Goodness-of-Fit Testing for the Newcomb-Benford Law With Application to the Detection of Customs Fraud

Barabesi, Lucio; Cerasa, Andrea; Cerioli, Andrea; Perrotta, Domenico

doi:10.1080/07350015.2016.1172014

Cited by 31 publications

(27 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The problem with this type of data is that it is subject to intentional manipulation, thus diminishing its reliability or suitability for data analysis. Benford’s law has already attracted interest in antifraud analysis [44] , [45] . For that reason, testing Benford’s law is particularly attractive for the detection of fraudulent self-reported COVID-19 data [44] , [45] .…”

Section: Discussionmentioning

confidence: 99%

“…Benford’s law has already attracted interest in antifraud analysis [44] , [45] . For that reason, testing Benford’s law is particularly attractive for the detection of fraudulent self-reported COVID-19 data [44] , [45] . Based on the empirical findings and the simulation Table 2 , BL test shows that the data from Japan is incorrect.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

COVID-19, flattening the curve, and Benford’s law

Lee

Han

Jeong

2020

Physica A: Statistical Mechanics and its Applications

View full text Add to dashboard Cite

For many countries attempting to control the fast-rising number of coronavirus cases and deaths, the race is on to “flatten the curve,” since the spread of coronavirus disease 2019 (COVID-19) has taken on pandemic proportions. In the absence of significant control interventions, the curve could be steep, with the number of COVID-19 cases growing exponentially. In fact, this level of proliferation may already be happening, since the number of patients infected in Italy closely follows an exponential trend. Thus, we propose a test. When the numbers are taken from an exponential distribution, it has been demonstrated that they automatically follow Benford’s Law (BL). As a result, if the current control interventions are successful and we flatten the curve (i.e., we slow the rate below an exponential growth rate), then the number of infections or deaths will not obey BL. For this reason, BL may be useful for assessing the effects of the current control interventions and may be able to answer the question, “How flat is flat enough?” In this study, we used an epidemic growth model in the presence of interventions to describe the potential for a flattened curve, and then investigated whether the epidemic growth model followed BL for ten selected countries with a relatively high mortality rate. Among these countries, South Korea showed a particularly high degree of control intervention. Although all of the countries have aggressively fought the epidemic, our analysis shows that all countries except for Japan satisfied BL, indicating the growth rates of COVID-19 were close to an exponential trend. Based on the simulation table in this study, BL test shows that the data from Japan is incorrect.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

COVID-19, flattening the curve, and Benford’s law

Lee

Han

Jeong

2020

Physica A: Statistical Mechanics and its Applications

View full text Add to dashboard Cite

show abstract

“…It is plausible that this heterogeneity in the manipulation of data hampered the functioning of two-digits goodness-of-fit testing on the overall catches. We believe that future studies should compare the effectiveness of more refined goodness-of-fit tests (Barabesi et al, 2018; Lesperance et al, 2016) and the combination of these tests with other approaches, such as machine learning or the inspection of last-digits (Badal-Valero et al, 2018; Beber and Scacco, 2012) for detecting anomalies in mixtures of data. We also believe that other statistical approaches for the detection of manipulated data should be tested in ecology and conservation, because the Benford’s law has some precise distributional preconditions.…”

Section: Discussionmentioning

confidence: 99%

A fish rots from the head down: how to use the leading digits of ecological data to detect their falsification

