8-2 Hypothesis Test Basics

To support (or not) a claim about a mean, proportion or standard deviation statisticians use a technique called hypothesis tests.

The steps may seem overwhelming at first but they really can be mastered with some practice.

First we need to learn all the components of a hypothesis test. Those components include:

1. Use the original problem statement to formulate three formal mathematical inequalities called the original claim, the null hypothesis and the alternate hypothesis.

2. Use sample data and the level of significance to compute a test statistic and critical region. An equally valid technique is to compute the p-value.

3. Use those results to reject (or not) the null hypothesis and draft a sentence or two describing your decision regarding the original claim based on the data and your statistical analysis.

Please also check page 381 Figure 8-8 (p-value method) and 8-9 (traditional method).

You do not need to learn both methods since they provide identical results.

Most students seem to prefer the traditional method since it relies on critical values and we have been introduced to critical values when working with confidence intervals but I will accept your work using either method.

Technique to be learned: Set up all the components of a formal hypothesis test regarding a claim about a parameter (μ , p or σ) of a single population.