Statistics in everyone's backyard: an impact study via citation network analysis

Lijia Wang, Xin Tong, Y. X. Rachel Wang

Abstract

The increasing availability of curated citation data provides a wealth of resources for analyzing and understanding the intellectual influence of scientific publications. In the field of statistics, current studies of citation data have mostly focused on the interactions between statistical journals and papers, limiting the measure of influence to mainly within statistics itself. In this paper, we take the first step towards understanding the impact statistics has made on other scientific fields in the era of Big Data. By collecting comprehensive bibliometric data from the Web of Science database for selected statistical journals, we investigate the citation trends and compositions of citing fields over time to show that their diversity has been increasing. Furthermore, we use the local clustering technique involving personalized PageRank with conductance for size selection to find the most relevant statistical research area for a given external topic of interest. We provide theoretical guarantees for the procedure and, through a number of case studies, show the results from our citation data align well with our knowledge and intuition about these external topics. Overall, we have found that the statistical theory and methods recently invented by the statistics community have made increasing impact on other scientific fields.

Introduction

The phrase "statisticians get to play in everyone's backyard" captures the interdisciplinary nature of the field. However, quantifying this external impact has historically been challenging. Most bibliometric studies in statistics focus inward, analyzing how journals cite each other to determine internal prestige. In the era of Big Data, where statistical methods underpin modern science, there is a pressing need to systematically measure the "export" of statistical innovation to other fields. This paper addresses this gap by analyzing a large-scale citation network to map the intellectual footprint of statistics across diverse scientific domains.

Methodology

The study employs a multi-faceted network analysis approach:

  • Descriptive Bibliometrics: Normalized citation counts are used to account for publication volume growth over time. Lorenz curves and Gini coefficients measure citation inequality and the concentration of influence.
  • Diversity Metrics: An adapted Gini index measures the diversity of citing fields, where a lower value signifies a broader, more interdisciplinary audience.
  • Local Clustering: The core technical contribution is a local clustering method using Adjusted Personalized PageRank (aPPR) combined with conductance minimization for size selection. This allows for the identification of relevant statistical research "communities" starting from specific external "seed" topics (e.g., genomics or economics) without needing to partition the entire global network.

Data and Network Construction

The researchers curated a dataset of 9,338 source articles published between 1995 and 2018 in five premier journals: Annals of Applied Statistics (AOAS), Annals of Statistics (AOS), Biometrika, Journal of the American Statistical Association (JASA), and Journal of the Royal Statistical Society: Series B (JRSSB). Using the Web of Science (WoS) database, they identified 264,356 citing papers, resulting in a directed citation network of 273,694 nodes.

Key Insights: The Export Boom and Topic Hubs

The analysis reveals a significant "export boom" in statistical citations. Fields such as Life Sciences (BIO) and Technology (TECH) have become primary consumers of statistical research.

A key finding is the divergence between "internal hits" and "external hits." Some papers are highly influential within the statistics community (theoretical works), while others serve as foundational methods for external fields despite lower internal rankings. For example, a paper on "trim and fill" methods for meta-analysis ranked 1,520th internally but 19th externally, highlighting its massive impact in psychiatry and psychology.

Furthermore, different statistical sub-fields cater to specific "customers." False Discovery Rate (FDR) research is a staple in biology, while Causal Inference is central to the social sciences and economics.

Quantitative Results

Rising Influence and Citation Inequality

Normalized citation counts show a consistent upward trend across all five journals, indicating that the impact per article is growing. JRSSB exhibits a particularly high impact per article. Lorenz curves further illustrate the distribution of citations, showing how a small number of "superstar" papers often drive a large portion of the total impact.

Normalized citation counts (a) show a steady upward trend for major journals, while Lorenz curves (b) illustrate citation inequality.

Increasing Interdisciplinary Diversity

The composition of citing fields has shifted over time, with a growing proportion of citations originating from outside the core statistics and mathematics communities. This is reflected in the "Gini concentration" of citing fields, which has been steadily dropping. This trend confirms that statistical research is reaching an increasingly diverse set of scientific disciplines.

The growth of citations from external fields (a) and the decreasing Gini concentration (b) demonstrate the expanding diversity of statistical impact.

Case Study Validation

The aPPR-based local clustering algorithm successfully identifies relevant hubs. In a case study on "single-cell transcriptomics," the algorithm correctly pinpointed the Benjamini-Hochberg paper on FDR as a central hub, aligning with its critical role in genomic differential expression testing.

Limitations and Future Outlook

While the study provides a robust framework for measuring impact, it has limitations. It treats all citations equally, regardless of their context (e.g., a passing mention versus a foundational use). The dataset, ending in 2018, does not capture recent surges in research related to the COVID-19 pandemic. Future work could expand the scope to include books and conference proceedings and integrate textual data (abstracts, titles) to further refine the network analysis.

Summary updated

Abstract

The increasing availability of curated citation data provides a wealth of resources for analyzing and understanding the intellectual influence of scientific publications. In the field of statistics, current studies of citation data have mostly focused on the interactions between statistical journals and papers, limiting the measure of influence to mainly within statistics itself. In this paper, we take the first step towards understanding the impact statistics has made on other scientific fields in the era of Big Data. By collecting comprehensive bibliometric data from the Web of Science database for selected statistical journals, we investigate the citation trends and compositions of citing fields over time to show that their diversity has been increasing. Furthermore, we use the local clustering technique involving personalized PageRank with conductance for size selection to find the most relevant statistical research area for a given external topic of interest. We provide theoretical guarantees for the procedure and, through a number of case studies, show the results from our citation data align well with our knowledge and intuition about these external topics. Overall, we have found that the statistical theory and methods recently invented by the statistics community have made increasing impact on other scientific fields.
BibTeX
@misc{wang2021statistics,
  author = {Lijia Wang and Xin Tong and Y. X. Rachel Wang},
  title = {{Statistics in everyone's backyard: an impact study via citation network analysis}},
  year = {2021},
  eprint = {2110.08605},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2110.08605},
}
Figure