Cerri

2018

Preprint

View full text Add to dashboard Cite

8Managing wildlife populations requires good data. Researchers and policy makers need 9 reliable population estimates and, in case of commercial or recreational harvesting, also 10 trustworthy information about the number of removed individuals. However, auditing 11 schemes are often weak and political or economic pressure could lead to data fabrication 12 or falsification. Time-series data and population models are crucial to detect anomalies, 13 but they are not always available nor feasible. Therefore, researchers need other tools 14 to identify suspicious patterns in ecological and environmental data, to prioritize their 15 controls. We showed how the Benford's law might be used to identify anomalies and 16 potential manipulation in ecological data, by testing for the goodness-of-fit of the lead-17 ing digits with the Benford's distribution. For this task, we inspected two datasets that 18 were found to be falsified, containing data about estimated large carnivore populations 19 in Romania and Soviet commercial whale catches in the Pacific Ocean. In both the two 20 datasets, the first and second digits numerical series deviated from the expected Ben-21 ford's distribution. In data about large carnivores, the first too digits, taken together, 22 also deviated from the expected Benford's distribution and were characterized by a high 23 Mean Absolute Deviation. In Soviet whale catches, while the single digits deviated from 24 the Benford's distribution and the Mean Absolute Deviation was high, the first two digits 25 were not anomalous. This controversy invites researchers to combine multiple measures 26 of nonconformity and to be cautious in analyzing mixtures of data. Testing the distribu-27 tion of the leading digits might be a very useful tool to inspect ecological datasets and 28 to detect potential falsifications, with great implications for policymakers and researchers 29 as well. For example, if policymakers revealed anomalies in harvesting data or popula-30 tion estimates, commercial or recreational harvesting could be suspended and controls 31 1 strengthened. On the other hand, revealing falsification in ecological research would be 32 crucial for evidence-based conservation, as well as for research evaluation.33 Introduction 34 Successful management of animal and plant populations requires informed decision-making. 35 Information about populations and their geographical distribution is crucial for design-36 ing effective networks of protected areas, identifying threats and integrating conservation 37 in policy making. Furthermore, as many animals and plants are traded, environmental 38 managers also need trustworthy information about the number and qualities of these indi-39 viduals which are removed from nature. During the last 20 years, conservation biology was 40 flooded with information. Digitalization enabled conservationists and agencies to store 41 and share their data (Hampton et al., 2013; Page et al., 2015). Advances in informatics 42 and the computational power of computers, allowed for an ...

show abstract

“…Benford-Like Digit Preferences. Benford's Law or the first digit law has been instrumental at catching fraud in various financial situations (Barabesi et al, 2018;Badal-Valero et al, 2018) and in small scale clinical trials (Al-Marzouki et al, 2005). The method presented here is designed with the potential to generalize and be applied to multiple sets of data of varying types and configurations (e.i.…”

Section: Methodsmentioning

confidence: 99%

Detecting fabrication in large-scale molecular omics data

Bradshaw

Payne

2020

Preprint

View full text Add to dashboard Cite

8Background: Fraud is a pervasive problem and can occur as fabrication, falsification, 9 plagiarism or theft. The scientific community is not exempt from this universal problem and 10 several studies have recently been caught manipulating or fabricating data. Current measures 11 to prevent and deter scientific misconduct come in the form of the peer-review process and on-12 site clinical trial auditors. As recent advances in high-throughput omics technologies have 13 moved biology into the realm of big-data, fraud detection methods must be updated for 14 sophisticated computational fraud. In the financial sector, machine learning and digit-preference 15 are successfully used to detect fraud. 16Results: Drawing from these sources, we develop methods of fabrication detection in 17 biomedical research and show that machine learning can be used to detect fraud in large-scale 18 omic experiments. Using the raw data as input, the best machine learning models correctly 19 predicted fraud with 84-95% accuracy. With digit frequency as input features, the best models 20 detected fraud with 98%-100% accuracy. All of the data and analysis scripts used in this project 21 are available at https://github.com/MSBradshaw/FakeData. 22Conclusions: Using digit frequencies as a generalized representation of the data, multiple 23 machine learning methods were able to identify fabricated data with near perfect accuracy. 24 25

show abstract

Goodness-of-Fit Testing for the Newcomb-Benford Law With Application to the Detection of Customs Fraud

Cited by 31 publications

References 27 publications

COVID-19, flattening the curve, and Benford’s law

COVID-19, flattening the curve, and Benford’s law

A fish rots from the head down: how to use the leading digits of ecological data to detect their falsification

Detecting fabrication in large-scale molecular omics data

Contact Info

Product

Resources

About