Source-linked AI summary
An empirical analysis of the use of alphabetical authorship in scientific publishing
Ludo Waltman
TL;DR
The paper asks how often and where alphabetical authorship is used, addressing the lack of evidence covering science as a whole. It empirically analyzes all scientific fields while distinguishing intentional from incidental alphabetical ordering, finding that intentional use declined to 3.7% in 2011.
Problem
Existing studies examined alphabetical authorship in specific fields, leaving no evidence covering science as a whole; authorship order also matters for assigning credit.
Method
The paper empirically analyzes alphabetical authorship across all scientific fields, distinguishing intentional ordering from incidental alphabetical order and estimating intentional use from bibliographic data.
Results
3.7% of publications were estimated to be intentionally alphabetical in 2011, down from 8.9% in 1981; use was most common in mathematics, business and finance, economics, and particle physics.
Takeaways & Limitations
Alphabetical authorship is declining, but remains especially common in mathematics, business and finance, economics, and physics, particles and fields.
Takeaways & Limitations
Results depend on Web of Science coverage, which changed over time, and on the definition used for alphabetical authorship.
Abstract
from arXiv · showhide
There are different ways in which the authors of a scientific publication can determine the order in which their names are listed. Sometimes author names are simply listed alphabetically. In other cases, authorship order is determined based on the contribution authors have made to a publication. Contribution-based authorship can facilitate proper credit assignment, for instance by giving most credits to the first author. In the case of alphabetical authorship, nothing can be inferred about the relative contribution made by the different authors of a publication. In this paper, we present an empirical analysis of the use of alphabetical authorship in scientific publishing. Our analysis covers all fields of science. We find that the use of alphabetical authorship is declining over time. In 2011, the authors of less than 4% of all publications intentionally chose to list their names alphabetically. The use of alphabetical authorship is most common in mathematics, economics (including finance), and high energy physics. Also, the use of alphabetical authorship is relatively more common in the case of publications with either a small or a large number of authors.
1. Introduction
The paper addresses how scientific authors determine authorship order and why that choice matters for assigning credit. It fills a field-wide evidence gap by empirically analyzing alphabetical authorship across all sciences.
- Contribution-based ordering can identify the first author as the most significant contributor.Alphabetical ordering provides no information about relative contributions.
- Authorship-order practices matter because they affect how credit can be assigned among coauthors.When names are alphabetical, the relative credit due to each author cannot be inferred from position.
- Earlier studies examined alphabetical authorship in specific scientific fields, but no study covered science as a whole.The paper presents its analysis as filling this literature gap.
- The analysis measures alphabetical-authorship frequency, changes over time, disciplinary differences, and partial alphabetical authorship.It distinguishes intentional alphabetical ordering from incidental alphabetical ordering produced by a non-alphabetical criterion.
2. Intentional vs. incidental alphabetical authorship
The paper distinguishes intentional alphabetical authorship from incidental alphabetical order that arises when contribution-based ordering happens to match the alphabet. It models and estimates intentional alphabetical authorship while correcting for this incidental component.
- Incidental alphabetical authorship occurs when contribution-based ordering happens to coincide with alphabetical order.Jones is the main contributor but is listed first because the contribution-based order also matches the alphabet.
- 50% is the assumed incidental-alphabetical probability for two-author publications, versus 2.8 × 10^-7 for ten-author publications.The calculation assumes contribution-based ordering and no correlation between author-name alphabetic position and contribution.
- The probability of incidental alphabetical authorship quickly decreases as the number of authors increases.Figure 1 illustrates this relationship under the stated assumptions.
- Correcting for incidental alphabetical authorship is necessary to interpret changes in observed alphabetical authorship over time.Without correction, a decline could reflect either less intentional alphabetical ordering or changes in incidental alphabetical ordering.
- The model treats intentional alphabetical authorship as unobserved while using observed author counts and alphabetical ordering to estimate its average probability.The estimator is shown to be unbiased, and for sufficiently large author counts it is close to the observed proportion of alphabetical publications.
3. Empirical analysis
Across 1981–2011, alphabetical authorship declined overall, but its intentional use varied substantially across disciplines and publication sizes. The analysis distinguishes intentional from incidental alphabetical ordering and shows that small and large author groups exhibit relatively more alphabetical authorship.
- Scope and definitions: The analysis covers multi-author articles, notes, and reviews indexed in Web of Science from 1981–2011.Single-author publications are excluded; “alphabetical publication” denotes a multi-author publication with alphabetically listed authors.
- General trends: 15.9% of publications were alphabetical in 2011, down from 32.2% in 1981, but this decline partly reflects more authors per publication.The average number of authors increased over the same period, reducing incidental alphabetical ordering.
- General trends: 3.7% of publications were intentionally alphabetical in 2011, down from 8.9% in 1981, showing that intentional and incidental alphabetical authorship both declined.The increase in average author count therefore explains only part of the overall decrease.
- Disciplinary differences: Intentional alphabetical authorship differed widely across 223 subject categories, being common in some fields and virtually absent in others.Alphabetical ordering in many categories was usually incidental rather than intentional.
- Disciplinary differences: More than 50% of publications were intentionally alphabetical in Mathematics, Business, finance, Economics, and Physics, particles & fields.Physics, particles & fields had an average of 18.8 authors per publication, much higher than the other listed categories.
4. Conclusions
The analysis finds declining alphabetical authorship, strong disciplinary variation, and differences by publication size, while emphasizing implications for credit assignment and methodological limitations.
- 8.9% of publications were intentionally alphabetical in 1981, declining to 3.7% in 2011.
- More than 50% of publications were intentionally alphabetical in Mathematics, Business, finance, Economics, and Physics, particles & fields during 2007–2011.Alphabetical hyperauthorship appears more common in physics than biomedical research.
- Partial alphabetical authorship is most common in natural-science fields with relatively many authors, while alphabetical authorship is relatively common with either few or many authors.
- Alphabetical authorship complicates credit assignment because author-list position may provide no information about individual contributions.This concern is especially relevant for multi-author publications and hyperauthorship.
- The findings depend on Web of Science coverage, the definition of alphabetical authorship, and assumptions used to estimate intentional alphabetical authorship.The authors note that actual use may be higher, particularly for publications with many authors.
Appendix
The appendix establishes that the estimator p̂ is unbiased for p by modeling intentional alphabetical ordering and incidental alphabetical ordering under non-alphabetical authorship.
- Appendix: The probability that publication i is listed alphabetically combines intentional alphabetical ordering with incidental alphabetical ordering under a non-alphabetical criterion.The latter probability is based on the assumption that all author-name orderings are equally likely.
- Appendix: The derivation computes the expected value of the individual estimator p̂_i using the probability model for alphabetical ordering.The appendix substitutes the probability expression into the expectation and simplifies the result.
- Appendix: The expected value of p̂ equals p, proving that p̂ is an unbiased estimator of p.