In our earlier topic on Better Understanding Continuous Distributions and Stochastic Modeling via the Normal Distribution, we set the stage for / got half way to some important and practical stuff on estimating population parameters (e.g. when we used the normal case to understand what’s really happening when we talk about continuous distributions).
And then in Better Understanding Statistical Inference by Estimating the Mean and its Confidence Invervals we broadened our understanding of what it means to express the precision of an estimating parameter from standard errors (which imply a 68.26% level of confidence) to confidence intervals for any level of confidence we might want.
In this topic you will use what you learned in those topics to learn about something called hypothesis testing.
As you’ll see, hypothesis testing is another task of inferential statistics in which we test assertions or claims made about population parameters using sample statistics. And it’s super practical. In fact, we’re going to learn about it by looking at the following pretty authentic examples you’ll run into in the geomatics context:
- A case where you suspect the bias in an EDM instrument is no longer what was published
- A case where you’ve measured something repeatedly and suspect a gross error in one or more of your measurements
- A case where you measure your way around a loop that should close and want to test the observed misclosure for possible gross errors