Exercise

Power analysis for a t-test

Now we're going to have another look at the same example, drawn from our Olympic dataset (Weight difference between Sports). A random subset of 100 samples from each group in this data yields these boxplots.

Boxplots of body weights of Olympic athletes from different sports (100 sampled for each)

Knowing the effect size of the difference between these two groups (see the relevant exercise for a refresher), you need to work out what the odds are that your t-test will pick up this difference. Your data is contained in athletesample. pandas and statsmodels.stats.power are loaded into your workspace as pd and pwr.

Instructions

100 XP
  • Set the parameters as follows: an effect size of 0.42145, 100 samples from each group and a standard alpha value of 0.05.
  • Run and print the power analysis.