unadjusted model for baseline cholesterol 22:31 Monday, April 19, 2004 1 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 unadjusted model for baseline cholesterol 22:31 Monday, April 19, 2004 2 The GLM Procedure Dependent Variable: CHOLEST Sum of Source DF Squares Mean Square F Value Pr > F Model 5 62153.369 12430.674 6.85 <.0001 Error 910 1652345.697 1815.765 Corrected Total 915 1714499.066 R-Square Coeff Var Root MSE CHOLEST Mean 0.036252 18.58863 42.61179 229.2358 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 3107.92135 1553.96067 0.86 0.4253 SEX 1 55736.25417 55736.25417 30.70 <.0001 TRETGRP*SEX 2 1300.86236 650.43118 0.36 0.6990 unadjusted model for baseline cholesterol 22:31 Monday, April 19, 2004 3 The GLM Procedure Level of -----------CHOLEST----------- TRETGRP N Mean Std Dev 3 305 229.219672 44.1896437 4 306 226.307190 39.6880097 5 305 232.190164 45.7095727 Level of -----------CHOLEST----------- SEX N Mean Std Dev 0 422 220.672986 39.9861280 1 494 236.550607 44.6758039 Level of Level of -----------CHOLEST----------- TRETGRP SEX N Mean Std Dev 3 0 140 221.050000 40.6714021 3 1 165 236.151515 45.9604292 4 0 152 219.730263 36.5471609 4 1 154 232.798701 41.6696281 5 0 130 221.369231 43.2494367 5 1 175 240.228571 45.9472997 unadjusted model for 6-month followup cholesterol 22:31 Monday, April 19, 2004 4 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. unadjusted model for 6-month followup cholesterol 22:31 Monday, April 19, 2004 5 The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 5 66255.585 13251.117 8.87 <.0001 Error 896 1338703.566 1494.089 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.047158 16.43850 38.65344 235.1397 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 4217.87705 2108.93853 1.41 0.2443 SEX 1 62499.17341 62499.17341 41.83 <.0001 TRETGRP*SEX 2 671.34895 335.67448 0.22 0.7988 unadjusted model for 6-month followup cholesterol 22:31 Monday, April 19, 2004 6 The GLM Procedure Level of -----------CHOLEST6---------- TRETGRP N Mean Std Dev 3 298 237.686088 39.7181238 4 303 234.625045 36.9956777 5 301 233.136835 41.6281778 Level of -----------CHOLEST6---------- SEX N Mean Std Dev 0 420 226.289377 36.3380418 1 482 242.851643 40.5298809 Level of Level of -----------CHOLEST6---------- TRETGRP SEX N Mean Std Dev 3 0 139 228.785843 36.4977807 3 1 159 245.466805 40.8812336 4 0 152 227.340100 34.1229691 4 1 151 241.958234 38.4129882 5 0 129 222.361324 38.5880009 5 1 172 241.218469 42.0986744 unadjusted model for baseline cholesterol - less interaction 7 22:31 Monday, April 19, 2004 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 unadjusted model for baseline cholesterol - less interaction 8 22:31 Monday, April 19, 2004 The GLM Procedure Dependent Variable: CHOLEST Sum of Source DF Squares Mean Square F Value Pr > F Model 3 60852.506 20284.169 11.19 <.0001 Error 912 1653646.559 1813.209 Corrected Total 915 1714499.066 R-Square Coeff Var Root MSE CHOLEST Mean 0.035493 18.57554 42.58179 229.2358 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 3478.54778 1739.27389 0.96 0.3836 SEX 1 55565.81757 55565.81757 30.65 <.0001 unadjusted model for baseline cholesterol - less interaction 9 22:31 Monday, April 19, 2004 The GLM Procedure Least Squares Means CHOLEST Standard LSMEAN TRETGRP LSMEAN Error Pr > ¦t¦ Number 3 228.578213 2.440978 <.0001 1 4 226.256041 2.434257 <.0001 2 5 231.035538 2.447131 <.0001 3 Least Squares Means for effect TRETGRP Pr > ¦t¦ for H0: LSMean(i)=LSMean(j) Dependent Variable: CHOLEST i/j 1 2 3 1 0.5007 0.4764 2 0.5007 0.1664 3 0.4764 0.1664 NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should be used. H0:LSMean1= CHOLEST Standard H0:LSMEAN=0 LSMean2 SEX LSMEAN Error Pr > ¦t¦ Pr > ¦t¦ 0 220.797463 2.074805 <.0001 <.0001 1 236.449064 1.917250 <.0001 unadjusted model for 6-month