The geography of novel and atypical research

TL;DR

This paper investigates the geography of creative science by measuring country-level novelty and atypicality in publications, using the Microsoft Academic Graph (MAG) and the Uzzi et al. (2013) framework. It introduces country-specific indicators r_N and r_A to assess over- or under-representation of novel and atypical work, and applies hypergeometric testing, fractional counting, and time-series clustering to reveal persistent national differences and emerging patterns, notably China’s rise in both novelty and atypicality. The analysis demonstrates discipline- and collaboration-dependent heterogeneity and validates Uzzi’s findings at a large scale, while highlighting policy-relevant implications for balancing risk-taking with traditional impact metrics. The study provides publicly available classifications and data to enable replication and further exploration of creativity in science, with implications for science policy and evaluation practices.

Abstract

The production of knowledge has become increasingly a global endeavor. Yet, location related factors, such as local working environment and national policy designs, may continue to affect what kind of science is being pursued. Here we examine the geography of the production of creative science by country, through the lens of novelty and atypicality proposed in Uzzi et al. (2013). We quantify a country's representativeness in novel and atypical science, finding persistent differences in propensity to generate creative works, even among developed countries that are large producers in science. We further cluster countries based on how their tendency to publish novel science changes over time, identifying one group of emerging countries. Our analyses point out the recent emergence of China not only as a large producer in science but also as a leader that disproportionately produces more novel and atypical research. Discipline specific analysis indicates that China's over-production of atypical science is limited to a few disciplines, especially its most prolific ones like materials science and chemistry.

Figures (7)

• Figure 1: Reproducing the relationship between atypical science and scientific impact, as reported in Uzzi-atypical-2013. The figure considers all papers published in the 1990s. (A) The compositions of the four groups of papers based on our implementation, which uses MAG, and based on Uzzi-atypical-2013, which used WoS, indicating that each group's relative share remains almost identical despite different datasets are used. (B) Hit rates of the four groups of papers in MAG and WoS, showing that the HC-HN group has the highest hit rate. Hit rate is defined as the percentage of papers in a group whose citations are ranked into the top 5% of all papers. The dashed horizontal line marks the baseline hit rate (5%).
• Figure 2: The persistent relationship between atypicality and scientific impact across decades. (A) Compositions of the four categories of MAG papers published in each decade, suggesting that atypical papers are the least common in science. (B) Hit rates of the four groups of papers by decade, indicating that atypical papers are most likely to be hits.
• Figure 3: The tendency $r_N$ of the production of novel papers by country/region during 1990--2019. Color encodes $r_N$.
• Figure 4: Ranking of countries/regions based on their $r_A$, the tendency to produce atypical papers, in the three decades. Ranking is within the 38 selected places, and color encodes $r_A$.
• Figure 5: Comparison of the tendency to generate novel and atypical science. Color represents decade, and scatter size indicates relative share among all papers published in the decade.