Chapter 6 Hypothesis Tests

Confidence intervals determine a range where our population mean resides given the characteristics of a sample and a desired level of confidence. Recall that the population parameter can be anywhere within the range dictated by a confidence interval.

Hypothesis testing is a similar inferential method, but it approaches the problem from the opposite direction.

1. You start with an unambiguous claim on the value of the population parameter. This claim is nonarbitrary, and dictated from either theory or a past observation.

2. You test to see if the sample statistics are consistent with the claim (or refute it)

The general idea is that you begin with some nonarbitrary statement on what value you believe (or do not believe) the population parameter to be. You then test if the characteristics of your sample suggest that it is likely or not that a population with your proposed parameter values generated a sample similar to the one you currently have.

If this seems a bit vague at the moment, it will hopefully be more concrete soon. The main thing to keep in mind is that hypothesis tests are quite simple and structured. Once you learn how to perform one hypothesis test - you can essentially perform them all. This chapter guides you through some basic steps that once mastered - you’ll have a powerful tool of statistical inference under your belt.