> dfbetas .mr (Intercept ) x1 x2 x3 1 -0. 246116897 1. 393566e-01 1.393566e-01 0 . 134708113 2 0 . 703903616 -3. 985652e-01 -3.985652e-01 -0 . 385270290 3 0 . 390545624 5. 502359e-01 -5.502359e-01 -0 . 566915870 4 -0. 606633752 -8. 546804e-01 8.546804e-01 0 . 880589310 5 0 . 002794219 3. 863346e-18 4. 021962e-18 -0 . 001484435 6 -0. 045976523 -4. 656027e-17 -5. 894186e-17 0 . 024425126 7 -0. 038506629 -4.546330e-17 -5.522508e-17 0 . 020456729 8 0 . 013238747 1. 516617e-17 1. 734227e-17 -0 . 007033112 9 -0. 480229093 1. 368152e+00 1. 368152e+00 -1 . 322514372 10 0.286856522 -8. 172417e-01 -8. 172417e-01 0 . 789981029 11 -0. 006476926 1. 364870e-02 -1. 364870e-02 0 . 014062452 12 -0. 091981581 1. 938310e-01 -1. 938310e-01 0 . 199706854 1. What is the interpretation of the DFBETAS value for observation 9 and predictor 3 ? 2. What will happen to this predictor's parameter estimate if observation 9 is omitted? Is it likely indicate strong evidence against Ho : B3 = 0? 3. How would you determine which points are influential, based on the DFBETAS?\f\f> mr=1m(y~x1+x2+x3) > summary (mr ) Call : Im ( formula = y ~ x1 + x2 + x3 ) Residuals : Min 10 Median 30 Max -23. 808 -17 . 193 0 . 904 8. 143 32 . 192 Coefficients : Estimate Std. Error t value P r ( > t ) (Intercept) 397 . 08738 62 . 75676 6. 327 0 . 000226 * * * x1 -110 . 75000 14 . 76248 -7.502 6 . 91e-05 * * * x2 15 . 58333 4.92083 3. 167 0 . 013258 * x3 -0 . 05829 0 . 02563 -2. 274 0 . 052565 . Signif. codes : 0 '* * * ' 0. 001 '* * ' 0.01 '* 0. 05 '. ' 0.1 1 Residual standard error: 20.88 on 8 degrees of freedom Multiple R-squared: 0. 8993, Adjusted R- squared: 0.8616 F-statistic: 23.83 on 3 and 8 DF, p-value: 0 . 0002422