# Citation models and research evaluation

**Authors:** V. A. Traag  
**arXiv:** [2207.11116](https://arxiv.org/abs/2207.11116)  
**Date:** 2022-07-22  
**Summary page:** https://www.summarizepaper.com/en/paper/2207.11116v1/

## Key points

- The article "Citation Models and Research Evaluation" explores the study of citations in science from multiple perspectives, including scientometrics and the science of science.
- The author provides a comprehensive review of literature on citations, citation distributions, and models of citations to understand their relevance to research evaluation and the role of metrics and indicators in that process.
- The article examines how citations relate to peer review and aims to integrate both literatures for a better understanding of each other.
- The fundamental problem in research evaluation is that research quality is unobservable which has implications for conclusions drawn from quantitative studies of citations and citation models.
- To address this issue, the term "indicators" is discussed as an important concept in this context which requires proper causal reasoning when used in practice because entering causal territory may invalidate even valid indicators.
- By combining citation models with proper causal reasoning while acknowledging the fundamental problem about unobservable research quality, progress can be made towards better research evaluation.
- This article serves as a draft for a chapter in the Handbook of Computational Social Science edited by Taha Yasseri forthcoming 2023 from Edward Elgar Publishing Ltd.

## Summary

In the article "Citation Models and Research Evaluation," V.A. Traag explores the study of citations in science from multiple perspectives, including scientometrics and the science of science. The author provides a comprehensive review of literature on citations, citation distributions, and models of citations to understand their relevance to research evaluation and the role of metrics and indicators in that process. The article also examines how citations relate to peer review and aims to integrate both literatures for a better understanding of each other. The fundamental problem in research evaluation is that research quality is unobservable which has implications for conclusions drawn from quantitative studies of citations and citation models. To address this issue, the term "indicators" is discussed as an important concept in this context which requires proper causal reasoning when used in practice because entering causal territory may invalidate even valid indicators. By combining citation models with proper causal reasoning while acknowledging the fundamental problem about unobservable research quality, progress can be made towards better research evaluation. This article serves as a draft for a chapter in the Handbook of Computational Social Science edited by Taha Yasseri forthcoming 2023 from Edward Elgar Publishing Ltd.