Get started with the jfa package

Introduction

Welcome to the ‘Get started’ page of the jfa package. jfa is an R package that provides Bayesian and classical statistical methods for audit sampling, data auditing, and algorithm auditing. This page points you to the vignettes accompanying each of these three subjects.

Audit sampling

Firstly, jfa facilitates statistical audit sampling. That is, the package provides functions for planning, performing, and evaluating an audit sample compliant with international standards on auditing.

• Audit sampling: Get started
• Creating a prior distribution for audit sampling
• Planning statistical audit samples
• Selecting statistical audit samples
• Evaluating statistical audit samples
• Evaluating audit samples with partial misstatements
• Walkthrough of the classical audit sampling workflow
• Walkthrough of the Bayesian audit sampling workflow

Data auditing

Secondly, jfa facilitates statistical data auditing. That is, the package includes functions for auditing data, such as testing the distribution of first digits of a data set against Benford’s law, or assessing whether a data set includes an unusual amount of repeated values.

• Data auditing: Get started
• Digit analysis

Algorithm auditing

Finally, jfa facilitates statistical algorithm auditing. That is, the package implements functions for auditing algorithms, such as computing fairness metrics and testing the equality of parity metrics across protected groups.

• Algorithm auditing: Get started
• Algorithmic fairness