followup cholesterol - less interaction 10 22:31 Monday, April 19, 2004 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. unadjusted model for 6-month followup cholesterol - less interaction 11 22:31 Monday, April 19, 2004 The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 3 65584.236 21861.412 14.66 <.0001 Error 898 1339374.915 1491.509 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.046681 16.42430 38.62006 235.1397 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 4019.88257 2009.94128 1.35 0.2604 SEX 1 62364.26951 62364.26951 41.81 <.0001 unadjusted model for 6-month followup cholesterol - less interaction 12 22:31 Monday, April 19, 2004 The GLM Procedure Least Squares Means CHOLEST6 Standard LSMEAN TRETGRP LSMEAN Error Pr > ¦t¦ Number 3 237.125700 2.238878 <.0001 1 4 234.652602 2.218668 <.0001 2 5 231.944010 2.233653 <.0001 3 Least Squares Means for effect TRETGRP Pr > ¦t¦ for H0: LSMean(i)=LSMean(j) Dependent Variable: CHOLEST6 i/j 1 2 3 1 0.4329 0.1011 2 0.4329 0.3899 3 0.1011 0.3899 NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should be used. H0:LSMean1= CHOLEST6 Standard H0:LSMEAN=0 LSMean2 SEX LSMEAN Error Pr > ¦t¦ Pr > ¦t¦ 0 226.224326 1.886447 <.0001 <.0001 1 242.923882 1.760458 <.0001 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 13 tests of homogeneity of variances The GLM Procedure Class Level Information Class Levels Values z 6 1 2 3 4 5 6 Number of observations 916 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 14 tests of homogeneity of variances The GLM Procedure Dependent Variable: CHOLEST Sum of Source DF Squares Mean Square F Value Pr > F Model 5 62153.369 12430.674 6.85 <.0001 Error 910 1652345.697 1815.765 Corrected Total 915 1714499.066 R-Square Coeff Var Root MSE CHOLEST Mean 0.036252 18.58863 42.61179 229.2358 Source DF Type I SS Mean Square F Value Pr > F z 5 62153.36879 12430.67376 6.85 <.0001 Source DF Type III SS Mean Square F Value Pr > F z 5 62153.36879 12430.67376 6.85 <.0001 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 15 tests of homogeneity of variances The GLM Procedure Bartlett's Test for Homogeneity of CHOLEST Variance Source DF Chi-Square Pr > ChiSq z 5 11.2622 0.0464 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 16 tests of homogeneity of variances The GLM Procedure Level of -----------CHOLEST----------- z N Mean Std Dev 1 140 221.050000 40.6714021 2 152 219.730263 36.5471609 3 130 221.369231 43.2494367 4 165 236.151515 45.9604292 5 154 232.798701 41.6696281 6 175 240.228571 45.9472997 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 17 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Moments N 916 Sum Weights 916 Mean 0 Sum Observations 0 Std Deviation 42.4952032 Variance 1805.84229 Skewness 0.47074728 Kurtosis 0.71444301 Uncorrected SS 1652345.7 Corrected SS 1652345.7 Coeff Variation . Std Error Mean 1.40408103 Basic Statistical Measures Location Variability Mean 0.0000 Std Deviation 42.49520 Median -2.9244 Variance 1806 Mode -30.1515 Range 311.64719 Interquartile Range 55.82359 NOTE: The mode displayed is the smallest of 2 modes with a count of 6. Tests for Location: Mu0=0 Test -Statistic- -----p Value------ Student's t t 0 Pr > ¦t¦ 1.0000 Sign M -24 Pr >= ¦M¦ 0.1204 Signed Rank S -7918 Pr >= ¦S¦ 0.3231 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 18 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Tests for Normality Test --Statistic--- -----p Value------ Shapiro-Wilk W 0.987676 Pr < W <0.0001 Kolmogorov-Smirnov D 0.034491 Pr > D <0.0100 Cramer-von Mises W-Sq 0.244507 Pr > W-Sq <0.0050 Anderson-Darling A-Sq 1.588224 Pr > A-Sq <0.0050 Quantiles (Definition 5) Quantile Estimate 100% Max 191.84848 99% 112.63077 95% 73.63077 90% 55.95000 75% Q3 25.89924 50% Median -2.92435 25% Q1 -29.92435 10% -54.05000 5% -64.15152 1% -85.22857 0% Min -119.79870 Extreme Observations ------Lowest------ -----Highest----- Value Obs Value Obs -119.7987 843 132.771 214 -110.3692 435 135.848 94 Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 19 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Extreme Observations ------Lowest------ -----Highest----- Value Obs Value Obs -99.2286 40 169.631 174 -93.2286 872 177.771 43 -92.7987 845 191.848 566 Histogram # Boxplot 190+* 1 0 .