Small area estimation

modeling

Small area estimation for crime analysis.

Author

David Buil-Gil

Published

November 11, 2021

According to David Buil-Gil, victimization surveys provide key information about crimes known and unknown to the police, and are the main source of data to analyse perceived safety and confidence in policing. These surveys, however, are only designed to allow aggregating responses and producing reliable direct estimates (i.e., weighted means or totals) at very large spatial scales, such as countries or states. Sample sizes are too small to produce direct estimates of adequate precision at the increasingly refined spatial scales of the criminology of place. Model-based small area estimation techniques may be used to improve the reliability of small area estimates produced from victimization surveys. Small area estimation techniques are applied to produce reliable estimates of parameters of interest (and their associated measures of error) for areas for which only small or zero sample sizes are available. This workshop will introduce theory and show a step-by-step exemplar study in R to illustrate the utility of small area estimation for crime and place research. You find materials here. And his presentation here

David Buil-Gil is a Lecturer in Quantitative Criminology at the Department of Criminology of the University of Manchester, and a core member of the Manchester Centre for Digital Trust and Society. His research areas cover geographic criminology, small area estimation applications in criminology, measurement error in criminological research, new methods for data collection, and open data.

Citation

For attribution, please cite this work as (Bull-Gil 2021)

References

Bull-Gil, D. 2021. NSC-R Workshop: Small Area Setimation for Crime Analysis.” NSCR. https://nscrweb.netlify.app/posts/2021-10-12-small_area_estimation/.