Page 208 - Becoming Metric Wise

198 Becoming Metric-Wise

Now, the 3-year impact factor, JIF 3 , will be larger than JIF 2 if
JIF , JIF
2 3
cð1Þ 1 cð2Þ cð1Þ 1 cð2Þ 1 cð3Þ
3 ,
pð1Þ 1 pð2Þ pð1Þ 1 pð2Þ 1 pð3Þ

3cð1Þðpð1Þ 1 pð2ÞÞ 1 cð1Þpð3Þ 1 cð2Þððpð1Þ 1 pð2ÞÞ 1 cð2Þpð3Þ
, ðpð1Þ 1 pð2ÞÞcð1Þ 1 ðpð1Þ 1 pð2ÞÞcð2Þ 1 ðpð1Þ 1 pð2ÞÞcð3Þ
3ðcð1Þ 1 cð2ÞÞpð3Þ , ðpð1Þ 1 pð2ÞÞcð3Þ
cð1Þ 1 cð2Þ pð1Þ 1 pð2Þ
3 ,
cð3Þ pð3Þ

Dividing both sides of the last inequality by 2 yields equation ( )below:

μ ð2Þ μ ð2Þ
p
c
, ðTÞ
cð3Þ pð3Þ
The assumptions p(3) # μ p (2) and c(3) . μ c (2) entail that
μ ð2Þ μ ð2Þ
p
c
, 1 ,
cð3Þ pð3Þ
so that inequality ( ), and hence the inequality JIF 3 . JIF 2 is always satis-

fied. This shows that JIF 3 , the impact factor calculated over a 3-year
period is, in the basic model, always larger than JIF 2 , the “standard” or
Garfield-Sher impact factor.

6.21 ADDITIONAL INFORMATION

Additional information about impact factors and their use in citation
analysis can be found in: Todorov and Gla ¨nzel (1988), Stegmann (1999),
Wilson (1999), Egghe and Rousseau (2002), Gla ¨nzel and Moed (2002),
Moed (2005a), Nicolaisen (2008) and Bar-Ilan (2008). Rousseau (2002b)
discusses aspects of journal evaluation and concludes that a whole battery
of indicators is necessary, see also (Haustein, 2012). In the same vein,
Bar-Ilan (2012) proposes journal report cards.
Regional citation indices, such as the already mentioned Chinese ones
and SciELO often provide a derived database of local impact indicators
( Jin et al., 2002). Also Publish or Perish (Google Scholar) provides means
to search per journal.