← Back to introduction

adapted pure h-index (fractional credit)

The adapted pure h-index (Chai et al. 2008) is very similar to the pure h-index (fractional credit), except that it estimates its own core rather than relying on the standard h-index core. For a given publication, if one wishes to assign all authors equal credit, or if one does not have information about authorship order, one can calculate an effective citation count as the number of citations divided by the square-root of the number of authors,

$$C^{*}_i = \frac{C_i}{\sqrt{A_i}}.$$

Publications are ranked according to these new citation counts and the h-equivalent value, he, is found as the largest rank for which the rank is less than the number of equivalent citations, or

$$h_e = \underset{i}{\max}\left(i \leq C^{*}_i\right).$$The adapted pure h-index is calculated by interpolating between this value and the next largest, as

$$h_{ap.\text{frac}}= \frac{\left(h_e+1\right)C^{*}_{h_e}-h_e C^{*}_{h_e +1}}{C^{*}_{h_e}-C^{*}_{h_e+1}+1}.$$

Example

Publications are ordered by adjusted number of citations, from highest to lowest.

Citations (Ci)592616111110343211100000
Authors (Ai)332134144411242111
Adjusted Citations (\(C^*_i\))34.0615.0111.3111.006.355.003.002.001.501.001.001.000.710.000.000.000.000.00
Rank (i)123456789101112131415161718
he = 5

The largest rank where \(i \leq C^*_i\) is 5. Interpolating between this and the next largest rank yields hap.frac = 5.5746.

History

Yearhap.frac
19970.0000
19982.0873
19993.2112
20004.7500
20015.5746
20026.8473
20038.4101
200412.5000
200514.0000
200615.3181
200717.0000
200820.0000
200921.1757
201023.6004
201126.0000
201227.8909
201328.8087
201431.1598
201531.3559
201631.6510
201731.8463
201832.7500
201933.2000
202034.6667
202136.0000
202236.7692
202336.9333
202437.2956
202537.3812

References