Transcript for:
Random Sampling in Statistics

vocabulary so in statistics when we say we have a random sample that's when everyone in the population has a chance to be in your sample so everyone in the population could be in your sample I'm still taking a sample though I'm still collecting data from let's say 300 people but every one of the millions of people in my population had a chance to be one of those 300 does that make sense so you're not it's not a census you're not collecting data from everybody you're still collecting a sample but everyone in the population has a chance to be in your sample and oftentimes we'd like it to be an equal chance though that doesn't always happen but but that's the idea trying to give everyone in your population a chance to be in your sample and that's what we've heard to as random okay so by the way walking down the mall and bumping into somebody and asking them a question is not random that's random the way you think but that's not random in statistics because not every all the millions of people in your population had a chance to bump into you at the mall so who you bump into is not random in statistics random means you made sure that every one of in your population had a chance to be included in your sample now that's very difficult actually so let's kind of look at it so the go-to random sample the most common collecting data method for people that know statistics is called the simple random sample simple random sample sometimes referred to as SRS this means you're selecting individual people or objects randomly okay that means again everyone in the population had a chance to be chosen now this usually means you have to somehow number your population in a lot of ways you have to have access to into your entire population and usually you give each person or object in your population a number so this comes into play where we use a lot of random number generators nowadays with in computer programs do random number generators the old days before computers you'd always see statisticians would have these huge books on their desks and and it would just be a book of random numbers and they would they would flip the book open whatever page it fell on they would they would just start writing those numbers that came up that's who they would pick but you have to sort of be able to give everyone in your population a number and that becomes that can be really difficult whereas the program examples are not easy but it's the the effort is worth it because this gives you a very very good way of collecting data that sort of minimizes bias simple random samples tend to not have very much bias in them if they really are a true simple random sample from the population in fact I kind of think of this is the second best if you can't do a census simple random sample that would be your second best now so let me give you an example suppose we have again the population of all the students at my college well I get every one every student at my college has a student ID number right so what I can do is I could just go ahead and and have a computer randomly selects do 90 numbers okay and if a computer randomly select student ID numbers it will have every possible student ID number will be available so again it will be a random sample now by the way just one one quick note don't think that you can pick randomly with your brain I always get somebody that says oh I can pick randomly I'll just I'll just go like this on the page and start pointing at things you know that's not random random means everyone in your population has to have an equal chance of being chosen so or at least a chance well in a simple random sample usually they do have an equal chance but but the one thing about it is that you're so in this case we would want the computer to pick randomly or like a random number generator or something like that okay another one let's suppose we're looking at the whole US population or any population of a government so of a country the population of their country so you could enter in the US we all have social security numbers so we could you can have a computer or randomly select Social Security numbers and then the government could sort of track down those people to get information from that way everybody has it has a chance again using a computer-generated random sample again having a computer randomly select numbers is sort of a very common technique for simple random samples by the way another way you can do it actually in modern days nowadays too you can have if you have access to a column that has maybe all the names of the people in your population this is very common with businesses like you have a listing of every single employee and your in your company which might be thousands and thousands of people in your population and suppose you can't get data from all of them you would just want to take a random sample so what they'll do is you can actually have a computer randomly select cells from the column so you can have a computer randomly select some names from this column of data so that's that's another way you could do it with the computer at least that now we know everybody has an equal chance of being chosen again don't think that you can choose randomly you can you have to use some kind of random number generator you could also do this thing where you take all the names of the people in your company put them in a big box or basket and then reach in and pick up pick out a name you know make sure all the all the all the pieces of paper about the same size and you reach in and grab pieces of paper like lotto style so that that would be a way of getting a random sample okay but a simple random sample is very good that that tends to minimize bias and is our our best sampling technique in terms of collecting data okay so we'll continue the discussion I'm going to go over a few more ways of collecting data in our next video this is Matt to show and intro stats