next video: • video 6.1. sampling di...
prior video: • video 5.3. three kinds...
closed captioning text:
This class is going to focus on inferential statistics, which, in all cases, we are going to be observing a sample, and making inferences about populations. All of the tests that we are going to look at assume random sampling. "Random sampling" is when you have a population and the individuals in the population are chosen to be in your sample such that every individual in the population has the same chance as everybody else of being in your sample.
"Random sampling" is everyone in the population has the same chance, or probability, of being in your sample. If this is what the population looks ... I am just making it so each little dot is a person ... and we do some random sampling to get our sample. So this is the population. This is the sample. If we end up... So these are individuals in our sample. They are a subset of the population.
If it is a random sample, every one of these people has an equal probability as everybody else of ending up in our sample. They would just be completely, randomly picked from the group to be brought into our sample. So all of the inferential statistics that we talked about in this class assume that the population here ... we use random sampling to get our sample. Then what that allows us to do, is once we have our sample, we measure our sample, and if it was a random sample then we can be pretty confident in our inferences from our sample back to our population.
In research, there is a concept that is kind of like sampling but a little bit different, and it is important that you know the distinction. So, random sampling is if everyone in the population has an equal chance of being in your sample. There is this other concept called "random assignment", and this is relevant for experiments. Each person in your sample has the same probability--as everybody else in your sample--of being in your experimental conditions. So this is just for experiments.
Here, you start off with a sample. You have already done your sampling from the population,
and you already have a sample of individuals already ... before random assignment happens. So, theoretically, you do random sampling to get your sample, then once you have your sample then you could do this random assignment. So you could have your participants show up to your lab, you flip a coin, and if it is heads ... if it is heads, they are in your caffeine condition. If it is tails, then they are drinking decaffeinated coffee.
Alright, so each of these individuals has an equal chance as every other individual, of being in your two experimental conditions. In this middle process here, this would be the random assignment. You could literally flip a coin, and that would produce a random assignment of individuals to your two different conditions. This is done in experiments, and this is assumed for all different research. This class is going to assume this for all of the inferential statistics that we do. Random assignment is not at all a central feature of this class. It is a really big deal in research methods class, because experiments, all true experiments, have random assignment,
and that balances out the characteristics of people in the caffeine group with those in the decaf group so the only difference between these two groups is what you have done to them.
The reason that is so critical, is that these experiments are the only way that scientists can get strong evidence for a causal claim. All the other kind of research that we do we can not make strong causal claims. We can just say that two variables are associated with each other. But, again this experimental method is something more for a research methods class, and we won't focus on random assignment in this class, but random sampling is going to be a big deal.