Preprint – submitted to Heliyon.
ISSN/ISBN: Not available at this time. DOI: 10.2139/ssrn.4516032
Abstract: The COVID-19 pandemic has brought about significant challenges worldwide, with governments and health organizations striving to accurately monitor and analyze infection data during the first wave. Traditional statistical methods have been widely employed for detecting anomalies in COVID-19 datasets. However, in recent years, the application of Benford’s Law has gained attention as a powerful tool for anomaly detection in various domains, including financial fraud detection and election result analysis. This study presents a novel application of Benford’s Law for COVID-19 anomaly detection, aimed at identifying potential inconsistencies or anomalies in COVID-19 data reporting during the first wave of the pandemic. In this study, we apply Benford’s Law for first digits with focus on the[52]ψ− Factor (ψ∗),[35] Statistic (α∗),[51] Statistic (ω∗) and Kolmogorov-Smirnov test to detect inconsistencies in world continental COVID-19 data. By comparing the observed distribution of leading digits in COVID-19 cumulative confirmed cases, cumulative deaths, cumulative recovered cases and cumulative active cases against the expected distribution posited by Benford’s Law, deviations from the expected patterns were identified. We used the deviation from the Newcomb–Benford law of anomalous numbers as a proxy for data accuracy. Such deviations may indicate data manipulation, reporting errors, or irregularities in the dataset, thereby drawing attention to potentially problematic areas that require further investigation. The study found that with the exception of Australia/Oceania continent which display severe nonconformity to the Benford’s law due to its inherent data structure, world continental COVID-19 data was dependable during the first wave of the pandemic. To improve the accuracy and reliability of anomaly detection processes, other anomaly detection methods such as density-based techniques and domain-specific knowledge are recommended in combination with the Newcomb-Benford test.
Bibtex:
@misc{,
author = {Edmund Fosu Agyemang and Joseph Agyapong Mensah and Eric Nyarko},
title = {How dependable is World Continental COVID-19 data? Disclosure of Inconsistencies in Daily Reportage Confirmed Cases, Recovered and Deaths During First Wave},
year = {2023},
url = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4516032},
doi = {10.2139/ssrn.4516032},
}
Reference Type: Preprint
Subject Area(s): Medical Sciences