Session Ready
Exercise

Bootstrapping confidence intervals

Having bootstrapped the distribution of the female-effect coefficient in the last exercise, you can now use it to estimate a confidence interval. It will allow you to make the following assessment about your data: "Given the uncertainty from imputation, we are 95% sure that the female-effect on earnings is between a and b", where a and b are the lower and upper bounds of the interval.

In the last exercise, you have run bootstrapping with R = 50 replicates. In most applications, however, this is not enough. In this exercise, you can use boot_results that were prepared for you using 1000 replicates. First, you will look at the bootstrapped distribution to see if it looks normal. If so, you can then rely on the normal distribution to calculate the confidence interval.

Instructions 1/3
undefined XP
  • 1
  • 2
  • 3
  • plot() and print() the bootstrapping results, boot_results.