A good loglinear model 8 The GENMOD Procedure Model Information Data Set WORK.SEX Distribution Poisson Link Function Log Dependent Variable count Number of Observations Read 16 Number of Observations Used 16 Class Level Information Class Levels Values m 2 divorced married ems 2 yes no pms 2 yes no g 2 women men Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 6 8.1535 1.3589 Scaled Deviance 6 8.1535 1.3589 Pearson Chi-Square 6 8.2820 1.3803 Scaled Pearson X2 6 8.2820 1.3803 Log Likelihood 3947.5236 Algorithm converged. Analysis Of Parameter Estimates Standard Wald 95% Parameter DF Estimate Error Confidence Limits Intercept 1 4.8244 0.0743 4.6788 4.9700 Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq Intercept 4215.91 <.0001 A good loglinear model 9 The GENMOD Procedure Analysis Of Parameter Estimates Standard Wald 95% Parameter DF Estimate Error Confidence Limits m divorced 1 -0.4718 0.0759 -0.6205 -0.3230 m married 0 0.0000 0.0000 0.0000 0.0000 ems yes 1 -4.0342 0.3567 -4.7333 -3.3352 ems no 0 0.0000 0.0000 0.0000 0.0000 pms yes 1 -1.1558 0.1529 -1.4555 -0.8560 pms no 0 0.0000 0.0000 0.0000 0.0000 g women 1 0.9670 0.0794 0.8114 1.1226 g men 0 0.0000 0.0000 0.0000 0.0000 m*ems divorced yes 1 2.3626 0.3868 1.6045 3.1207 m*ems divorced no 0 0.0000 0.0000 0.0000 0.0000 m*ems married yes 0 0.0000 0.0000 0.0000 0.0000 m*ems married no 0 0.0000 0.0000 0.0000 0.0000 m*pms divorced yes 1 1.0033 0.1716 0.6669 1.3397 m*pms divorced no 0 0.0000 0.0000 0.0000 0.0000 m*pms married yes 0 0.0000 0.0000 0.0000 0.0000 m*pms married no 0 0.0000 0.0000 0.0000 0.0000 ems*pms yes yes 1 2.5376 0.4569 1.6420 3.4332 ems*pms yes no 0 0.0000 0.0000 0.0000 0.0000 ems*pms no yes 0 0.0000 0.0000 0.0000 0.0000 ems*pms no no 0 0.0000 0.0000 0.0000 0.0000 pms*g yes women 1 -1.3106 0.1530 -1.6104 -1.0108 pms*g yes men 0 0.0000 0.0000 0.0000 0.0000 pms*g no women 0 0.0000 0.0000 0.0000 0.0000 pms*g no men 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced yes yes 1 -1.7955 0.5121 -2.7991 -0.7919 m*ems*pms divorced yes no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced no yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced no no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married yes yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married yes no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married no yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married no no 0 0.0000 0.0000 0.0000 0.0000 Scale 0 1.0000 0.0000 1.0000 1.0000 Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq m divorced 38.65 <.0001 m married . . ems yes 127.94 <.0001 ems no . . pms yes 57.11 <.0001 pms no . . g women 148.38 <.0001 A good loglinear model 10 The GENMOD Procedure Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq g men . . m*ems divorced yes 37.31 <.0001 m*ems divorced no . . m*ems married yes . . m*ems married no . . m*pms divorced yes 34.17 <.0001 m*pms divorced no . . m*pms married yes . . m*pms married no . . ems*pms yes yes 30.84 <.0001 ems*pms yes no . . ems*pms no yes . . ems*pms no no . . pms*g yes women 73.42 <.0001 pms*g yes men . . pms*g no women . . pms*g no men . . m*ems*pms divorced yes yes 12.29 0.0005 m*ems*pms divorced yes no . . m*ems*pms divorced no yes . . m*ems*pms divorced no no . . m*ems*pms married yes yes . . m*ems*pms married yes no . . m*ems*pms married no yes . . m*ems*pms married no no . . Scale NOTE: The scale parameter was held fixed. Observation Statistics Observation count m ems pms g Pred Xbeta Std HessWgt