Open jacolb22 opened 4 months ago
Statistical inference can be tricky if we take the theoretical-based approach. It is much more insightful to explore the simulation-based approach we covered here.
Make sure that you look a simulation-based approaches on YouTube first. Once you have a pretty solid notion of what is going on when we do Null Hypothesis Testing, then you can consider the theoretical approach --which is the more traditional road (and for which you will find much more on YouTube).
I learned how to identify response and explanatory variables, including the "unit of observation" that the tidy dataset is built around. Additionally, I learned to think about what diagnostics I want to run on a dataset to make sure all values are valid. What is a valid value? What should I do if certain values are invalid? If I leave some values in, what are the consequences? I also learned to think about which modeling process I would use, like a t-test or linear regression, or others. This was the hardest part, asI hadn't seriously looked at some of these things since high school. Deriving the p-values would also be very difficult, again, because it has been so much time. Like the last lab, I will probably take to youtube to refresh my memory on some of these things, because it may be useful in the future.
@francojc