# SRM Checks

## SRM - Sample ratio mismatch

Sample ratio mismatch (SRM) occurs when the observed allocation of **unique** users between test groups differs from the expected allocation or "split" of the test. For additional background, go to the [Statsig blog post on sample ratio mismatch](https://www.statsig.com/blog/sample-ratio-mismatch).

SRM is a signal of unknown bias in the test. Because diagnosing the cause of the imbalance isn't always possible, it isn't always easy to determine how much the bias affects results.

## SRM checks

Statsig runs SRM checks on all experiments and feature gates as part of Health Checks (go to [Health Checks documentation](https://docs.statsig.com/experiments/monitor) for details). Statsig uses a Chi-squared test to determine whether the user split between groups indicates a sample ratio mismatch.

{% figure %}
![SRM health check results interface](https://docs.statsig.com/images/experiments/srm-checks-health.png)
{% /figure %}

Statsig automatically analyzes data by common dimensions logged by the Statsig SDK to identify potential drivers of SRM. These dimensions include sdk\_type, sdk\_version, reason, is\_bot, browser\_name, browser\_version, os, os\_version, and region.

{% figure %}
![SRM dimension analysis breakdown](https://docs.statsig.com/images/experiments/srm-checks-dimensions.png)
{% /figure %}