10.2 Simulation of normally-distributed data
For demonstration purposes, let us start with setting up normally distributed data. This is ideal if you do not yet have a very good understanding of your data. You may have a vague idea about the performance of your test, but rely more on general ideas of what you would have to expect from certain outcomes the results of the data collection may have.
If you have already collected observational data, there are specific ways to work with your collected data sample. For instance, there may be simulations you could run using different distribution assumptions that are closer to the data that you may already have collected. Below, I will give a hint where you maybe want go other ways in
R in that direction and I may even add that in a future post.