I.  Goal:

• To develop an understanding of hypothesis tests.

• Students will explore the components of hypothesis tests.
• Students will be able to distinguish between right-tailed, left-tailed, and two-tailed tests.
• Students will use tabulated sample data to conduct a hypothesis test.
• Students will use raw sample data to conduct a hypothesis test.
• Students will discover that different raw samples produce different results, sometimes.

• Computers with Fathom Dynamic Software by Key Curriculum Press (Ideally each student would have his/her own computer but groups of three per computer would work well.)
• A Fathom file to use that contains raw data of your choice.  I intend to use a file containing the semester grades of the students I have had in class the previous three semesters called SemesterGrades.ftm.
• Student Instruction Page

• Have you ever had a theory about the world around you that you wanted to test?  When people do research, they often need to back up their data with results.  Today we are going to learn how to test theories based on data.

1. Today we are going to explore hypothesis tests.
2. Students should follow instructions on the student instruction page.
3. This exploration is intended to be only an introduction to hypothesis tests.  The questions on the student page are designed to have students focus on ideas corresponding to hypothesis tests but not about how to actually perform them.  Some ideas that should be discussed after the activity include:
1. What pieces of information do you need to perform a hypothesis test?
2. What does the information in the alternative hypothesis correspond to?  (Your conjecture.)
3. What about the alternative hypothesis determines the placement of the red tails?
4. What does the red area under the curve correspond to in the results?
5. What kind of language is used when reporting the results of the hypothesis test?
6. What determined the outcome of your hypothesis test?

• This lesson began with a question of how to show other people the validity of our results.  Can you always be certain that your sample provides accurate results?  Why or why not?