Lower Upper Resraw Reschi Resdev StResdev StReschi Reslik 1 17 divorced yes yes women 18.672199 2.9270357 0.1675495 18.672199 13.445503 25.930679 -1.672199 -0.386982 -0.392984 -0.569711 -0.561009 -0.565166 2 54 divorced no yes women 47.302905 3.8565717 0.1209237 47.302905 37.321349 59.954017 6.6970952 0.973739 0.9520202 1.7145564 1.7536713 1.7417054 3 36 divorced yes no women 38.4 3.6480575 0.1390905 38.4 29.237293 50.434218 -2.4 -0.387298 -0.391442 -0.771987 -0.763816 -0.765925 A good loglinear model 11 The GENMOD Procedure Observation Statistics Observation count m ems pms g Pred Xbeta Std HessWgt Lower Upper Resraw Reschi Resdev StResdev StReschi Reslik 4 214 divorced no no women 204.31698 5.3196726 0.0634378 204.31698 180.42892 231.36773 9.6830189 0.6774208 0.6721731 1.5942906 1.6067374 1.604532 5 28 divorced yes yes men 26.327801 3.2706254 0.1586349 26.327801 19.292312 35.92898 1.6721992 0.3258973 0.3225356 0.555222 0.5610089 0.5590627 6 60 divorced no yes men 66.697095 4.2001614 0.1082347 66.697095 53.948252 82.458697 -6.697095 -0.820036 -0.834368 -1.78432 -1.753671 -1.760419 7 17 divorced yes no men 14.6 2.6810215 0.148917 14.6 10.904217 19.548401 2.4 0.6281087 0.6119818 0.7442047 0.763816 0.7506104 8 68 divorced no no men 77.683019 4.3526367 0.0827916 77.683019 66.047102 91.368906 -9.683019 -1.098621 -1.122718 -1.64198 -1.606737 -1.62331 9 4 married yes yes women 6.2240707 1.8284242 0.2692903 6.2240707 3.6715905 10.551029 -2.224071 -0.891479 -0.954517 -1.288656 -1.203551 -1.250961 10 25 married no yes women 27.80083 3.3250659 0.1441388 27.80083 20.958808 36.876436 -2.80083 -0.5312 -0.540516 -0.831651 -0.817317 -0.823403 11 4 married yes no women 5.7963305 1.7572251 0.3542259 5.7963305 2.8949128 11.605685 -1.796331 -0.746122 -0.790706 -1.514163 -1.428787 -1.452567 12 322 married no no women 327.48679 5.7914477 0.0518714 327.48679 295.82904 362.53236 -5.486792 -0.303195 -0.304047 -0.88194 -0.879467 -0.879761 13 11 married yes yes men 8.7759397 2.1720139 0.263836 8.7759397 5.2325816 14.718761 2.2240603 0.7507576 0.7219908 1.1574293 1.2035457 1.1858144 14 42 married no yes men 39.19917 3.6686556 0.133672 39.19917 30.164428 50.939966 2.80083 0.4473509 0.4421762 0.8078631 0.8173174 0.8144965 15 4 married yes no men 2.2038132 0.7901891 0.3581984 2.2038132 1.0921335 4.4470685 1.7961868 1.2099408 1.0846512 1.2807332 1.4286725 1.3242419 16 130 married no no men 124.51321 4.8244118 0.0743017 124.51321 107.63902 144.0327 5.4867923 0.491712 0.4881654 0.8731238 0.8794671 0.8774891 Loglinear model equal to logit model 12 The GENMOD Procedure Model Information Data Set WORK.SEX Distribution Poisson Link Function Log Dependent Variable count Number of Observations Read 16 Number of Observations Used 16 Class Level Information Class Levels Values m 2 divorced married ems 2 yes no pms 2 yes no g 2 women men Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 3 0.6978 0.2326 Scaled Deviance 3 0.6978 0.2326 Pearson Chi-Square 3 0.7013 0.2338 Scaled Pearson X2 3 0.7013 0.2338 Log Likelihood 3951.2514 Algorithm converged. Analysis Of Parameter Estimates Standard Wald 95% Parameter DF Estimate Error Confidence Limits Intercept 1 4.8850 0.0836 4.7211 5.0489 Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq Intercept 3413.29 <.0001 Loglinear model equal to logit model 13 The GENMOD Procedure Analysis Of Parameter Estimates Standard Wald 95% Parameter DF Estimate Error Confidence Limits m divorced 1 -0.6997 0.1327 -0.9599 -0.4395 m married 0 0.0000 0.0000 0.0000 