Answer: performing cross-validation.

## Explanation **Data snooping** refers to the practice of extensively searching through data to find statistically significant patterns that may be spurious. **Mitigation Methods:** **A. Cross-validation ✓** - **Most effective**: Tests model performance on out-of-sample data - **How it works**: Divides data into training and validation sets - **Benefits**: - Reduces overfitting - Provides realistic performance estimates - Helps identify spurious relationships **B. Point-in-time data** - **Purpose**: Prevents look-ahead bias - **Limitation**: Doesn't address data snooping directly - **Use case**: Ensures only information available at decision time is used **C. Revised macroeconomic data** - **Purpose**: Provides more accurate historical data - **Limitation**: May introduce look-ahead bias if not handled properly - **Not a solution**: Can actually facilitate data snooping by providing cleaner data to mine **Why Cross-validation is Best:** - Directly addresses the core problem of overfitting - Provides objective performance metrics - Standard practice in statistical modeling and machine learning - Helps distinguish between genuine patterns and random noise **Conclusion:** Cross-validation is the most direct and effective method for mitigating data snooping.

Author: LeetQuiz Editorial Team