0.9 Ch. 9: hypothesis testing of single mean and single proportion

Hypothesis testing is done constantly in business, education, and medicine to name just a few areas. To perform a hypothesis test, you set up two contradictory hypotheses and use data to support one of them. Introduce the students to hypothesis testing by an example. Use a table to show the outcomes. Use $H_o$ as the null hypothesis and $H_a$ as the alternate hypothesis. Go over the language "reject $H_o$ " and "do not reject $H_a$ ".

Conducting a hypothesis test

Sometimes the data favors neither hypothesis. In this case, we say the test is inconclusive.

A hypothesis test may be left-tailed, right-tailed or two-tailed. What the test is concerned with generally determines what type of test is being done.

Associated with the null hypothesis is a pre-conceived $\alpha$ . $\alpha =P\text{(Type I error)}$ .Students sometimes have a difficult time when there is no pre-conceived $\alpha$ . We use $\alpha =0\text{.}\text{05}$ if there is none.

The data is used to calculate the p-value. The $p$ -value is the probability that the information (data) will happen purely by chance when the null hypothesis is true. If we reject the null hypothesis, then we believe the information did not happen purely by chance with the current null hypothesis. Therefore, we believe that the null hypothesis is not true.

The decision (to reject or not reject) is based on whether $\alpha >p$ -value or -value.

The example in the book concerning Jeffrey, an eight-year old swimmer, is a good first example to do with the class. They can follow along in the book and then complete the problem that follows (bench press problem). By filling in the blanks, they are led through the steps of hypothesis testing.

In the beginning, the students have the most difficulty in determining which test to use (test of a single mean - normal or Student-t or test a binomial proportion) and the type (left-, right-, or two-tailed). We do several examples (usually we choose some homework problems) in class with the students. If a single mean Student-t is done, the assumption is that the population from which the data is taken is normal. In reality, this would have to be shown to be true.

Here is a series of solution sheets that can be copied and used by the students to do the hypothesis testing problems. A solution sheet makes it clearer to the student what the steps to the tests are.

Go over the solution for "Fido's Fleas", a binomial proportion hypothesis testing problem written as a poem. The problem is at the end of the text portion of the chapter. The solution on a solution sheet follows the poem.

If you use the TI-83/84 series, there are functions to perform the different hypotheses tests. They can be found in STAT TESTS. Z-Test (normal test) does a test of a single mean when the population standard deviation is known; T-test (Student-t test) does a test of a single mean when the population standard deviation is not known; 1-PropZTest (normal test) does a test of a single proportion. The examples in the book contain TI-83/84 calculator instructions, in detail.

Assign practice

Assign homework

Assign projects