0.0000 ems yes 1 -3.7052 0.3972 -4.4837 -2.9266 ems no 0 0.0000 0.0000 0.0000 0.0000 pms yes 1 -1.1729 0.1550 -1.4767 -0.8691 pms no 0 0.0000 0.0000 0.0000 0.0000 g women 1 0.8824 0.0977 0.6909 1.0740 g men 0 0.0000 0.0000 0.0000 0.0000 m*ems divorced yes 1 2.3960 0.3879 1.6358 3.1563 m*ems divorced no 0 0.0000 0.0000 0.0000 0.0000 m*ems married yes 0 0.0000 0.0000 0.0000 0.0000 m*ems married no 0 0.0000 0.0000 0.0000 0.0000 m*pms divorced yes 1 1.0995 0.1787 0.7493 1.4497 m*pms divorced no 0 0.0000 0.0000 0.0000 0.0000 m*pms married yes 0 0.0000 0.0000 0.0000 0.0000 m*pms married no 0 0.0000 0.0000 0.0000 0.0000 ems*pms yes yes 1 2.3463 0.4996 1.3672 3.3255 ems*pms yes no 0 0.0000 0.0000 0.0000 0.0000 ems*pms no yes 0 0.0000 0.0000 0.0000 0.0000 ems*pms no no 0 0.0000 0.0000 0.0000 0.0000 pms*g yes women 1 -1.3341 0.1770 -1.6809 -0.9872 pms*g yes men 0 0.0000 0.0000 0.0000 0.0000 pms*g no women 0 0.0000 0.0000 0.0000 0.0000 pms*g no men 0 0.0000 0.0000 0.0000 0.0000 ems*g yes women 1 -0.5049 0.2920 -1.0773 0.0674 ems*g yes men 0 0.0000 0.0000 0.0000 0.0000 ems*g no women 0 0.0000 0.0000 0.0000 0.0000 ems*g no men 0 0.0000 0.0000 0.0000 0.0000 m*g divorced women 1 0.3089 0.1458 0.0231 0.5947 m*g divorced men 0 0.0000 0.0000 0.0000 0.0000 m*g married women 0 0.0000 0.0000 0.0000 0.0000 m*g married men 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced yes yes 1 -1.7999 0.5130 -2.8052 -0.7945 m*ems*pms divorced yes no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced no yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms divorced no no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married yes yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married yes no 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married no yes 0 0.0000 0.0000 0.0000 0.0000 m*ems*pms married no no 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g yes yes women 1 0.1030 0.4230 -0.7261 0.9322 ems*pms*g yes yes men 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g yes no women 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g yes no men 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g no yes women 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g no yes men 0 0.0000 0.0000 0.0000 0.0000 Loglinear model equal to logit model 14 The GENMOD Procedure Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq m divorced 27.79 <.0001 m married . . ems yes 87.01 <.0001 ems no . . pms yes 57.25 <.0001 pms no . . g women 81.53 <.0001 g men . . m*ems divorced yes 38.16 <.0001 m*ems divorced no . . m*ems married yes . . m*ems married no . . m*pms divorced yes 37.87 <.0001 m*pms divorced no . . m*pms married yes . . m*pms married no . . ems*pms yes yes 22.06 <.0001 ems*pms yes no . . ems*pms no yes . . ems*pms no no . . pms*g yes women 56.82 <.0001 pms*g yes men . . pms*g no women . . pms*g no men . . ems*g yes women 2.99 0.0838 ems*g yes men . . ems*g no women . . ems*g no men . . m*g divorced women 4.49 0.0342 m*g divorced men . . m*g married women . . m*g married men . . m*ems*pms divorced yes yes 12.31 0.0005 m*ems*pms divorced yes no . . m*ems*pms divorced no yes . . m*ems*pms divorced no no . . m*ems*pms married yes yes . . m*ems*pms married yes no . . m*ems*pms married no yes . . m*ems*pms married no no . . ems*pms*g yes yes women 0.06 0.8076 ems*pms*g yes yes men . . ems*pms*g yes no women . . ems*pms*g yes no men . . ems*pms*g no yes women . . ems*pms*g no yes men . . Loglinear model equal to logit model 15 The GENMOD Procedure Analysis