Fairer Citation Based Metrics

Ognjen ArandjelovićPublishing 2016 · DOI

Abstract: I describe a simple modification which can be applied to any citation count based index (e.g. Hirsch’s h-index) quantifying a researcher’s publication output. The key idea behind the proposed approach is that the merit for the citations of a paper should be distributed amongst its authors according to their relative contributions. In addition to producing inherently fairer metrics I show that the proposed modification has the potential to normalize partially for the unfair effects of honorary authorship and thus discourage this practice.

Introduction

In the high-stakes world of academic publishing, metrics like the h-index serve as the primary currency for grants, promotions, and professional prestige. However, a significant flaw persists in how these metrics calculate impact: they treat authorship as a binary state. Whether a researcher is the lead author who conducted the bulk of the study or a senior colleague who provided a minor review, both receive equal citation credit. This "positive sum" logic incentivizes the unethical practice of "honorary authorship," leading to citation inflation and obscuring individual contributions.

In "Fairer Citation Based Metrics," Ognjen Arandjelović proposes a mathematical adjustment to distribute citation credit more equitably based on author rank.

The Proposed Adjustment

The central proposal is a "Contribution Weighted Citations" approach. Instead of calculating an index based on raw citation counts, the index is derived from adjusted counts where credit is weighted by the inverse of the researcher's rank in the author list:

$$\text{Adjusted Citations}_i = \frac{\text{Citations}_i}{\text{Author Rank}_i}$$

This model operates on the principle that author rank is a viable proxy for relative contribution. Because author rank is a zero-sum game—to move one person up, another must move down—this system introduces a natural self-policing mechanism. Authors are incentivized to ensure the list accurately reflects effort to protect their own weighted impact.

Comparison of total citation contribution between simple counting and the proposed rank-based adjustment.

As shown in the figure above, the standard "Simple count" method allows the total credit awarded for a single paper to grow linearly with the number of authors. The proposed adjustment significantly dampens this inflation, ensuring that the total credit distributed remains closer to the actual impact of the work.

Case Study and Validation

The paper validates this approach using two primary datasets: a 50-year longitudinal case study of a highly successful researcher and an analysis of 11 world-renowned computer scientists with h-indexes exceeding 110.

The longitudinal study reveals a clear "seniority shift." As the subject researcher became more senior, their average author rank increased (moving toward the end of the list), while their publication volume reached physically improbable levels—peaking at 117 papers in a single year.

Longitudinal analysis of a researcher's career showing (a) h-index divergence, (b) publication volume, and (c) increasing average author rank.

Under standard metrics, the researcher's h-index continues to grow superlinearly. However, the modified metric (shown in plot 'a' above) moderates this growth, reflecting the transition from a primary contributor to a supervisorial or honorary role. When applied to the group of 11 top computer scientists, the average h-index dropped from 118 to 69 (a 42% reduction), and total citation counts fell by an average of 60%.

Discussion and Limitations

While the $1/\text{rank}$ weighting provides a robust heuristic for many STEM fields, it has notable limitations. In disciplines such as Mathematics or Economics, where authors are typically listed alphabetically, rank does not convey contribution. Furthermore, in fields where the "Last Author" position is reserved for the Principal Investigator (PI), this model may unfairly penalize senior leadership.

Future work could explore more complex weighting schemes, such as those based on the CRediT (Contributor Roles Taxonomy), to better reflect non-linear contribution distributions.

Conclusion

By weighting citations by author rank, we can create a more transparent and fair assessment of scientific impact. This simple modification not only provides a more accurate picture of individual achievement but also serves as a mathematical deterrent against the practice of honorary authorship, ultimately preserving the integrity of academic metrics.

Summary updated

Abstract

Abstract I describe a simple modification which can be applied to any citation count based index (e.g. Hirsch’s h-index) quantifying a researcher’s publication output. The key idea behind the proposed approach is that the merit for the citations of a paper should be distributed amongst its authors according to their relative contributions. In addition to producing inherently fairer metrics I show that the proposed modification has the potential to normalize partially for the unfair effects of honorary authorship and thus discourage this practice.
BibTeX
@inproceedings{arandjelovic2016fairer,
  author = {Ognjen Arandjelović},
  title = {{Fairer Citation Based Metrics}},
  booktitle = {Publishing},
  year = {2016},
  doi = {10.1007/s12109-016-9461-8},
}
Figure