* 2 0 . 130+* 4 0 .** 5 0 .****** 23 ¦ 70+*********** 44 ¦ .****************** 69 ¦ .******************************* 124 +-----+ 10+***************************************** 162 ¦ + ¦ .********************************************* 177 *-----* .************************************* 145 +-----+ -50+*************************** 106 ¦ .********** 38 ¦ .**** 14 ¦ -110+* 2 0 ----+----+----+----+----+----+----+----+----+ * may represent up to 4 counts Analysis of baseline cholesterol 22:31 Monday, April 19, 2004 20 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Normal Probability Plot 190+ * ¦ * ¦ 130+ ** ¦ ***+ ¦ *****+ 70+ ***** ¦ ***** ¦ ****** 10+ +***** ¦ ****** ¦ ****** -50+ ******** ¦ ******+ ¦*****++ -110+*+ +----+----+----+----+----+----+----+----+----+----+ -2 -1 0 +1 +2 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 21 tests of homogeneity of variances The GLM Procedure Class Level Information Class Levels Values z 6 1 2 3 4 5 6 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 22 tests of homogeneity of variances The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 5 66255.585 13251.117 8.87 <.0001 Error 896 1338703.566 1494.089 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.047158 16.43850 38.65344 235.1397 Source DF Type I SS Mean Square F Value Pr > F z 5 66255.58541 13251.11708 8.87 <.0001 Source DF Type III SS Mean Square F Value Pr > F z 5 66255.58541 13251.11708 8.87 <.0001 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 23 tests of homogeneity of variances The GLM Procedure Bartlett's Test for Homogeneity of CHOLEST6 Variance Source DF Chi-Square Pr > ChiSq z 5 8.8624 0.1147 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 24 tests of homogeneity of variances The GLM Procedure Level of -----------CHOLEST6---------- z N Mean Std Dev 1 139 228.785843 36.4977807 2 152 227.340100 34.1229691 3 129 222.361324 38.5880009 4 159 245.466805 40.8812336 5 151 241.958234 38.4129882 6 172 241.218469 42.0986744 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 25 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Moments N 902 Sum Weights 902 Mean 0 Sum Observations 0 Std Deviation 38.5460442 Variance 1485.79752 Skewness 0.20182415 Kurtosis -0.0478482 Uncorrected SS 1338703.57 Corrected SS 1338703.57 Coeff Variation . Std Error Mean 1.28344288 Basic Statistical Measures Location Variability Mean 0.00000 Std Deviation 38.54604 Median -1.92780 Variance 1486 Mode . Range 234.28714 Interquartile Range 53.14313 Tests for Location: Mu0=0 Test -Statistic- -----p Value------ Student's t t 0 Pr > ¦t¦ 1.0000 Sign M -14 Pr >= ¦M¦ 0.3687 Signed Rank S -3591.5 Pr >= ¦S¦ 0.6466 Tests for Normality Test --Statistic--- -----p Value------ Shapiro-Wilk W 0.996478 Pr < W 0.0412 Kolmogorov-Smirnov D 0.022197 Pr > D >0.1500 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 26 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Tests for Normality Test --Statistic--- -----p Value------ Cramer-von Mises W-Sq 0.070274 Pr > W-Sq >0.2500 Anderson-Darling A-Sq 0.50466 Pr > A-Sq 0.2114 Quantiles (Definition 5) Quantile Estimate 100% Max 136.2699 99% 90.8901 95% 65.5652 90% 49.3750 75% Q3 26.3288 50% Median -1.9278 25% Q1 -26.8144 10% -48.9277 5% -62.0858 1% -82.4718 0% Min -98.0173 Extreme Observations ------Lowest----- -----Highest----- Value Obs Value Obs -98.0173 843 103.288 453 -97.4007 845 109.546 503 -96.2127 272 112.714 67 -94.1347 245 126.429 43 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 27 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Extreme Observations ------Lowest----- -----Highest----- Value Obs Value Obs -90.4523 435 136.270 174 Missing Values -----Percent Of----- Missing Missing Value Count All Obs Obs . 