Exercise

Performance ratings: accounting for job levels

Now that you're warmed up on logistic regression, you can use it to statistically test what you saw in an earlier graph - are women less likely to be labeled a high performer, even when taking differences in job level into account? You can test this by adding job_level into the glm() formula and looking at the new p-value. The process is the same as adding a new variable to a linear regression.

Instructions 1/3

undefined XP
    1
    2
    3
  • Using glm(), run a multiple logistic regression to test whether men are statistically more likely to be labeled high performers in this dataset when job level is also taken into account. Assign the result to logistic_multiple.