Quantifying science and innovation

Author:

He, Zhongyang

Graduate Program:

Energy and Mineral Engineering

Degree:

Doctor of Philosophy

Document Type:

Dissertation

Date of Defense:

June 05, 2019

Committee Members:

Zhen Lei, Dissertation Advisor/Co-Advisor
Zhen Lei, Committee Chair/Co-Chair
Seth Adam Blumsack, Committee Member
Joel Reid Landry, Committee Member
Zhongyi Yuan, Outside Member

Keywords:

Innovation
Patent
Econometrics
Computational Social Science
Novelty

Abstract:

Science and technology is an essential engine of economic growth. The increasing availability of data offers many opportunities to examine and understand activities, mechanisms, and processes in science and innovation. In my dissertation, I investigate two novel and interesting questions that could be of important implications for science and technology policies: citation dynamics of atypical papers, and failed innovative endeavors. Citations, widely accepted as a quality metric of research outputs, exhibit a short-term rise-and-fall trajectory for most papers. However, many important discoveries deviate from this pattern. In this study, we identify and classify atypical papers whose citation trajectories do not fit this pattern from two large datasets. For some papers, we verify that their pattern can be explained by a later than usual immediacy parameter within prior framework. For remaining atypical papers, we propose a second act model that can accurately describe their two-peak citation dynamics. This model provides a mechanistic framework to understand atypical papers and can be applied to identification of the time of exogenous shocks. Many innovative endeavors, such as grant proposals and patent applications, fail to get funded or granted. While successful cases are widely examined, failed ones are largely ignored. Here we fill this gap by analyzing failed endeavors from three diverse large-scale datasets, including patent applications, NIH grant proposals, and Kickstarter projects. We first classify diamonds-in-the-rough (DIRs) and lumps-of-coal within failed endeavors based on a new quantitative metric. We find that DIRs are similar to the most impactful successes in patent and grant systems, and there exist two different transition patterns from failure to success. We devise a theoretical model that highlights the possible and interesting discrepancy between evaluation and recognition processes of an innovation. The model not only explains empirical findings, but reveals important insights regarding evaluation and recognition in innovative systems, which could have important policy implications.