Data can sometimes reveal patterns that provide insights into complex issues, especially around diversity, equity, and inclusion. Conducting a pay equity analysis will help you assess whether pay differences exist between people in different demographic groups.
Besides incorporating acceptable variables like job title, grade, department, hire date, education attained, and salary, pay equity analysis incorporates data on unacceptable variables like gender, race, and age. Next-level information includes internal factors like disciplinary actions, performance reviews, and accumulated job expertise, plus external factors like industry-wide pay standards. Because that data resides in disparate sources and isn’t always numeric, gathering data is the most time-consuming step in the process.
Regression modeling can help determine whether pay differences exist between people in different demographic groups performing similar work. It reveals whether unacceptable variables (i.e., those representing potential causes of bias, like gender) may have a statistically significant effect in lowering an individual’s pay.
The model starts with an anonymized data set including the factors that should drive an employee’s compensation, as well as the potential causes of bias that shouldn’t matter. Exploratory data analysis (EDA) highlights correlations among variables and identifies any errors in the data set. A linear regression tool then fits a model to the data set and points to significant and insignificant factors. The results show the company any factors that don’t fit the model. These factors can then be addressed with a compensation expert.
With Alteryx, you can:
- Pull in employee demographic data directly from HR platforms like Workday
- Use Alteryx Intelligence Suite to perform Exploratory Data Analysis (EDA) to uncover outliers in your data, find correlations, and identify errors
- Develop a regression model right in an Alteryx Designer workflow that reveals meaningful patterns in data that impact pay equity