Using the same groups of four students each, formulate a total of five questions on single and multiple regression. Do not attempt questions (yet) on logistic regression or mixed models. Follow the guidelines for the questions on ANOVA.
Due: Apr. 15, 2011.
Formulate some "quiz" questions on ANOVA -- single ANOVA, factorial ANOVA and repreated measures. The questions should be in English and suitable for undergraduates. Avoid the error of asking only difficult questions, and try to focus on information that is important for students wishing to apply anova.
It would be most useful if the questions were formulated in markup language. Here's an example of questions (in Dutch) in yaml (yet another markup language). I would use them in that case!
Some ideas: make sure that students ask the right questions when they see ANOVA results reported on (and that they can criticize poor reporting), make sure they know about preconditions for application, test whether they can formulate null hypotheses and alternatives, and test whether they can cull the important information from SPSS output (or output from another program).
Perhaps the best way to do this is in groups, where at least one member of the group can speak Dutch and at least one has seen and used .html