A machine learning approach to quantify gender bias in collaboration practices of mathematicians - Original research article

Aug 5, 2026

Upload by Jiven

A machine learning approach quantifies gender bias in mathematics collaboration using coauthorship-based signals. The analysis focuses on two collaboration dimensions: the number of distinct coauthors and the number of single-authored papers. Controlling for potential confounders such as seniority and total publication volume, the results show slightly larger collaboration networks for female mathematicians and significantly fewer single-authored publications compared with male colleagues. The findings validate earlier descriptive observations while providing more precise statistical models for how gender relates to collaboration patterns.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

AccessFree to download

Introduction

Shift from individual to group effort in mathematics

Coauthorship trends and career implications

Role and persistence of single-authored publications

Methods and key focus areas

Distinct coauthors as collaboration network size

Single-authored papers as evidence of maturity

TYPE Original Research PUBLISHED 18 January 2023 DOI 10. 3389/fdata.2022.989469 OPEN ACCESS EDITED BY Xi Niu, University of North Carolina at Charlotte, United States REVIEWED BY Alexander V. Mantzaris, University of Central Florida, United States Xueru Zhang, The Ohio State University, United States Riyi Qiu, University of North Carolina at Charlotte, United States *CORRESPONDENCE Helena Mihaljevi [helena. mihaljevic@htw-berlin.de](helena. mihaljevic@htw-berlin.de) †These authors have contributed equally to this work SPECIALTY SECTION This article was submitted to Data Mining and Management, a section of the journal Frontiers in Big Data RECEIVED 08 July 2022 ACCEPTED 28 December 2022 PUBLISHED 18 January 2023 CITATION Steinfeldt C and Mihaljevi H (2023) A machine learning approach to quantify gender bias in collaboration practices of mathematicians. Front. Big Data 5:989469 . doi: 10.3389/fdata.2022.989469 COPYRIGHT © 2023 Steinfeldt and Mihaljevi . This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. A machine learning approach to quantify gender bias in collaboration practices of mathematicians Christian Steinfeldt† and Helena Mihaljevi*† Department 4-Computer Science, Communication and Economics, Hochschule für Technik und Wirtschaft Berlin, University of Applied Sciences, Berlin, Germany Collaboration practices have been shown to be crucial determinants ofscientiﬁc careers. We examine the e􀀀ect of gender on coauthorship-based collaboration in mathematics, a discipline in which women continue to be underrepresented, especially in higher academic positions. We focus on two key aspects of scientiﬁc collaboration—the number of di􀀀erent coauthors and the number of single authorships. A higher number of coauthors has a positive e􀀀ect on, e.g., the number of citations and productivity, while single authorships, for example, serve as evidence of scientiﬁc maturity and help to send a clear signal of one’s proﬁciency to the community. Using machine learning-based methods, we show that collaboration networks of female mathematicians are slightly larger than those of their male colleagues when potential confounders such as seniority or total number of publications are controlled, while they author signiﬁcantly fewer papers on their own. This conﬁrms previous descriptive explorations and provides more precise models for the role of gender in collaboration in mathematics. KEYWORDS collaboration networks, machine learning, gender in mathematics, regression-based analysis, authorship, scientiﬁc publishing, single-authored publications, coauthorship 1. Introduction Nowadays, research is built as group e􀀓ort, in which individuals collaborate through joint discussions of ideas and methods, oral and written presentations, and the integration of obtained feedback into further work (Ductor et al., 2018) . It is thus not so surprising that the notion of mathematics as a discipline pursued by individual geniuses is considered outdated. But even historically, mathematics o􀀓ers a range of examples of fruitful collaborations. The presumably most prominent such example are Hardy and Littlewood, who jointly wrote about 100 papers of great importance for pure mathematics in England in the 􀀂rst half of the twentieth century (Wilson, 2002) . Paul Erds, one of the most proli􀀂c mathematicians in history, helped to transform the discipline into asocial activity by collaborating with more than 500 coauthors. The public forum-based collaboration on the Hales-Jewett theorem, driven by Tim Gowers’ Polyma

FAQs

What two aspects of collaboration does the study analyze?

The study examines (1) the number of different coauthors and (2) the number of single-authorships in mathematicians’ publishing records.

How does the study assess gender differences?

It uses machine learning-based methods and controls for potential confounders such as seniority and total number of publications.

What is the main finding about female mathematicians versus male colleagues?

Female mathematicians have slightly larger collaboration networks, but they author significantly fewer papers on their own.

Aug 5, 2026

Upload by Jiven