Sample Ratio Mismatch Calculator
Calculate and analyze sample ratio mismatches in your experimental data.
Calculate Sample Ratio Mismatch
Table of Contents
What is Sample Ratio Mismatch?
Sample Ratio Mismatch (SRM) occurs when the observed ratio of samples in different groups significantly differs from the expected ratio. This can indicate issues with randomization or data collection in experiments.
- Indicates potential randomization issues
- Can affect experiment validity
- Should be monitored in A/B tests
- Requires statistical testing
Detecting SRM
Chi-Square Test
Most common method
Z-Test
For large samples
Visual Inspection
Initial screening
Interpreting Results
Interpretation Guidelines
- p-value < α: Significant mismatch
- p-value ≥ α: No significant mismatch
- Consider sample size impact
- Check for systematic bias
Common Examples
Example 1 No Significant Mismatch
Expected: 0.5, Observed: 0.48, n=1000
Result: Not significant (p > 0.05)
Example 2 Significant Mismatch
Expected: 0.5, Observed: 0.35, n=1000
Result: Significant (p < 0.05)
Example 3 Small Sample Size
Expected: 0.5, Observed: 0.45, n=100
Result: Not significant (p > 0.05)