Statistics and Population Values
Introduction
Welcome to Drawing Conclusions from a Sample, the fourth course in this learning path! In the first three courses, you built a strong foundation: identifying populations and samples, choosing trustworthy samples, and computing summary measures like the mean, median, and MAD. Now we take the next step and ask the really exciting question: what can a sample actually tell us about the population it came from?
This first lesson focuses on a simple but powerful distinction: the difference between a value we compute from a sample and the true value for the entire population.
The Gap Between What We Measure and What We Want to Know
Imagine a coffee-shop owner who wants to know the average amount every customer spends per visit. Measuring every single transaction over the life of the business is not realistic, so instead she looks at a random sample of 50 recent receipts and computes the mean. That mean is a real, concrete number, but it is based on only 50 transactions, not all of them.
This small scenario highlights a gap that shows up in almost every real-world use of data. We have numbers from a sample, but we want to know about the whole population. Keeping those two things separate in our minds is the key idea of this lesson.
Population Values: The Numbers We Rarely Know
A population value is the result we would get if we could measure every single member of the population and then do our calculation. For example, if the coffee-shop owner could average all customer transactions ever, that result would be the population mean.
Population values are sometimes called population parameters. They are fixed and definite: there is one true average spending amount across all customers, whether or not we ever learn what it is. In practice, population values are almost always unknown because collecting data from an entire population is too costly, too slow, or simply impossible.
Sample Statistics: The Numbers We Actually Compute
A sample statistic is a value calculated from the data in a sample. As you may recall from the previous course, we already practiced computing several of these: the sample mean, the sample median, the range, and the mean absolute deviation. Each of those calculations used a subset of the population, not the whole thing.
Returning to the coffee-shop example, the mean of those 50 receipts is a sample statistic. It is a single number we can write down and use, and it is based entirely on the data we collected.
