The power company wants to develop a model to predict the amount of electricity used by a house. An engineer of a company proposed the following multiple regression model y = P + x1B1 + x2B2 + x3B3, where y = the amount of electricity (in specific units) used by a house per day, x1 = the heated area of a house in square meters, x2 = the mean outside temperature in degrees Celsius, x3 = the mean number of hours of sunlight per day. To estimate the multiple regression model, a random sample of 14 houses was selected from the houses in a certain area. The MINITAB multiple regression program was used, and the following computer output was obtained:
Predictor Coef SE Coef Constant 357.00 2235.00 x1 0.8082 0.3178 x2 -16.61 28.23 x3 -0.016 0.014 S = 267.7 R-Sq = 82.0% Analysis of variance Source DF SS Regression 3 3271175 Residual Error 10 716459 Total 13 3987635
T 0.16 5.87 -0.59 -1.143
MS 1090392 71646
F 15.22
a. In the context of the given problem, interpret the regression coefficients.
b. Do the data provide sufficient evidence to conclude at the 5% significance level that the model is useful in predicting the amount of electricity (in specific units) used by a house per day?
c. Do the data provide sufficient evidence to conclude at the 5% significance level that the amount of electricity (in specific units) used by a house per day and the mean outside temperature in degrees Celsius are negatively linearly related?