Homework 2

Chapter 5, Question 5

A linear regression model of the solar radiation model parameter D on latitude (D = -1.366 + .0053 LATITUDE) explains very little of the variation in latitude (R-squared = 0.03). The ANOVA table suggests that the model does not significantly explain the data (p>0.4). The residual plot suggests that a curvilinear relationship may exist between the variables (see RWG p. 53).

Chapter 5, Question 6

A second-degree polynomial regression of D on latitude explains the relationship between these variables much better. The model explains 54% of the total variance in the data (R-squared = 0.54) and the coefficients of the model are significant at the p<.0001 level (the F-test in the ANOVA table gives a p-value of 7.8 E-05) allowing us to reject the null hypothesis (Beta1 = Beta2 = 0 (see RWG p.80)).

Chapter 5, Question 7

From eqn. 5.11 (RWG p. 152): Maximum latitude = -0.295/2(-0.003) = 49.2 From the regression model: Maximum D = -8.2167 + 0.2952(49.2) - 0.0030(49.2)^2 = -0.95