Bowling A student runs an experiment to test how different factors affect her bowling performance. She uses 3 levels for the weight of the ball (low, medium, and high) and 2 approaches (standing and walking), throwing 4 balls at each condition and choosing the conditions in random order. She counts the number of pins knocked down as the response variable.

1. What are the null and alternative hypotheses for the main effects?

2. How many degrees of freedom does each factor sum of squares have? How about the Error Sum of Squares?

3. Should she consider fitting an interaction term to the model? Why or why not? How many degrees of freedom would it take?