The regression model indicating how family income in 2010 is associated with par
ID: 3220235 • Letter: T
Question
The regression model indicating how family income in 2010 is associated with parents' age, parents' education, and family income in 2008 is displaye in the table below. (Note: You may assume linear effects of continuous variables when answering the following questions)
QUESTION 1: What are the net effects of parents' age, parents' eduction, and family income in 2008 on family income in 2010? (In othe rwords, describe the effects of each, holding the others constant.) Also discuss statistical significance of each predictor.
QUESTION 2: Which of the 3 predictors (parents' age, parents' education, and family income in 2008) has the strongest effect on famliy income in 2010? Please also provide a brief description why.
Source Number of obs 14,979 df SS MS F (3 14975) 4413.98 Model l 1.9713e +11 3 6.5710e +10 Prob F Residual 2.2293e-11 14,975 14886791.5 R-squared 0.4693 Adj R-squared 0.4692 Total I 4. 2006e-11 14,978 28045119.6 Root MSE 3858.3 faminc10 l coef. Std. Err. t P It C95% Conf. Interval par age I 56.74059 4.423846 12.83 0.000 48.06931 65.41 187 par educ l 339.5092 12.69315 26.75 0.000 314.6291 364.3894 faminc08 I 5381605 5258411 006285 85.63 0.000 5504799 cons 4374.184 215.1895 20.33 0.000 4795.981 3952.386Explanation / Answer
SOLUTION 1: If the variables, parents education and family income are fixed, then, for each change of 1 unit in parents age, there will be an increase of 56.74 units in the family income for the year 2010.
If the variables, parents age and family income are fixed, then, for each change of 1 unit in parents education,there will be an increase of 339.51 units in the family income for the year 2010.
If the variables, parents age and parents education are fixed, then, for each chage of 1 unit in family income_2008, there will be an increase of 0.5382 units in the family income for the year 2010.
Since, p-values, for all the three independent variables is less than 0.05, we can conclude that all three independent variables are significant in predicting family income for year 2010.
SOLUTION 2: Since, the regression coefficient for parents education is 339.51, as we know earlier, for each change of 1 unit in parents education,there will be an increase of 339.51 units in the family income for the year 2010 (when other two independent variables are fixed). Hence, parents education seems to have the strongest effect on family income for 2010.