Lecture 6. Resampling Methods

Date: 2023-03-07

This lecture goes into more details about Resampling. In order to keep relevant notes together, I went back and added those extra materails to the previous lecture instead of putting them here.

---

However, here is an interesting question: Which sampling strategy leads to independent samples? Sampling with replacement or without replacement?

Answer: Sampling with replacement leads to independent samples.

When you sample with replacement, after drawing each sample, you "replace" it back into the population. This means that the chance of drawing any specific sample remains constant across all draws, making each draw independent of the others.

On the other hand, when you sample without replacement, once a sample is drawn, it's not returned to the population. This means the probability of drawing any other specific sample changes with each draw, making the draws dependent on each other.