Side-life-insurance is a company that provides compensation cover to purchasers of the insur- ance in the event of a car accident that makes a person unable to work. Side-life-insurance serves in Kentucky (ky) and Michigan (mi). Since the number of weeks of the pay-out to people who are injured is too few, many high earning individuals were withdrawing from this insurance pro- gram. As a result, Side-life-insurance increased compensation cover for high earning individuals. logweeks is logarithm of number of weeks compensation is covered afchnge is a dummy =1 if after policy change in benefits highearn is a dummy =1 if an individual is high earner afchngehighearn is an interaction term=afchnge"highearn The following result was found from a random sample of data from both before and after policy change in both Michigan and Kentucky The following result was found from a random sample of data from both before and after policy change in both Michigan and Kentucky . regress logweeks afchnge highearn highlpre Source 55df MS Number of obs- FL 3,7146) - 44.10 Model 222.508722 3 74.169574 Prob > F 0.0000 Residual 12018.6019 7146 1.68186425 R-squared 0.0182 Adj R-squared. 0.0178 Total 12241.1107 7149 1.71228293 Root MSE 1.2969 7150 Logweeks Coef. Std. Err. t P>It [95% Conf. Interval] afchnge highearn highlpre cons .0974329 .030724 -3.014077 .6514712 .532954 . 1044874 1.164927 .024424 3.17 -4.63 5.10 47.70 0.002 0.000 0.000 0.000 .0372048 -4.291154 .3281277 1.117048 . 157661 -1.737001 .7377803 1.212805 1 1a) Comment on overall effectiveness of the policy. Did it affect the target beneficiaries? Com- ment on the small R which suggests the regression is useless. 1b) The researcher was interested in looking at the two states (ky and mi) separately and found the following result. The top part of the table is for Michigan and the bottom is for Kentucky. regress logweeks afchnge highearn highlpre if mi==1 Source FO3, 1520) - R-squared Adj R-squared SS df MS Model Residual 31.4888756 2882.86596 3 1520 10.4962919 1.89662234 Number of obs - 1524 5.53 Prob > F 0.0009 = 0.0108 0.0089 Root MSE 1.3772 Total 2914.35483 1523 1.91356194 Coef. Std. Err. t P>t! (95% Conf. Intervall Logweeks afchnge highearn highlpre cons 1551489 .7333176 -.0718054 1.386888 .0708783 3.178922 . 4812673 .0527739 2.19 0.23 -0.15 26.28 0.029 0.818 0.881 0.000 .0161193 -5.50222 -1.615824 1.28337 .2941785 6.968855 .8722128 1.490405 regress logweeks afchnge highearn highlpre if kynl Source SS df MS Model Residual 203. 149135 9043.8568 3 5622 67.7163785 1.60865471 Number of obs 5626 FC 3, 5622) - 42.10 Prob > F 0.0000 R-squared = 0.0220 Adj R-squared - 0.0214 Root MSE 1.2683 Total 9247.00594 5625 1.64391217 Logweeks Coef. Std. Err. t P>It! afchnge highearn highlpre cons .0855711 -3.326372 .5968738 1.088804 .0338656 .9412415 .1528008 .0274542 2.53 -3.53 3.91 39.66 0.012 0.000 0.000 0.000 (95% Conf. Intervall .0191815 . 1519607 -5.171569 -1.481176 .2973253 .8964223 1.034983 1.142625 Compare and contrast the two states results 1c) The researcher wanted to focus on Kentucky and included other variables such as gender, industry type, marital status, age and others. highlpre is interaction term =afchnge*highearn. male is dummy = 1 if the individual is male, married is dummy =1 if the individual is married. afchngehighearnmarried is an interaction term =afchnge highearn*married. hosp is a dummy variable if the individual is hospitalized Comment on the new variables and the R2 regress logweeks afchnge highearn highlpre afchngehighearnmarried male married age hosp if ky==1 Source SS df Number of obs - 5360 Fl 8, 5351) - 142.14 Model 1526.53722 8 190.817152 Prob > F Residual 7183.73003 5351 1.34250234 R-squared Adj R-squared - 0.1740 Total 8710.26724 5359 1.6253531 Root MSE . 1.1587 0.0000 - 0.1753 Logweeks Coef. Std. Err. t P>It1 [95% Conf. Interval] afchnge highearn highlpre afchngehighearnmarried male married age hosp cons 0510588 -2.23386 .389771 .1757785 ..0681235 .6477297 .0064607 1.1009 .6555451 .0387543 .8815614 .1433655 .0615866 .0414233 .0381707 .0013444 .0360648 .0631791 1.32 0.188 -2.53 0.011 2.72 0.007 2.85 0.004 -1.64 0.100 1.25 0.211 4.81 0.000 30.53 0.000 10.38 0.000 ..0249155 -3.96208 . 1087162 .0550438 .. 1493301 ..0271004 .6038252 1.030199 .5316883 .1270331 ..5056407 .6708258 .2965133 .0130831 .1225597 .0090963 1.171602 .7794019 1d) Redoing 1c) above for Michigan the following result is found regress logweeks afchnge highearn highlpre afchngehighearnmarried male married age hosp if ni==1 Source SS 55 df MS Number of obs- 1484 Fl FI 8, 1475) - 28.94 Model 382.21964 8 47.777455 Prob > F = 0.0000 Residual 2434.87679 1475 1.65076392 R-squared = 0.1357 Adj R-squared = 0.1310 Total 2817.09643 1483 1.899593 Root MSE = 1.2848 logweeks Coef. Std. Err. . . t P>It [95% Conf. Intervall afchnge .097517.0769156 highearn 3.761461 3.009743 highlpre 5484503.4558464 afchngehighearnmarried .1656494 1431021 male 2931887.0918341 married ..0782431.0771634 age .0133218.0028908 hosp 1.023929.0766651 _cons .976849.1362236 1.27 0.205 1.25 0.212 -1.20 0.229 1.16 0.247 -3.19 0.001 -1.91 0.311 4.61 0.000 13.36 0.000 7.170.000 ..0533585 .2483925 -2.142371 9.665294 -1.442627 .345726 ..1150558 .4463547 ..4733281 - 1130494 ..2296048 .0731186 .0076512 .6189924 .873545 1.174314 7096363 1.244062 Comment on this result and compare it with the Michigan result in 1b)