CHAMPAIGN, Ill. — When researchers make bold rhetorical assertions in their studies about the novelty of their work — regardless of its actual uniqueness — their papers receive more citations from other scholars, a new study found. Serving as a key mechanism for establishing scientific authority, the use of novelty claims is particularly pronounced in science, technology, engineering and medicine — disciplines where the pressure to achieve firsts is particularly intense — and these claims are linked with the existing gender gaps in scientific impact within these fields, according to study author and University of Illinois Urbana-Champaign sociology professor Jina Lee.
“Researchers’ claims about the novelty or originality of their work and its contributions to their field — often through the use of terms such as ‘innovative’ or ‘groundbreaking’ — are a form of rhetorical framing and self-promotion that they use to establish themselves as authorities,” Lee said. “There have been many studies about gender gaps in self-promotion and how those tend to reproduce gender gaps in career progress and career rewards. But far less attention has been given to these novelty claims of authority and, in this study, I show how these claims function as gendered professional practices across STEM and other disciplines to shape scientific impact.”
People are also reading…
Published in the journal Gender and Society, the study is based on Lee’s review of more than 106,000 scholarly papers. In the study, Lee defined novelty as “the extent that research genuinely departs from existing knowledge through unusual combinations of intellectual foundations, new theories or methods, or findings that substantially advance understanding.” However, true novelty exists independently of study authors’ framing — i.e., work that is highly novel may be described modestly by some scholars, while others make bold claims about research that consists of only incremental contributions, Lee said.
Lee noted that disciplines such as STEM — where much of the work is incremental advancements and breakthrough discovery is the singular gold standard — heighten the competition among researchers to be the first to make such discoveries, whereas the humanities, social sciences and other non-STEM disciplines provide scholars with diverse pathways for achieving recognition, such as using new approaches or theories and researching understudied areas.
“STEM fields tend to focus excessively on novelty, originality, breakthrough and innovation, while the humanities and social sciences tend to value multiple types of contributions,” Lee said. “There also is a cultural expectation in STEM that men are more creative, innovative and better at making these achievements. And the evaluative criteria that are typically associated with masculinity are more important in STEM.”
Lee randomly sampled peer-reviewed papers in an open access database and linked them with datasets from SciSciNet, a repository for studying science research. Spanning six decades, the final sample included papers in STEM and other fields that were written by teams of researchers. Lee focused on the first and last authors of the papers due to their significance in academic publishing: First authors typically contribute the most to studies, while last authors tend to provide funding, research facilities and intellectual guidance, she said.
Women were the first or last authors of slightly more than 41% and 36%, respectively, of the non-STEM papers. Among those in the STEM disciplines, women were 27% of the first authors and 17% of the last authors, Lee found.
To assess the authors’ scholarly influence and impact within their field, Lee counted the number of citations each paper received within 10 years of its publication. She also explored how papers’ citations and impact differed by the first or last author’s gender across the full sample and in the STEM and non-STEM groups, whether papers written by men received more citations, and if gender gaps in impact were mediated by gender differences in novelty claims. To ascertain authors’ gender, Lee relied on SciSciNet data and precalculated gender inference and cultural consensus scores of names using a database that incorporates data from more than 150 countries.
Using a dictionary that Lee and her then-co-authors developed for a previous study, Lee searched the papers’ abstracts for 16 terms frequently used to signal novelty — such as new, original and unprecedented — and tallied them for each paper. Non-STEM papers averaged slightly more of these novelty claims — 0.85 compared with 0.74 for those in STEM fields, Lee found.
The number of citations that papers received varied considerably across the full sample — ranging from none to 3,057 — and averaged 31.47 citations per paper, Lee found. Novelty claims were significantly associated with citation counts — representing a 2.9% increase in the number of citations per claim, Lee said. However, papers with women as first authors contained significantly fewer — 4.8% fewer — novelty claims compared with those that had men as the first authors.
Without including all the control variables — including team size, women’s representation in the field, and journal and author metrics — the base model indicated that women as first authors had 8.2% greater citation counts compared with those on which men were the first authors. Conversely, Lee found that papers with women as the last author had 3.6% fewer citations, even when controlling for papers’ conventionality.
However, these initial gender effects reversed course when Lee implemented the full set of control variables. Papers with women as first authors had a decrease of 2.3% in expected citation counts compared with those that had men as first authors. Although Lee found a small positive effect for papers with women as the last authors, the effect was not significant.
Across disciplines Lee found contrasting effects between gender and scientific impact when she compared papers in STEM disciplines with those in other fields. In the humanities and social sciences, women first or last authors had no significant effects on citation counts — suggesting there was no evidence that gender affected scientific impact, she wrote.
“(However,) STEM papers with women as first authors contained 5.4% fewer novelty claims and received 3.1% fewer citations, while those with women as last authors had 3.4% fewer novelty claims but received 1.6% more citations,” Lee said. “I controlled for the actual novelty and typicality of the papers using an established measure, and rhetorical claims of novelty still had a statistically significant effect. The most important takeaway is that this only happened in STEM fields — and not in the humanities and social sciences, where multiple types of gender-neutral contributions are valued and the culture is less masculine dominated. In STEM disciplines, gender gaps in impact are significantly mediated by gender differences in the use of novelty claims.”
