Now there are four types of sampling that I want to discuss. The first one is going to be simple random selection. So let's start off by talking about simple random sampling. That is where every member of a population has an equal likelihood of being chosen. So, I do have to have a master list and add a number to each individual. And then have some form of random number generator that can be done on paper, if possible. These days, of course, we just use computers to generate random numbers. And those random numbers, each of them from that master list has an equal likelihood of being chosen and would choose from a population. Then there's systematic sampling. Now there we're going to be very rigorous about it and choose our individuals specifically, but also from that master list. So how's this going to work? That is where we're just going to as a system just choose every individual equally spaced, every 10th individual, every 20th individual on the list. In the end, we've got to a little equation. We call it every kth, a lowercase k, every kth individual. And the sample size we're going to call lowercase i. And the population, uppercase N. So to choose the kth member we'll just be N divided by i. So if we look at a little example if we have a population of 10,000 and we want 1,000 from those. So 10,000 divided by 1,000 it's 10, so we're going to choose from that list exactly every 10th individual. Just to remember though that this can also introduce bias. Imagine you work in a clinic, and you're going to include in your data collection every 10th patient who comes into the clinic. But that might mean that it's a patient who comes in quite late in the day, if it's every 10th one, and you only see a small amount of patients everyday. And there might be a reason why those patients come in late. They might be sicker or elderly in some way, and it's difficult for them to come to a hospital, or to the clinic early in the morning. So you can introduce bias by this sort of sampling. Then there's the cluster random sampling. This is usually where we don't have a master list. We don't have a master list of the population, but at least they clustered somehow. So as an example I mention you wanted to do some research on healthcare workers at clinics in a city. You don't have a master list of those health care workers, but you do have a master list of those clinics. Now, those clinics can be chosen at random, and then every worker at that clinic would form part of a sample. Now lastly, there's the stratified random sampling. That's where there's a mutually exclusive trait in the population. Which immediately separates them and we can take participants from each of those groups. The usual example there is just gender, if you were to divide the population between male and female for instance. You could then start taking samples from each of those mutually exclusive groups. And that's usually done where we want to keep the group sizes equal in number.