PDDModel is a statistical simulation model developed to evaluate the effectiveness of Police, Drugs, & Diversion (PDD) interventions. The primary outcomes evaluated are:
- Reoffending Rate: Measures whether a participant reoffends within a year of PDD intervention.
- Drug Treatment Rate: Assesses whether a participant enters into drug treatment within a year of PDD intervention.
- Hospitalisation Count: Counts the number of times a participant is hospitalized for a drug-related incident within a year of PDD intervention.
These outcomes were simulated using binomial distributions for reoffending and treatment rates, and a negative binomial distribution for hospitalisation count, thereby introducing between-individual variance.
The model simulates comparative scenarios between police forces implementing PDD interventions and those that do not. The parameters of the distributions are set on a per-force basis, allowing the introduction of between-force variance through baseline rates.
In the simulation process, both force data and individual-level data per force are randomized. The results are generated by averaging over thousands of such instances and fitting models for each instance and outcome.
Here's an overview of the main files and directories in this repository:
data/: Directory containing data files used in the simulations.figures/: Directory for storing generated figures from the simulations or data analysis.outputs/: Directory for storing the outputs of Rmd runs.simdata/: Directory containing simulated data.markdown_functions.R: R script file with functions for markdown.model_check_deps.R: R script file for checking dependencies in the model.model_config.R: R script file for model configuration.model_config_modification.R: R script file for modifying the model configuration.model_hospital_events.R: R script file for simulating hospital events in the model.model_participants.R: R script file for simulating participants in the model.model_police_forces.R: R script file for simulating police forces in the model.model_power_vs_police_force_variance.R: R script file for simulating power vs police force variance in the model.model_simulation.R: R script file containing functions for running the model simulation.model_test.R: R script file for testing the model.report.Rmd: R Markdown file for the project report.sim1.R,sim2.R,sim3.R,sim4.R,sim5.R: R script files for different simulation scenarios.sim1.Rmd,sim2.Rmd,sim3.Rmd,sim4.Rmd,sim5.Rmd: R Markdown files for different simulation scenarios.
Additional files in the repository include:
.gitattributes: Git attributes file so that the repository is not marked as mainly HTML by github..gitignore: Specifies intentionally untracked files that Git should ignore.README.md: This file, providing an overview of the repository.