Overview
Employee turnover represents one of the larger Human Resources related costs. Because there are costs to recruiting and retaining employees, high turnover indicate that those expenditures are ineffective and essentially wasted. Identifying where the turnover problems are in an organization can help contain costs and optimize the budget.
Measures
{% Separation \; Rate = \frac{Number \; Employees \; who \; Left}{Number \; of \; Employees} %}
{% New \; Separation \; Rate = \frac{Number \; New \; Employees \; who \; Left}{Number \; of \; New \; Employees} %}
Hypothesis Tests
Once a measure of turnover has been computed, the next question that arises, is whether that turnover rate is high or low. This is done by comparing the rate to the rate of other companies or departments. Even though the samples will differ, it may be instructive to run a test to determine if it is likely that the two statistics could have come from populations drawn from identical distributions, that is, whether the difference between the statistics is due to chance, or from a material difference in the distributions.
- Comparing Two Samples - compares only two samples to determine if they come from the same distribution. This can be done to compare the turnover rate of one department from the rest of the company, for example.
- Analysis of Variance - ANOVA - is used to analyze more than two samples. In the case of turnover, it can be used to analyze the turnover rates of multiple departments to identify if there are outliers.