From the course: Data Literacy: Exploring and Describing Data

Unlock the full course today

Join today to access over 24,700 courses taught by industry experts.

Inference as describing populations

Inference as describing populations

- [Instructor] A population is everybody that you might be concerned about. And it might be that you have this wonderful, huge, diverse audience at least potentially out there waiting for you. On the other hand, your sample, the people you actually have data about, the experience that you've had might be a much smaller part of that. And this gets it to the issue of how much can you learn about the larger population when all you have is data from a smaller sample? And this is the problem of inference and attempting to describe populations that are only partially observed. Now, they're few things that go into this that make it possible to make that conceptual leap. Number one is sampling. You need to have representative sampling. That is the sample of people that you're dealing with needs to be representative a good standing, a good picture of the larger population if going to get anywhere, random samples or something like that. Even though conceptually, they're easy to describe…

Contents