Scientific Novelty, Complexity and Uncertainty: What makes Science Risky?
Nicola Melluso, Katholieke Universiteit, Leuven
Decision-making in science, particularly regarding funding, is inherently shaped by the uncertainty surrounding the future impact of research endeavors. While understanding the factors influencing the impact of scientific research is essential, accurately identifying the specific features that drive research's potential remains elusive. This study addresses this gap by studying the risk profile of scientific research directly utilizing Natural Language Processing to analyze the textual content of the entire population of scientific publications from the last century. I develop and validate two novel text-based metrics: novelty and complexity. Using Nobel Prizes papers, which arguably pioneered new scientific ideas with a major impact on scientific progress, I demonstrate that novelty and complexity strongly influence future impact. Crucially, I find that the interplay between novelty and complexity significantly influences the risk profile of a research endeavor. High novelty coupled with high complexity offers the potential for breakthrough discoveries, but also leads to greater uncertainty, higher likelihood of failure and delayed recognition. I further show that top scientists consistently embrace such risk, demonstrating a willingness to pursue highly novel and highly complex research compared to their peers. These findings provide empirical evidence for a critical tension in science: overly risk-averse strategies could unintentionally stifle breakthroughs in science.
Zoom link to join
https://unu-merit-eu.zoom.us/j/88182790733?pwd=ZitsWXpxUmFVdmw4YTYwNW9wV0I4QT09
About the speaker
Nicola Melluso is a researcher at KU Leuven University in Belgium. His research interests mainly focus on Innovation Management and the Science of Science.
Venue: Room 0.18, Boschstraat 24, Maastricht (UNU-MERIT) and Online
Date: 11 April 2024
Time: 12:00 - 13:00 CEST