Smart Cities, Smart Citizens: A Reproducible Simulation Study with Empirical Benchmarking of AI Infrastructure, Behavioral Nudges, and Citizen Satisfaction in Municipal Management

This record contains the replication and benchmarking materials for the article "Smart Cities, Smart Citizens: A Reproducible Simulation Study with Empirical Benchmarking of AI Infrastructure, Behavioral Nudges, and Citizen Satisfaction in Municipal Management". It includes the canonical synthetic dataset used in the manuscript, Python scripts for exact dataset regeneration, statistical analysis, and a Monte Carlo evaluation of the estimation strategy, a dependency file, a reproducibility report, a basic Jamovi project file, and a structured non-sensitive summary of the Santander municipal benchmarking case. The Monte Carlo experiment (10,000 replications) characterises coefficient recovery, statistical power, Type I error, and confidence-interval coverage under the known data-generating process, and shows that the estimator is unbiased and well calibrated. Additional scripts provide a sample-size/power analysis and a sensitivity analysis under a realistic (assumed) correlation structure among the predictors. The materials support transparent review, practical computational reproducibility, and inspection of the empirical benchmarking component. The Santander case is used as a light external plausibility and policy-relevance benchmark, not as a full causal validation dataset.