1. Paying new hires fairly
When a company has a hard time filling a certain position, one option is to increase the offered salary to make the job more attractive.
2. Paying new hires fairly
If the company does not increase the salary of current employees at the same rate that the market rate for new hires increases, the current employees will be paid less than the new hires. This is true even in jobs where the additional experience with the company should make the current employees more valuable. When the long-tenured employees learn that they are underpaid, they may leave the company. Or, they may stay put, and feel resentful toward their employer; or they may even file a lawsuit if they believe the difference in pay is due to discrimination.
3. Checking wage equality
The solution is to do routine checks of both current employee and new hire salaries. This kind of analysis is usually done by compensation professionals, who are trained to understand nuances of market rates and job classifications. However, the general principles in this example can be used to test many kinds of differences between employee groups.
4. The data
In this chapter, you'll find out whether new hires are getting paid more than current employees. The dataset is based on past datasets in this course, with new_hire and job_level variables added in. As always, this is synthetic data generated for educational purposes.
5. Introducing broom::tidy()
As you analyze the pay dataset, you'll use the tidy() function from the broom package. tidy() takes the messy output of a statistical test, such as a t-test, chi-squared test, or regression, and returns the output in a nicely-formatted data frame.
Notice that the numbers in both the tidy and the regular output are the same, but the tidy version has column names. This means you can access the p-value of the test directly and use it somewhere else. You can use the pull() function to output the p-value from a tidied statistical test.
6. Let's practice!
Now, go take a look at the pay data!