the last type of random sampling we're going to discuss is cluster sampling now cluster sampling seems uh quite similar to stratified sampling but it's a little different okay just like with stratified sampling we divide our population into groups in this case we call them clusters divide the population into groups called clusters and I say that we divide it but really in real life most the time um these groups are divided we get these clusters naturally okay and we work with these clusters for convenience okay once we have our clusters we want to randomly sample a few of the Clusters and Sample everyone from those clusters okay every member from those clusters so randomly select a few clusters and randomly sample everyone or every every member from those clusters okay now this is a little different than stratified sampling stratified sampling we also had groups we called them strata but in stratified sampling we sample a few members from every group with cluster sampling we sample every member from a few groups okay do you kind of see the difference so let's draw a picture okay say we have [Music] okay so these black circles are going to be our groups okay good enough maybe I'll just take out the last one okay in each group there are members of our population okay there are some numbers here there are some members here in this cluster there are some numbers in this cluster there are some members in this cluster there are some members in this cluster okay often these clusters are divided by timer space so that it's easier so maybe maybe all these people are in California and all these people are in Washington DC if we split them up then it'd be hard to interview just a few people here and just a few people here say you have to interview them in person you'd have to be in California for these interviews and Washington DC for these ones over here okay so that could be kind of difficult so often we use cluster sampling to simplify things um when uh the the different clusters or groups are are segmented in time or space or both okay so what do we do once we have our clusters we randomly select a few clusters say in this case we select two okay so I'm going to randomly choose uh say this cluster here and this cluster over here okay and we then sample every member in those clusters so this one this one this one this one this one this one this one this one this one this one this one this one this one and that class should get sampled everyone in that cluster gets sampled and this one this one this one this one this one this one this one this one this one this one this one gets handled every member in that cluster gets sampled and no one in the other no member in the other clusters get sampled at all okay that is cluster sampling so let's take a look at an example let's look at this a bit more carefully so let's get a picture here so say we have our group of students the kind of the same group of students we've been dealing with in the last few examples okay but this time they're sitting at individual tables okay the tables we could view as clusters instead of going about the whole room and selecting a uh students randomly it might be easier to take everyone from the table and then everyone at another table right that's cluster sampling so what we're going to do is we are going to use cluster sampling to randomly sample six students okay in this case that would mean we need to sample two clusters so the idea here is that the thing we're really picking isn't the students it's the Clusters and then the members that are in those clusters are the ones we sample okay but the thing that we're really focused on at first are the Clusters okay so we are going to uh so we need to randomly choose two clusters now in this case there are 10 tables the tables are our clusters so we need to choose two of the ten tables so let's pull out our calculator okay blow that up just a little more we'll use our sample function we want to any table from 1 to 10 should be in our group okay and we want to choose two of those tables so notice this sample function is choosing the tables not the students directly you click the evaluate button make it a little smaller so you could actually see there we go and the tables that were chosen are tables 3 and table five so this table and this table okay so those are the tables we've sampled and so the six students we have in our sample are Hubler fontecha legany Lundquist Cunningham and then Selmo they are the students that are in our clusters they are the students that we have sampled