Solo 3 de 20: Evaluating Fare Evasion Intervention in Bogotá's TransMilenio

The project creates a pre- and post-treatment design to evaluate the impact of the "Solo 3 de 20" informational campaign. The campaign aims to reduce fare evasion in Bogotá's TransMilenio public transportation system by educating passengers about the prevalence of evasion rates.

Table of Contents

1. Overview
2. Installation
3. Usage
4. Project Structure
5. Dependencies
6. Contact

Overview

• Goal: Evaluate the impact of an educational campaign on reducing fare evasion at selected TransMilenio stations.
• Baseline Fare Evasion Rates:
  • Treatment Stations (Troncal 26): 14.58%
  • Control Stations (Troncal 80): 13.04%
• Expected Treatment Effect: A 2% reduction in fare evasion at treatment stations.
• Methodology: Pre-test and post-test design comparing treatment and control stations.

Installation

Follow these steps to run the project on your machine:

Clone the Repository

git clone https://github.com/yourusername/Solo3de20.git
cd Solo3de20

Install Required R Packages

Ensure the required R libraries are installed:

install.packages(c("DeclareDesign", "estimatr", "fabricatr"))

Usage

Run the following steps to execute the analysis:

1. Open R or RStudio.
2. Load the main analysis script:

   source("fare_evasion_experiment.R")

3. Inspect results in the console, including simulated data and diagnostic outcomes.

Output Example

The script outputs simulated fare evasion data before and after the intervention and calculates:

• Change Score: Difference in fare evasion rates (Post-test - Pre-test).
• Post-Test Only: Treatment effect measured using post-treatment rates.

Project Structure

The project directory is organized as follows:

Solo3de20/
├── data/                    # Placeholder for input datasets (if needed)
├── results/                 # Placeholder for generated results
├── fare_evasion_experiment.R # Main R script for running the analysis
├── README.md                # Project documentation (this file)
└── LICENSE                  # License information

Dependencies

The project requires the following R packages:

• DeclareDesign: For defining the experimental design.
• estimatr: For robust linear regression.
• fabricatr: For creating and manipulating data.

Contact

For questions or feedback, please contact:

• Author: Juan Diego López Rodríguez
• Email: jdlr@alumni.upenn.edu
• GitHub: [https://github.com/jdlr9]

Thank you for exploring the impact of "Solo 3 de 20"!