Analysis of Collaboration in CS Prizewinning with a Nobel-Turing Comparison

TL;DR

This study analyzes how collaboration behaviors influence prizewinning in computer science, revealing that prizewinners tend to collaborate more and earlier with other winners, and that collaboration increases the odds of winning.

Contribution

It provides the first comprehensive analysis of collaboration patterns among CS prizewinners and compares them to Nobel-Turing winners, highlighting the impact of collaboration on prize success.

Findings

Prizewinners collaborate earlier and more with other winners.

Post-win collaboration increases likelihood of winning.

General CS prizes see more collaboration than specialized ones.

Abstract

In the scientific community, prizes play a pivotal role in shaping research trajectories by conferring credibility and offering financial incentives to researchers. Yet, we know little about the relationship between academic collaborations and prizewinning. By analyzing over 100 scientific prizes and the collaboration behaviors of over 5,000 prizewinners in CS, we find that prizewinners collaborate earlier and more frequently with other prizewinners than researchers who have not yet received similar recognition. Moreover, CS researchers across age groups collaborate more with prizewinners after winning their first prize, and collaborating with prizewinners after their first win increases the likelihood of the collaborator winning an award. We find that recipients of general CS prizes collaborate more than recipients of more specialized prizes, who collaborate less frequently. With Coarsened Exact Matching (CEM) and regression, we find an increase in prizewinning odds with strength of prizewinner collaboration. We examine the context of recent Nobel Prizes going to CS researchers by showing how an increasing share of Physics awards go to Physics-CS collaborations, and contrast Nobel-Turing winning author's trajectories. Our findings shed light on the relationship between prizewinning and collaboration.