Case Studies
============

Comprehensive case studies demonstrating Causal Agent in real-world scenarios across different domains. Each case study includes complete analysis workflows, from data preparation to result interpretation.

.. toctree::
   :maxdepth: 2

   education_policy_analysis
   healthcare_treatment_effects
   marketing_campaign_evaluation
   economic_policy_impact
   technology_product_features

Case Study Structure
--------------------

Each case study follows a consistent structure:

1. **Problem Statement**: Clear research question and context
2. **Data Description**: Dataset characteristics and variables
3. **Method Selection**: Why specific methods were chosen
4. **Analysis Workflow**: Step-by-step implementation
5. **Results Interpretation**: What the findings mean
6. **Limitations and Caveats**: Honest assessment of constraints
7. **Alternative Approaches**: Other methods that could be used

Featured Case Studies
---------------------

**Education Policy Analysis: Learning Mindset Intervention**
   Comprehensive walkthrough of how Causal Agent analyzes a randomized educational intervention, showing complete agent decision-making process.
   
   * **Domain**: Education research
   * **Method**: Linear Regression with Covariates
   * **Dataset**: Growth mindset intervention study (12,490 students)
   * **Key Learning**: Decision tree navigation for experimental data, precision optimization with covariates
   * **Agent Focus**: How randomization simplifies method selection, robustness checking across specifications

**Healthcare Treatment Effects: Hospital Treatment Analysis**
   Detailed analysis of observational healthcare data showing how Causal Agent handles selection bias and validates propensity score methods.
   
   * **Domain**: Medical research
   * **Method**: Propensity Score Matching
   * **Dataset**: Hospital patient records (3,504 patients)
   * **Key Learning**: Selection bias detection, propensity score model development, balance assessment
   * **Agent Focus**: Method exclusion logic, comprehensive robustness analysis, clinical interpretation

**Economic Policy Impact: Minimum Wage Analysis**
   Complete regression discontinuity analysis showing how Causal Agent identifies and exploits policy discontinuities for causal identification.
   
   * **Domain**: Labor economics
   * **Method**: Regression Discontinuity Design
   * **Dataset**: State-level employment data (2,847 county-month observations)
   * **Key Learning**: Running variable detection, discontinuity validation, bandwidth selection
   * **Agent Focus**: Geographic discontinuity identification, assumption testing, policy interpretation

**Marketing Campaign Evaluation: Instrumental Variables Analysis**
   Comprehensive IV analysis demonstrating how Causal Agent identifies valid instruments and handles endogeneity in marketing data.
   
   * **Domain**: Marketing analytics
   * **Method**: Instrumental Variables
   * **Dataset**: Customer behavior and advertising data (8,742 customers)
   * **Key Learning**: Instrument validation, endogeneity testing, continuous treatment effects
   * **Agent Focus**: Instrument detection and validation, first-stage analysis, business ROI interpretation

**Technology Product Features: A/B Testing Analysis**
   Complete A/B test analysis showing how Causal Agent optimizes experimental analysis for precision and business decision-making.
   
   * **Domain**: Product analytics
   * **Method**: Linear Regression with Covariates
   * **Dataset**: Mobile app engagement experiment (15,847 users)
   * **Key Learning**: Randomization validation, specification optimization, subgroup analysis
   * **Agent Focus**: Precision optimization, business significance testing, long-term monitoring strategy

Learning Objectives
-------------------

After completing these case studies, you will be able to:

* Identify appropriate causal inference methods for different research contexts
* Understand how to prepare and validate data for causal analysis
* Implement complete analysis workflows using Causal Agent
* Interpret results correctly and communicate findings effectively
* Recognize limitations and potential biases in causal analyses
* Apply best practices for reproducible causal inference research

Datasets and Resources
----------------------

* **Downloadable Datasets**: All case study data available for practice
* **Code Repository**: Complete analysis scripts and notebooks
* **Supplementary Materials**: Additional readings and references
* **Discussion Forums**: Community discussion for each case study