Page 130 - Becoming Metric Wise

121
Publication and Citation Analysis

Table 5.2 Data for the author credit problem
Art1 Art2 Art3 Art4 Art 5
A-B-C-F A-C-B-D C-E D C-D

Table 5.3 Solution of the problem
Credit Rank Credit Rank Credit (fractional) Rank
(first (normal
author counts)
count)

A 2 1 2 3 (1/4) 1 (1/4) 5 0.5 3
B 0 4 2 3 (1/4) 1 (1/4) 5 0.5 3
C 2 1 4 1 (1/4)1 (1/4)1 (1/2)1 2
(1/2)5 1.5
D 1 3 3 2 (1/4)1 11 (1/2)5 1.75 1
E 0 4 1 5 1/2 3
F 0 4 1 5 1/4 6

Source: Data in Table 5.2. One could also give an average rank, but that is not essential here. In that
case first author count leads to three authors at rank 5 (instead of rank 4).

author counts, normal counts and complete-normalized (fractional)
counts. Authors are given as in the by-line of their articles.
The solution is given in Table 5.3.
Counting contributions of universities or countries is even more com-
plicated. One method consists of giving an equal credit for every address
that occurs in the by-line, irrespective of the number of collaborators
with that address.
An example
Assume that an article has three authors: one from country C 1 and
two from country C 2 , belonging to the same department. The author
from country C 1 is first author, while one of the other authors is cor-
responding author.
Applying First Author Count gives one credit to country C 1 , and no
credits to country C 2 .
Applying major contribution count gives one credit to country C 1 and
one credit to country C 2 ; or alternatively (when normalizing), half a
credit to each country.
Applying Normal Counts gives country C 1 one credit, but now country
C 2 receives two credits.
Applying fractional counting (we illustrate the two main cases).