Normalizing Citation Counts

Higher total citation counts are obviously biased toward more senior faculty who have had more time to publish and be cited. To account for this a variety of metrics have been developed to adjust for career length, output, and other publication level factors. These include the Hirsch-Index (h-index), e-index, g-index, hl-annual index, hl-norm index, and age-weighted citation rates (AWCR, AWCRpA, and AW-Index). Normalizing can involve increasing amounts of information including individual publication ages and citation counts (see: https://harzing.com/pophelp/metrics.htm for more discussion). An example includes the m-quotient which is the h-index divided by the number of years since graduation. H-indices grow much more slowly than do citation counts and are positively correlated with career length while also indicating author productivity. In addition to career length bias, there is also the effect of multiple co-authors. More co-authors means greater potential for self-citation and greater visibility generated by multiple authors on a particular publication. This is usually a concern between disciplines which have different co-authoring patterns and citation levels (see Ioannidis, Boyack, and Wouters, 2016).

The following tables show how normalized citation metrics can produce different rankings. Tables 1 through 4 show planning faculty by total citation counts and three normalized indices. As shown, the greatest variably occurs when citation totals are normalized by years since degree (squared), years since degree, and the m-quotient, with standard deviations of 256.4, 128.8, and 89.7 respectively. The columns “Diff.Cites” show how the particular faculty ranking varies from the Total Citation ranking (Table 1). For example, Ralph Buehler (Virginia Tech) and Jason Cao (University of Minnesota) each move up 38 places in the citation ranking controlling for career length (see Table 2). Indices involving publication level data, such as AWCR and co-author shares of citations, can only be performed for approximately half of all planning faculty and could not be included. These analyses were performed for faculty with Google Scholar Citation profiles and also showed significant differences in ranking.

Table 1: Planning Faculty, Total Citation Counts

Table 2: Planning Faculty, Citation Counts Normalized by Years Since Terminal Degree