On the prevalence and scientific impact of duplicate publications in different scientific fields (1980-2007)

TL;DR

This study analyzes the prevalence and impact of duplicate publications across scientific fields from 1980 to 2007, finding they are rare but tend to have lower impact and are often published simultaneously.

Contribution

It provides a comprehensive analysis of duplicate publication prevalence and impact across all research fields over nearly three decades, using metadata-based identification.

Findings

Duplicate papers are about 1 in 2000 publications.

Most duplicates are published within a year of each other.

Duplicate papers tend to have lower citation counts.

Abstract

The issue of duplicate publications has received a lot of attention in the medical literature, but much less in the information science community. This paper aims at analyzing the prevalence and scientific impact of duplicate publications across all fields of research between 1980 and 2007, using a definition of duplicate papers based on their metadata. It shows that in all fields combined, the prevalence of duplicates is one out of two-thousand papers, but is higher in the natural and medical sciences than in the social sciences and humanities. A very high proportion (>85%) of these papers are published the same year or one year apart, which suggest that most duplicate papers were submitted simultaneously. Furthermore, duplicate papers are generally published in journals with impact factors below the average of their field and obtain a lower number of citations. This paper provides clear evidence that the prevalence of duplicate papers is low and, more importantly, that the scientific impact of such papers is below average.