!! Forum request 10/10/24 for more precision on !! p-values when reported as ">0.5" !! !! Piece-wise exponential regression Additive excess relative risk T0 * (1 + T1 + T2 + ...) all_can is used for cases py10k is used for person years Parameter Summary Table # Name Estimate Std.Err. Test Stat. P value -- ---------------------------- ---------- --------- ---------- -------- Log-linear term 0 1 %CON..................... 4.790 0.03605 132.9 < 0.001 3 sex_2.................... -0.3414 0.01428 -23.9 < 0.001 5 byrcat_2................. 0.01111 0.03603 0.3082 > 0.5 6 byrcat_3................. 0.09065 0.03602 2.516 0.0119 7 byrcat_4................. 0.1066 0.03799 2.807 0.005 8 byrcat_5................. 0.01180 0.3556 0.03317 > 0.5 9 lage70................... 5.808 0.04701 123.5 < 0.001 Linear term 1 10 dgy...................... 0.002907 0.01513 0.1921 > 0.5 Records used 13870 Deviance 16618.78 AIC 16634.78 Pearson Chi2 7419467.7 Degrees of freedom 13862 null @ ! set this model to the null model param 10 = 0 @ ! create a new model with dgy = 0 fit @ ! fit the simpler model Iter Step Deviance 0 0 16618.897 1 0 16618.817 2 0 16618.817 Piece-wise exponential regression Additive excess relative risk T0 * (1 + T1 + T2 + ...) all_can is used for cases py10k is used for person years Parameter Summary Table # Name Estimate Std.Err. Test Stat. P value -- ---------------------------- ---------- --------- ---------- -------- Log-linear term 0 1 %CON..................... 4.792 0.03445 139.1 < 0.001 3 sex_2.................... -0.3414 0.01428 -23.91 < 0.001 5 byrcat_2................. 0.01105 0.03603 0.3066 > 0.5 6 byrcat_3................. 0.09060 0.03602 2.515 0.0119 7 byrcat_4................. 0.1066 0.03798 2.806 0.00502 8 byrcat_5................. 0.01172 0.3556 0.03297 > 0.5 9 lage70................... 5.808 0.04701 123.5 < 0.001 Linear term 1 10 dgy...................... 0.000 Fixed 0.1924 > 0.5 Records used 13870 Deviance 16618.817 AIC 16632.817 Pearson Chi2 7424728.3 Degrees of freedom 13863 !! Note that the deviance is smaller !! Do an LRT test comparing the models, this is the p-value on parameter 10 lrt @ LR statistic -0.03702 Degrees of freedom -1 P value > 0.50 !! Get the precision on the p-value using the chi-squared test chisq 1 #_LRT @ Chi square = 0.0370243 df = 1 P = 0.847416 !! P-value with full precision !! get back to the base model param 10 free @ param 4 = 0 @ !! make sure that the reference parameter shows and is set to 0 fit @ Iter Step Deviance 0 0 16618.817 1 0 16618.780 2 0 16618.780 Piece-wise exponential regression Additive excess relative risk T0 * (1 + T1 + T2 + ...) all_can is used for cases py10k is used for person years Parameter Summary Table # Name Estimate Std.Err. Test Stat. P value -- ---------------------------- ---------- --------- ---------- -------- Log-linear term 0 1 %CON..................... 4.790 0.03605 132.9 < 0.001 3 sex_2.................... -0.3414 0.01428 -23.9 < 0.001 4 byrcat_1................. 0.000 Aliased 5 byrcat_2................. 0.01111 0.03603 0.3082 > 0.5 6 byrcat_3................. 0.09065 0.03602 2.516 0.0119 7 byrcat_4................. 0.1066 0.03799 2.807 0.005 8 byrcat_5................. 0.01180 0.3556 0.03317 > 0.5 9 lage70................... 5.808 0.04701 123.5 < 0.001 Linear term 1 10 dgy...................... 0.002907 0.01513 0.1921 > 0.5 Records used 13870 Deviance 16618.78 AIC 16634.78 Pearson Chi2 7419467.7 Degrees of freedom 13862 !! Set that to the null model null @ param 5=0 @ !! set a categorical parameter to 0 (param 5; 'byrcat_2') fit @ Iter Step Deviance 0 0 16619.547 1 0 16618.876 2 0 16618.875 Piece-wise exponential regression Additive excess relative risk T0 * (1 + T1 + T2 + ...) all_can is used for cases py10k is used for person years Parameter Summary Table # Name Estimate Std.Err. Test Stat. P value -- ---------------------------- ---------- --------- ---------- -------- Log-linear term 0 1 %CON..................... 4.800 0.01779 269.9 < 0.001 3 sex_2.................... -0.3410 0.01422 -23.98 < 0.001 4 byrcat_1................. 0.000 Fixed -0.3082 > 0.5 5 byrcat_2................. 0.000 Fixed 0.3082 > 0.5 6 byrcat_3................. 0.08083 0.01674 4.828 < 0.001 7 byrcat_4................. 0.09670 0.0201 4.811 < 0.001 8 byrcat_5................. 0.001696 0.3541 0.00479 > 0.5 9 lage70................... 5.807 0.0469 123.8 < 0.001 Linear term 1 10 dgy...................... 0.002868 0.01513 0.1895 > 0.5 Records used 13870 Deviance 16618.875 AIC 16632.875 Pearson Chi2 7404296.2 Degrees of freedom 13863 lrt @ LR statistic -0.09525 Degrees of freedom -1 P value > 0.50 !! p-value for byrcat_2 chisq 1 #_LRT @ Chi square = 0.095253 df = 1 P = 0.757602