Applying Benford's Law for Assessing the Validity of Social Science Data

Provides a method and workflow, based on Benford's Law, for assessing the validity of self-reported social science data.

Michael A. Long (Author), Paul B. Stretesky (Author), Kenneth J. Berry (Author), Janis E. Johnston (Author), Michael J. Lynch (Author)

9781009123075, Cambridge University Press

Hardback, published 23 November 2023

224 pages
23.5 x 15.5 x 1.7 cm, 0.494 kg

'Long and co-authors offer a systematic and accessible approach, 'Benford agreement analysis', to dealing with data validity concerns. The examples and discussion of when to use the approach make this book equally valuable for methodologists and empirical social scientists, and will work very well in research methods and statistics courses.' Andrew Jorgenson, University of British Columbia, Canada

Benford's Law is a probability distribution for the likelihood of the leading digit in a set of numbers. This book seeks to improve and systematize the use of Benford's Law in the social sciences to assess the validity of self-reported data. The authors first introduce a new measure of conformity to the Benford distribution that is created using permutation statistical methods and employs the concept of statistical agreement. In a switch from a typical Benford application, this book moves away from using Benford's Law to test whether the data conform to the Benford distribution, to using it to draw conclusions about the validity of the data. The concept of 'Benford validity' is developed, which indicates whether a dataset is valid based on comparisons with the Benford distribution and, in relation to this, diagnostic procedure that assesses the impact of not having Benford validity on data analysis is devised.

1. Introduction
2. Validity and self-reported data
3. Benford's law and assessing conformity
4. Data characteristics and the workflow of Benford agreement analysis
5. Benford agreement analysis of the Sea Around Us project fish landings data
6. Benford agreement analysis of US and global COVID-19 new-cases data
7. Assessing impacts of problematic Benford validity
8. Conclusion.

Subject Areas: Psychology [JM]