Of Parameter Estimates Standard Wald 95% Parameter DF Estimate Error Confidence Limits ems*pms*g no no women 0 0.0000 0.0000 0.0000 0.0000 ems*pms*g no no men 0 0.0000 0.0000 0.0000 0.0000 Scale 0 1.0000 0.0000 1.0000 1.0000 Analysis Of Parameter Estimates Chi- Parameter Square Pr > ChiSq ems*pms*g no no women . . ems*pms*g no no men . . Scale NOTE: The scale parameter was held fixed. Loglinear model equal to logit model 16 The GENMOD Procedure Model Information Data Set WORK.LSEX Distribution Binomial Link Function Logit Response Variable (Events) divorced Response Variable (Trials) n Number of Observations Read 8 Number of Observations Used 8 Number of Events 494 Number of Trials 1036 Class Level Information Class Levels Values ems 2 yes no pms 2 yes no g 2 women men Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 3 0.6978 0.2326 Scaled Deviance 3 0.6978 0.2326 Pearson Chi-Square 3 0.7013 0.2338 Scaled Pearson X2 3 0.7013 0.2338 Log Likelihood -663.3590 Algorithm converged. Analysis Of Parameter Estimates Standard Wald 95% Chi- Parameter DF Estimate Error Confidence Limits Square Intercept 1 -0.6997 0.1327 -0.9599 -0.4395 27.79 g women 1 0.3089 0.1458 0.0231 0.5947 4.49 g men 0 0.0000 0.0000 0.0000 0.0000 . pms yes 1 1.0995 0.1787 0.7493 1.4497 37.87 pms no 0 0.0000 0.0000 0.0000 0.0000 . ems yes 1 2.3960 0.3879 1.6358 3.1563 38.16 ems no 0 0.0000 0.0000 0.0000 0.0000 . ems*pms yes yes 1 -1.7999 0.5130 -2.8052 -0.7945 12.31 Loglinear model equal to logit model 17 The GENMOD Procedure Analysis Of Parameter Estimates Standard Wald 95% Chi- Parameter DF Estimate Error Confidence Limits Square ems*pms yes no 0 0.0000 0.0000 0.0000 0.0000 . ems*pms no yes 0 0.0000 0.0000 0.0000 0.0000 . ems*pms no no 0 0.0000 0.0000 0.0000 0.0000 . Scale 0 1.0000 0.0000 1.0000 1.0000 Analysis Of Parameter Estimates Parameter Pr > ChiSq Intercept <.0001 g women 0.0342 g men . pms yes <.0001 pms no . ems yes <.0001 ems no . ems*pms yes yes 0.0005 ems*pms yes no . ems*pms no yes . ems*pms no no . Scale NOTE: The scale parameter was held fixed. Observation Statistics Observation divorced n ems pms g Pred Xbeta Std HessWgt Lower Upper Resraw Reschi Resdev StResdev StReschi Reslik 1 17 21 yes yes women 0.7866505 1.3048522 0.3150002 3.5244617 0.6654024 0.8723881 0.4803402 0.2558601 0.2593777 0.321648 0.3172859 0.3201293 2 54 79 no yes women 0.6701106 0.7086851 0.1762825 17.46393 0.5898095 0.7415792 1.0612665 0.2539531 0.2548534 0.3768688 0.3755374 0.3761468 3 36 40 yes no women 0.8813451 2.0052302 0.3841734 4.183036 0.7776916 0.9403746 0.7461953 0.3648436 0.3737419 0.6042028 0.5898175 0.5953628 Loglinear model equal to logit model 18 The GENMOD Procedure Observation Statistics Observation divorced n ems pms g Pred Xbeta Std HessWgt Lower Upper Resraw Reschi Resdev StResdev StReschi Reslik 4 214 536 no no women 0.403522 -0.390811 0.0845813 129.01091 0.3643374 0.4439775 -2.287802 -0.201421 -0.201538 -0.726029 -0.72561 -0.725642 5 28 39 yes yes men 0.7302651 0.9959682 0.3022189 7.6821408 0.5995595 0.8303777 -0.48034 -0.173304 -0.172493 -0.315801 -0.317286 -0.316844 6 60 102 no yes men 0.5986399 0.3998011 0.1658526 24.507558 0.5186753 0.6736777 -1.061266 -0.214375 -0.214079 -0.375019 -0.375537 -0.375368 7 17 21 yes no men 0.8450569 1.6963461 0.3898533 2.7496502 0.7175311 0.9213219 -0.746195 -0.450001 -0.437253 -0.573108 -0.589817 -0.580149 8 68 198 no no men 0.3318798 -0.699695 0.1327395 43.903648 0.2769062 0.3918528 2.2878019 0.3452774 0.3442934 0.7235421 0.72561 0.7251423