Bibliometric author evaluation through linear regression on the coauthor network

TL;DR

This paper introduces a linear regression model to evaluate individual author contributions within coauthored works, aiming to improve credit assignment for academic recognition and funding decisions.

Contribution

It presents a simple, statistically estimable model based on constant 'author ability' to assess individual contributions in coauthored publications.

Findings

The model's rankings align well with fractional citation counts.

Noticeable differences highlight the model's unique perspective.

The approach is validated on a large coauthor network from CiteSeerX.

Abstract

The rising trend of coauthored academic works obscures the credit assignment that is the basis for decisions of funding and career advancements. In this paper, a simple model based on the assumption of an unvarying "author ability" is introduced. With this assumption, the weight of author contributions to a body of coauthored work can be statistically estimated. The method is tested on a set of some more than five-hundred authors in a coauthor network from the CiteSeerX database. The ranking obtained agrees fairly well with that given by total fractional citation counts for an author, but noticeable differences exist.