Exercise

Comparing randomization CIs and t-based CIs

As with hypothesis testing, if technical conditions hold (technical conditions are discussed more in the next chapter), the CI created for the slope parameter in the t-distribution setting should be in line with the CI created using bootstrapping. Create a CI for the slope parameter and compare it to the one created using the bootstrap percentile interval from the previous chapter.

Note that the bootstrap and t-intervals will not be exactly the same because they use different computational steps to arrive at the interval estimate.

Instructions 1/2

undefined XP

1

2

Calculate the confidence level and the lower and upper percentile cutoffs from alpha.