14 1.53 100.00 Analysis of Month 6 cholesterol 22:31 Monday, April 19, 2004 28 distribution of residuals from saturated (interaction) model The UNIVARIATE Procedure Variable: res Histogram # Boxplot Normal Probability Plot 135+* 1 0 135+ * .* 1 0 ¦ * .* 1 0 ¦ * .* 2 0 ¦ * .** 6 ¦ ¦ *** .***** 13 ¦ ¦ ****+ .***** 15 ¦ ¦ ***+ .******** 22 ¦ ¦ ***+ .********* 27 ¦ ¦ *** .************ 36 ¦ ¦ *** .************************ 70 ¦ ¦ **** .**************************** 83 +-----+ ¦ **** .********************** 66 ¦ ¦ ¦ *** .******************************** 94 ¦ + ¦ ¦ **** .********************************** 102 *-----* ¦ *** .**************************** 82 ¦ ¦ ¦ **** .**************************** 82 +-----+ ¦ **** .********************* 62 ¦ ¦ *** .****************** 52 ¦ ¦ **** .********** 30 ¦ ¦ *** .********** 30 ¦ ¦ **** .***** 13 ¦ ¦ **** .*** 7 ¦ ¦****+ -95+** 5 ¦ -95+*+ ----+----+----+----+----+----+---- +----+----+----+----+----+----+----+----+----+----+ * may represent up to 3 counts -2 -1 0 +1 +2 regression of 6 month on baseline cholesterol 22:31 Monday, April 19, 2004 29 The REG Procedure Model: MODEL1 Dependent Variable: CHOLEST6 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 1 933632 933632 1782.77 <.0001 Error 900 471327 523.69675 Corrected Total 901 1404959 Root MSE 22.88442 R-Square 0.6645 Dependent Mean 235.13972 Adj R-Sq 0.6642 Coeff Var 9.73227 Parameter Estimates Parameter Standard Variable DF Estimate Error t Value Pr > ¦t¦ Intercept 1 63.52631 4.13527 15.36 <.0001 CHOLEST 1 0.75024 0.01777 42.22 <.0001 covariance adjusted model with interaction 22:31 Monday, April 19, 2004 30 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. covariance adjusted model with interaction 22:31 Monday, April 19, 2004 31 The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 6 947400.487 157900.081 308.86 <.0001 Error 895 457558.664 511.239 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.674326 9.615810 22.61059 235.1397 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 8296.1071 4148.0535 8.11 0.0003 SEX 1 5957.6038 5957.6038 11.65 0.0007 TRETGRP*SEX 2 2.5697 1.2849 0.00 0.9975 CHOLEST 1 881144.9018 881144.9018 1723.55 <.0001 Standard Parameter Estimate Error t Value Pr > ¦t¦ Intercept 63.58925805 B 4.61289633 13.79 <.0001 TRETGRP 3 6.75779178 B 2.48823428 2.72 0.0067 TRETGRP 4 5.92964977 B 2.52460546 2.35 0.0191 TRETGRP 5 0.00000000 B . . . SEX 0 -5.37436591 B 2.65346578 -2.03 0.0431 SEX 1 0.00000000 B . . . TRETGRP*SEX 3 0 0.26316353 B 3.71898465 0.07 0.9436 TRETGRP*SEX 3 1 0.00000000 B . . . TRETGRP*SEX 4 0 0.11806041 B 3.70058939 0.03 0.9746 covariance adjusted model with interaction 22:31 Monday, April 19, 2004 32 The GLM Procedure Dependent Variable: CHOLEST6 Standard Parameter Estimate Error t Value Pr > ¦t¦ TRETGRP*SEX 4 1 0.00000000 B . . . TRETGRP*SEX 5 0 0.00000000 B . . . TRETGRP*SEX 5 1 0.00000000 B . . . CHOLEST 0.74217131 0.01787691 41.52 <.0001 NOTE: The X'X matrix has been found to be singular, and a generalized inverse was used to solve the normal equations. Terms whose estimates are followed by the letter 'B' are not uniquely estimable. covariance adjusted model without interaction 22:31 Monday, April 19, 2004 33 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. covariance adjusted model without interaction 22:31 Monday, April 19, 2004 34 The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 4 947397.918 236849.479 464.32 <.0001 Error 897 457561.234 510.102 