Many textbooks now come with an online component for self-assessment, study, and quizzes. Teachers are experimenting with the new software components, and wondering whether they have any effect on students' final grades. A teacher who has three sections of Introductory Statistics is wondering about three variations of use of the software, as follows:
a. Assigning quizzes for marks
b. Creating chapter quizzes and recommending them, but not assigning any marks
c. Not creating any quizzes, but relying on the sample tests built in to the software (no marks assigned) There are 45 students in each of the classes being compared. The Data Analysis output for Anova: Single Factor is also shown in Exhibit 11.21 on the next page.
Is there evidence, at the 5% level of significance, that differences in the use of the online software are associated with differences in final grades? Why should we be cautious in interpreting the results in this case? You should already have checked conditions in Develop Your Skills 11.1, Exercise 5.

Anova: Single Factor SUMMARY Groups Count Sum Average Variance Assigned Quzor Marks Quizzes, No Marks Sample Tests Only 45 45 2551 56.68889 226.5828 45 2433 54.06667 218.0182 ANOVA Source of Variation MS P value Between Groups Within Groups 6666.844 28894.89 2 3333.422 15.22801 1.12E-06 132 218.9007 Total 35561.73 134