Please explain what the MATLAB code is for the following questions: Source | SS df MS Number

Please explain what the MATLAB code is for the following questions:

 

Transcribed Image Text:

Source | SS df MS Number of obs 69,450 F (3, 69446) 3000.65 Model | 2280534.24 Residual | 17593273.6 ------ 3 760178.079 69,446 Prob > F = 0.0000 253.337465 R-squared = 0.1148 ----- Adj R-squared = 0.1147 Total | 19873807.8 69,449 286.16406 Root MSE 15.917 syst Coefficient Std. err. t P>|t| [95% conf. interval] age | bmi alc | cons .4542968 .008985 50.56 0.000 .4366862 .4719074 3.392563 .8791044 .0120806 .1274243 72.77 0.000 .8554265 .9027824 26.62 0.000 3.142811 3.642314 77.42089 .5623379 137.68 0.000 76.31871 78.52307 (la) Using MATLAB, write a small program that calculates standard errors associated with your coefficient estimates. These should reproduce the second column of the STATA output above. Note that these are obtained from the matrix a(X'X)-. The MATLAB commands sqrt and diag may be useful here. (1b) Given the standard errors in (1a) and the B vector, use MATLAB to reproduce calculation of the t- statistics in STATA (3rd column of output). Note that these are statistics associated with testing the null hypotheses that the population regression coefficient equals zero. Source | SS df MS Number of obs 69,450 F (3, 69446) 3000.65 Model | 2280534.24 Residual | 17593273.6 ------ 3 760178.079 69,446 Prob > F = 0.0000 253.337465 R-squared = 0.1148 ----- Adj R-squared = 0.1147 Total | 19873807.8 69,449 286.16406 Root MSE 15.917 syst Coefficient Std. err. t P>|t| [95% conf. interval] age | bmi alc | cons .4542968 .008985 50.56 0.000 .4366862 .4719074 3.392563 .8791044 .0120806 .1274243 72.77 0.000 .8554265 .9027824 26.62 0.000 3.142811 3.642314 77.42089 .5623379 137.68 0.000 76.31871 78.52307 (la) Using MATLAB, write a small program that calculates standard errors associated with your coefficient estimates. These should reproduce the second column of the STATA output above. Note that these are obtained from the matrix a(X'X)-. The MATLAB commands sqrt and diag may be useful here. (1b) Given the standard errors in (1a) and the B vector, use MATLAB to reproduce calculation of the t- statistics in STATA (3rd column of output). Note that these are statistics associated with testing the null hypotheses that the population regression coefficient equals zero.