Overall Notes

• There is one version of this assessment provided in the Teacher Resources Materials (TRM). Use this assessment as-is or use it as a model to create your own. 

• If you are making your own assessment, consider using the test bank and the ExamView software.

• Consider allowing students to use the four-step template from this lesson for a significance test. 

• Students will need the Chi-square independence applet (part of the Traditional Inference applets at www.statsmedic.com/applets to find the P-value.

 

Common Student Errors

• When writing hypotheses, students sometimes refer to the sample. Hypotheses should always be written about the population.

• When calculating expected counts for a two-way table, students will incorrectly calculate expected counts for the total row and/or total column. Expected counts should only be calculated for the individual values in a two-way table.

• Students will sometimes round expected counts to whole numbers. Because expected counts are just means, they should not be rounded. 

• When checking the Large Counts condition, students will incorrectly check that all the observed counts are at least 5. They should be checking that all the expected counts are at least 5. 

• Students will often forget to include context when writing a conclusion for a significance test. 

• Students will sometimes incorrectly "accept the null", when they should be "failing to reject the null".

• When rejecting the null hypothesis, students will say that we have "proved" the alternative hypothesis is true. We don't know for sure that the alternative is true (we could be making a Type I error), so students should be saying there is "convincing evidence for the alternative hypothesis".