data-forensics

Data authenticity checks apply seven forensic tests like GRIM, SPRITE, and Benford's Law to raw datasets to detect fabrication. The workflow outputs a report with a GREEN, YELLOW, or RED risk rating and recommended actions for integrity assurance.

Benford's Law is a forensic test analyzing the distribution of leading digits in a dataset to detect data inconsistencies. It flags deviations from expected digit frequencies, helping identify potentially fabricated or manipulated numerical values in research and financial datasets.

You can run forensic tests on SPSS and Stata files alongside CSV and Excel formats. These tests assess data integrity in medical, financial, or research datasets prior to downstream analyses or manuscript generation.

Data integrity validation using GRIM and SPRITE tests checks if reported summary statistics are mathematically consistent with given sample sizes. These tests identify impossible combinations of means, standard deviations, and sample sizes in raw datasets.

Detecting duplicate patterns and terminal digit anomalies is best done through dedicated forensic tests that scan raw datasets for repeated rows and suspicious last-digit distributions. These tests flag potential data fabrication and generate a risk-rated report.

Automated data forensics should not be solely relied upon when datasets are extremely small or lack sufficient numerical variance for statistical tests like Benford's Law or Variance Uniformity. Forensic tests flag risk levels but require human interpretation for edge cases.