[Paper Review] A complementary index to quantify an individual's scientific research output

Pablo Diniz Batista, Mônica G. Campiteli+2 more|arXiv (Cornell University)|Sep 6, 2005

scientometrics and bibliometrics research9 citations

TL;DR

This paper proposes a complementary citation index, hI = h²/Nₜ, to normalize the h-index across scientific disciplines by accounting for the number of co-authors (Nₜ) in an h researcher's most cited papers. Unlike the h-index, hI enables cross-field comparison through perfect data collapse in rank plots across diverse Brazilian research communities.

ABSTRACT

The number h of papers with at least h citations has been proposed to evaluate individual's scientific research production. This index is robust in several ways but yet strongly dependent on the research field. We propose a complementary index hI = h^2/N_t, with N_t being the total number of authors in the considered h papers. A researcher with index hI has hI papers with at least hI citation if he/she had published alone. We have obtained the rank plots of h and hI for four Brazilian scientific communities. Contrasting to the h-index curve, the hI index present a perfect data collapse into a unique allowing comparison among different research areas.

Motivation & Objective

• Address the limitation of the h-index in being highly dependent on research field, which hinders fair comparison across disciplines.
• Develop a normalized metric that accounts for co-authorship in highly cited papers to improve fairness in research evaluation.
• Enable meaningful cross-disciplinary comparison of individual scientific output by collapsing h-index rank plots into a single universal curve.
• Provide a robust, field-independent index that reflects both citation impact and collaborative effort in research output.

Proposed method

• Define a new index hI = h²/Nₜ, where h is the h-index and Nₜ is the total number of authors in the h papers with at least h citations.
• Apply the hI index to rank-ordered citation data from four distinct Brazilian scientific communities to test its performance.
• Generate rank plots of h and hI to visually assess data collapse across research fields.
• Interpret the hI index as representing the number of papers a researcher would need to have with at least hI citations if publishing alone.
• Use the hI index to compare individual research output across fields by evaluating the degree of data collapse in rank plots.
• Validate the index's robustness by demonstrating consistent scaling behavior across diverse scientific communities.

Experimental results

Research questions

• RQ1 Can a modified citation index reduce field-dependent bias in evaluating individual research output?
• RQ2 How does accounting for co-authorship in highly cited papers affect the comparability of research impact across disciplines?
• RQ3 Does the proposed hI index lead to a universal data collapse in rank plots across different scientific communities?
• RQ4 To what extent does hI preserve the robustness of the h-index while improving cross-field comparability?
• RQ5 Can hI serve as a reliable alternative to the h-index for individual research evaluation in multi-disciplinary contexts?

Key findings

• The hI index successfully achieves perfect data collapse in rank plots across four Brazilian scientific communities, indicating universal applicability.
• Unlike the h-index, which shows divergent curves across fields, the hI index produces a single, unified curve, enabling direct cross-field comparison.
• The hI index normalizes for co-authorship by incorporating Nₜ, the total number of authors in the h most cited papers, reducing field-specific bias.
• The index maintains the core principle of the h-index while introducing a field-independent normalization through the h²/Nₜ formula.
• The hI index allows researchers to be compared as if they had published alone, with hI representing the number of papers they would need to have with hI citations.
• The method demonstrates that citation impact can be fairly compared across disciplines when co-authorship is factored into the evaluation metric.