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.674324 9.605111 22.58543 235.1397 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 8367.0068 4183.5034 8.20 0.0003 SEX 1 5958.2449 5958.2449 11.68 0.0007 CHOLEST 1 881813.6810 881813.6810 1728.70 <.0001 Standard Parameter Estimate Error t Value Pr > ¦t¦ Intercept 63.53142634 B 4.49309175 14.14 <.0001 TRETGRP 3 6.87575430 B 1.84699011 3.72 0.0002 TRETGRP 4 5.97963517 B 1.84297731 3.24 0.0012 TRETGRP 5 0.00000000 B . . . SEX 0 -5.24688626 B 1.53522087 -3.42 0.0007 SEX 1 0.00000000 B . . . CHOLEST 0.74218512 0.01785058 41.58 <.0001 covariance adjusted model without interaction 22:31 Monday, April 19, 2004 35 The GLM Procedure Dependent Variable: CHOLEST6 NOTE: The X'X matrix has been found to be singular, and a generalized inverse was used to solve the normal equations. Terms whose estimates are followed by the letter 'B' are not uniquely estimable. covariance adjusted model without interaction 22:31 Monday, April 19, 2004 36 The GLM Procedure Least Squares Means CHOLEST6 Standard LSMEAN TRETGRP LSMEAN Error Pr > ¦t¦ Number 3 237.554881 1.309361 <.0001 1 4 236.658762 1.298398 <.0001 2 5 230.679127 1.306619 <.0001 3 Least Squares Means for effect TRETGRP Pr > ¦t¦ for H0: LSMean(i)=LSMean(j) Dependent Variable: CHOLEST6 i/j 1 2 3 1 0.6271 0.0002 2 0.6271 0.0012 3 0.0002 0.0012 NOTE: To ensure overall protection level, only probabilities associated with pre-planned comparisons should be used. H0:LSMean1= CHOLEST6 Standard H0:LSMEAN=0 LSMean2 SEX LSMEAN Error Pr > ¦t¦ Pr > ¦t¦ 0 232.340814 1.112980 <.0001 0.0007 1 237.587700 1.037504 <.0001 test of homogeneity of slope assumption 22:31 Monday, April 19, 2004 37 The GLM Procedure Class Level Information Class Levels Values TRETGRP 3 3 4 5 SEX 2 0 1 Number of observations 916 NOTE: Due to missing values, only 902 observations can be used in this analysis. test of homogeneity of slope assumption 22:31 Monday, April 19, 2004 38 The GLM Procedure Dependent Variable: CHOLEST6 Sum of Source DF Squares Mean Square F Value Pr > F Model 11 949578.014 86325.274 168.71 <.0001 Error 890 455381.137 511.664 Corrected Total 901 1404959.151 R-Square Coeff Var Root MSE CHOLEST6 Mean 0.675876 9.619810 22.62000 235.1397 Source DF Type III SS Mean Square F Value Pr > F TRETGRP 2 2473.0844 1236.5422 2.42 0.0898 SEX 1 128.3500 128.3500 0.25 0.6166 TRETGRP*SEX 2 773.6197 386.8099 0.76 0.4698 CHOLEST 1 843507.6849 843507.6849 1648.56 <.0001 CHOLEST*TRETGRP*SEX 5 2177.5266 435.5053 0.85 0.5136 Standard Parameter Estimate Error t Value Pr > ¦t¦ Intercept 52.72183529 B 9.30821559 5.66 <.0001 TRETGRP 3 31.14615213 B 13.20657054 2.36 0.0186 TRETGRP 4 12.83870096 B 14.17122638 0.91 0.3652 TRETGRP 5 0.00000000 B . . . SEX 0 0.86879861 B 13.95066157 0.06 0.9504 SEX 1 0.00000000 B . . . TRETGRP*SEX 3 0 -19.50272534 B 19.96448626 -0.98 0.3289 TRETGRP*SEX 3 1 0.00000000 B . . . test of homogeneity of slope assumption 22:31 Monday, April 19, 2004 39 The GLM Procedure Dependent Variable: CHOLEST6 Standard Parameter Estimate Error t Value Pr > ¦t¦ TRETGRP*SEX 4 0 4.23952618 B 20.84831608 0.20 0.8389 TRETGRP*SEX 4 1 0.00000000 B . . . TRETGRP*SEX 5 0 0.00000000 B . . . TRETGRP*SEX 5 1 0.00000000 B . . . CHOLEST 0.78757764 B 0.03821815 20.61 <.0001 CHOLEST*TRETGRP*SEX 3 0 -0.04539822 B 0.06146088 -0.74 0.4603 CHOLEST*TRETGRP*SEX 3 1 -0.10270912 B 0.05458299 -1.88 0.0602 CHOLEST*TRETGRP*SEX 4 0 -0.07456143 B 0.06322588 -1.18 0.2386 CHOLEST*TRETGRP*SEX 4 1 -0.02836967 B 0.05927012 -0.48 0.6323 CHOLEST*TRETGRP*SEX 5 0 -0.02449821 B 0.05989109 -0.41 0.6826 CHOLEST*TRETGRP*SEX 5 1 0.00000000 B . . . NOTE: The X'X matrix has been found to be singular, and a generalized inverse was used to solve the normal equations. Terms whose estimates are followed by the letter 'B' are not uniquely estimable.