 self-citations =
  >gives (p=3D0.2) a DCS of 0.8x100 + 0.8x 0.2x100 =3D 80 + 16 =3D 96.
  > A 50-50 mix would lead to a DCS of 75, all 100 non-self-citations a DCS =
  >of 100 and all 100 self-citations a DCS of 0.
  >

  I will also concede that the time and effort required to check the
"legitimacy" of each self-citation is probably prohibitive for
  most studies, especially those dealing with hugh datasets. But the solution
does not lie, in my humble opinion, in the
  automatic assumption of guilt implicit in research methodologies that
require the discounting or omission of all self-citations.

  Perhaps those who are researching citation motivation may be able to shed
more light on this situation.

  In any case, my thanks to Quentin Burrell for getting the discussion started
again!

  Eric Ackermann
  School of Information Sciences
  University of Tennessee-Knoxville

Eric Ackermann
eackerma at utk.edu

Vernon Hughes' law of low-level statistics, "named for an atomic
physicist of note at Yale. His law states that despite the fact
that a three-sigma effect [i.e., confidence interval] appears to
have a 99.73 percent chance of being right, it will be wrong half
the time. This is real life."

-from Gary Taubes' book "Bad Science, The Short Life and Weird
Times of Cold Fusion" (NY: Random House,1993: p. 84).