Continuing from Exercise 14, the coach responds to the players by claiming that shooting accuracy is more important than time on the ice. He adds Shoot% (% of shots on goal) to the model.

a) The coefficient of TOI/G in this model is quite different from the TOI/G coefficient in the previous model of Exercise 14. Why the change?
b) Find a 95% confidence interval for the coefficient of TOI/G. How does it compare with the one you found in Exercise 14?
c) The t-ratio associated with the TOI/G coefficient is now much smaller and the corresponding P-value much larger.

 

Data From Exercise 14

The players on the team in Exercise 2 point out to the coach that they can’t shoot if they are not on the ice. They add the variable Time OnIce/Game (TOI/G) (in minutes per game) to the regression:

Response variable is: Goals R squared = 95.7% s = 0.8850 with 65 -4 = 61 degrees of freedom t-ratio Variable Intercept Shots TOV/G Shoot% Coefficient -9.11025 SE(Coeff) 0.8057 0.111264 0.0057 -0.040008 0.0656 0.872339 0.0356 P-value -11.3