Senior scientists produce disruptive research less often than their junior colleagues, but their disruptions are more likely to persist and be consolidated by later research. That is according to an analysis of the publication records of more than 200,000 scientists, which finds that experience helps scientists produce disruptive work that takes longer to be superseded (Proc. Natl. Aca. Sci. 123 e2626763123).
Scientific advances usually build incrementally upon prior research, but progress can often be driven by disruptive research that overturns prior knowledge and leads to new lines of enquiry. One way to decide if a scientific paper is disruptive is if other articles cite it but don’t also reference studies on which the original paper builds, suggesting that it has displaced rather than consolidated earlier work.
Previous research has found that the tendency to produce disruptive work declines over a scientist’s career. But Nan Deng from the National Science Library of the Chinese Academy of Sciences and colleagues now argue that the conventional measure of disruption fails to capture how well disruption endures.
To explore this, they developed a new “persistent disruption index”, which measures whether a disruptive paper is later consolidated or displaced by the work that cites it. The team looked at how much a paper disrupts the work carried out in the references and how much that paper is disrupted by its citing papers.
They define a persistent disruption paper as sitting in the top 5% for reference disruption and the bottom 5% for citation disruption. This, they say, filters out so-called “nonreal disruption” – work that overturns previous research but is then itself quickly overturned.
Disruptive research
To explore how the persistence of disruption changes with career stage, Deng and colleagues analysed bibliographic data from SciSciNet on 206,130 scientists whose careers span more than 40 years.
Have you published a disruptive paper? New machine-learning tool helps you check
Read more
The analysis showed that as scientific careers progress, the probability of producing a disruptive paper decreases, while the probability of publishing a persistent disruption paper increases. This means that although scientists disrupt less often as they age, their disruptive research is more likely to endure and be built upon.
The team also found the disruptive work of experienced scientists received more attention from both the media and colleagues, which may go some way to increase the likelihood that their disruptive work persists.
The researchers add that evaluation and funding systems focused only on immediately visible disruption may overlook the durability of disruptive ideas, and how experience contributes to such work.