Skip to main content
SearchLoginLogin or Signup

Adjusting for Measurement Error in Police Recorded Crime Counts Using Bayesian Statistics

Training materials for short course ‘Adjustment Methods for Data Quality Problems: Missing Data, Measurement Error, and Misclassification.’

Published onJan 02, 2024
Adjusting for Measurement Error in Police Recorded Crime Counts Using Bayesian Statistics
·

Abstract

In this workshop we will see how to adjust for different forms of measurement error affecting police recorded crime counts using ‘rjags’. More broadly, this workshop could also be of interest to researchers seeking to adjust for typical forms of measurement error seen in count and duration data, such as telescoping bias or other recall errors present in retrospective survey questions (see Pina-Sánchez et al., 2014). We are going to show how we can overcome these limitations using adjustments based on Bayesian statistics. The key idea behind this approach is to estimate simultaneously the outcome model of interest (which we use to explore the causes or consequences of crime) and the measurement error model (where we describe how well does police data reflect the true extent of crime). This is a practical exercise of the short course ‘Adjustment Methods for Data Quality Problems: Missing Data, Measurement Error, and Misclassification.’

Comments
0
comment
No comments here
Why not start the discussion?