Finding similar employees
Finding comparable employees may be helpful if, for example, you want to compare the compensation of an individual employee to other similar employees. The system provides a powerful feature to help you identify similar employee and their compensation, and prevent legal action from employees due to unfair compensation.
In this article we will cover two main topics:
Using the Find Similar Employees feature
To use the Find Similar Employees feature, proceed as follows:
Start by running an analysis and then scroll down to the employee list.
To find similar employees to a specific individual (e.g., an employee with ID number 313), navigate to the Selection column on the far left of the employee list. Click the icon next to the employee's row and select Find Similar Employees, as shown on the following picture:
Find Similar Employees feature
In the Record limit box that appears, specify the number, or percentage, of employees you want to compare. You can also set a minimum number for each gender, to be included in the comparison. Once set, click Apply, as shown in the following illustration:
Record limit box
4. After clicking Apply, you’ll see that a filter has been applied, and the employee list will show the selected number of employees, plus the employee you started with (the one you're comparing against). The employee overview graph will also be updated to display only these selected employees.
5. A column with a Similarity Score has now been added for each employee in the employee list, ranking them from the most similar to the least similar based on the selected employee’s characteristics, on a scale of 0 to 100, where 100 is the most similar and 0 the least, as shown on the following picture:
Similarity score column
Now you can review the most similar employees and compare their compensation or other relevant data points.
Understanding how the similarity score is computed
The similarity score is used to identify comparable employees within the dataset. The score is between 0 and 100 and is computed based on two inputs from the regression model:
The regression weighting (or coefficient) of each variable in the model. It can be found in Regression Results section and Regression Coefficient column , within your Pay Equity analysis (at the bottom of the page).
-
Computing the distances: between variable values for each employees compared to the selected employee. .
For numerical variables the distance is the absolute difference of the numbers weighted by their regression coefficient.
For categorical variables the distance is the absolute difference between the regression weight of each variable value.
To ensure the similarity score is distributed and meaningful within the dataset, the distance is compared to the maximum occurring distance in the employee dataset:
If the distance between all variables is 0, the score will be 100.
If the distance between employees is maxed on all variables, the score will be 0.
Generally, we do not share formulas, but we hope this gives you a deeper understanding of the basis for computing.
Finally, you can click Reset All Filters and Pins to return to the full employee dataset. You will also be able to view the similarity scores for all employees in relation to the selected employee. If you want to end the comparison, click Clear Similarity Scores to remove the scores and reset the view.