(4)Multiple Correlation.Agresti & Finlay p. 358, problem 11.10 (5 points each, a and b)

For recent UN data for several nations, a regression of carbon dioxide use (CO2, as a measure of air pollution) on gross domestic product (GDP) has a correlation of 0.786. With life expectancy as a second explanatory variable, the multiple correlation is 0.787.

(a)Explain how to interpret the multiple correlation.

(b)For predicting CO2, did it help much to add life expectancy to the model? Does this mean that life expectancy is very weakly correlated with CO2? Explain.

(5)Multiple Regression.Agresti & Finlay p. 359, problem 11.14 (5 points each, a-c)

Table 11.15 comes from a regression analysis of y = number of children in family, = mother's educational level in years (MEDUC), and = father's socioeconomic status (FSES), for a random sample of 49 college students at Texas A&M University.

(a)Write the prediction equation. Interpret parameter estimates.

(b)Report SSE. Use it to explain the least sq

uares property of this prediction equation.

Explain why it is not possible that = 0.40.

Table 11.15

Sum of Squares

Regression 31.8

Residual 199.3

b

(Constant) 5.25

MEDUC -0.24

FSES 0.02