LEM: log-linear and event history analysis with missing data. Developed by Jeroen K. Vermunt (c), Tilburg University, The Netherlands. Version 1.2 (July 10, 1998). *** INPUT *** * Generated probabilities from 2PL with theta exp(1); * * a2= 1.7 b2= -1.05; * a3= .7 b3= -.35; * a4= 1.2 b4= .35; * a5= 1.7 b5= 1.05; * lat 1 man 4 dim 9 2 2 2 2 lab X A B C D mod X {wei(X)} A|X {A spe(A,1a,X,c,-1)} B|X {B spe(B,1a,X,c,-1)} C|X {C spe(C,1a,X,c,-1)} D|X {D spe(D,1a,X,c,-1)} sta wei(X) [0.000022345844 0.002789141321 0.049916406765 0.244097502895 0.406349206349 0.244097502895 0.049916406765 0.002789141321 0.000022345844] des [ -4.512745863399 -3.205429002856 -2.076847978677 -1.023255663789 0.000000000000 1.023255663789 2.076847978677 3.205429002856 4.512745863399 -4.512745863399 -3.205429002856 -2.076847978677 -1.023255663789 0.000000000000 1.023255663789 2.076847978677 3.205429002856 4.512745863399 -4.512745863399 -3.205429002856 -2.076847978677 -1.023255663789 0.000000000000 1.023255663789 2.076847978677 3.205429002856 4.512745863399 -4.512745863399 -3.205429002856 -2.076847978677 -1.023255663789 0.000000000000 1.023255663789 2.076847978677 3.205429002856 4.512745863399 ] nco dat gen4exp1.dat *** STATISTICS *** Number of iterations = 298 Converge criterion = 0.0000009951 Seed random values = 5128 X-squared = 2874.6934 (0.0000) L-squared = 2913.8902 (0.0000) Cressie-Read = 2886.6228 (0.0000) Dissimilarity index = 0.0200 Degrees of freedom = 7 Log-likelihood = -2206012.81261 Number of parameters = 8 (+1) Sample size = 1000000.0 BIC(L-squared) = 2817.1817 AIC(L-squared) = 2899.8902 BIC(log-likelihood) = 4412136.1493 AIC(log-likelihood) = 4412041.6252 Eigenvalues information matrix 8.99E+0005 7.33E+0005 5.79E+0005 4.32E+0005 2.25E+0005 1.22E+0005 60896.5460 26181.8096 *** FREQUENCIES *** A B C D observed estimated std. res. 1 1 1 1 11509.664 14418.414 -24.224 1 1 1 2 53030.991 53597.705 -2.448 1 1 2 1 22404.420 21774.623 4.268 1 1 2 2 124239.930 121357.226 8.275 1 2 1 1 9649.856 9135.306 5.384 1 2 1 2 48891.256 45988.550 13.536 1 2 2 1 19925.670 18112.954 13.469 1 2 2 2 131637.580 136727.859 -13.766 2 1 1 1 6493.986 5252.910 17.124 2 1 1 2 40510.749 38167.691 11.993 2 1 2 1 15213.978 14473.690 6.153 2 1 2 2 154549.960 158881.328 -10.866 2 2 1 1 5987.049 5218.963 10.632 2 2 1 2 47293.848 51568.202 -18.823 2 2 2 1 16119.867 18938.392 -20.481 2 2 2 2 292541.200 286386.190 11.501 *** PSEUDO R-SQUARED MEASURES *** * P(A|X) * baseline fitted R-squared entropy 0.6807 0.5517 0.1896 qualitative variance 0.2438 0.1865 0.2350 classification error 0.4213 0.2781 0.3398 -2/N*log-likelihood 1.3614 1.1033 0.1896/0.2051 likelihood^(-2/N) 3.9017 3.0142 0.2275/0.3059 * P(B|X) * baseline fitted R-squared entropy 0.6827 0.6470 0.0523 qualitative variance 0.2448 0.2278 0.0693 classification error 0.4280 0.3728 0.1289 -2/N*log-likelihood 1.3655 1.2940 0.0523/0.0667 likelihood^(-2/N) 3.9175 3.6475 0.0689/0.0926 * P(C|X) * baseline fitted R-squared entropy 0.5311 0.4875 0.0822 qualitative variance 0.1735 0.1581 0.0888 classification error 0.2234 0.2174 0.0267 -2/N*log-likelihood 1.0623 0.9750 0.0822/0.0803 likelihood^(-2/N) 2.8929 2.6511 0.0836/0.1278 * P(D|X) * baseline fitted R-squared entropy 0.3408 0.3101 0.0903 qualitative variance 0.0958 0.0891 0.0699 classification error 0.1073 0.1068 0.0042 -2/N*log-likelihood 0.6817 0.6201 0.0903/0.0580 likelihood^(-2/N) 1.9772 1.8591 0.0597/0.1208 *** LOG-LINEAR PARAMETERS *** * TABLE XA [or P(A|X)] * effect beta std err z-value exp(beta) Wald df prob A 1 -0.2084 0.0015 -137.328 0.8119 2 0.2084 1.2317 18859.04 1 0.000 spe(A,1a) [X 1] 1 -0.6304 0.0057 -111.039 0.5324 12329.62 1 0.000 * TABLE XB [or P(B|X)] * effect beta std err z-value exp(beta) Wald df prob B 1 -0.1561 0.0011 -142.108 0.8555 2 0.1561 1.1689 20194.58 1 0.000 spe(B,1a) [X 1] 1 -0.2851 0.0023 -123.675 0.7520 15295.59 1 0.000 * TABLE XC [or P(C|X)] * effect beta std err z-value exp(beta) Wald df prob C 1 -0.6968 0.0017 -411.125 0.4982 2 0.6968 2.0072 169024.14 1 0.000 spe(C,1a) [X 1] 1 -0.3810 0.0030 -126.860 0.6832 16093.45 1 0.000 * TABLE XD [or P(D|X)] * effect beta std err z-value exp(beta) Wald df prob D 1 -1.1917 0.0027 -435.400 0.3037 2 1.1917 3.2925 189573.47 1 0.000 spe(D,1a) [X 1] 1 -0.4246 0.0036 -116.340 0.6540 13534.90 1 0.000 *** LATENT CLASS OUTPUT *** X 1 X 2 X 3 X 4 X 5 X 6 X 7 X 8 0.0000 0.0028 0.0499 0.2441 0.4063 0.2441 0.0499 0.0028 A 1 0.9949 0.9740 0.9004 0.7054 0.3973 0.1536 0.0459 0.0114 A 2 0.0051 0.0260 0.0996 0.2946 0.6027 0.8464 0.9541 0.9886 B 1 0.9056 0.8198 0.7051 0.5674 0.4226 0.2900 0.1830 0.1053 B 2 0.0944 0.1802 0.2949 0.4326 0.5774 0.7100 0.8170 0.8947 C 1 0.8855 0.7406 0.5471 0.3512 0.1988 0.1022 0.0485 0.0211 C 2 0.1145 0.2594 0.4529 0.6488 0.8012 0.8978 0.9515 0.9789 D 1 0.8099 0.5839 0.3499 0.1803 0.0845 0.0372 0.0156 0.0060 D 2 0.1901 0.4161 0.6501 0.8197 0.9155 0.9628 0.9844 0.9940 X 9 0.0000 A 1 0.0022 A 2 0.9978 B 1 0.0529 B 2 0.9471 C 1 0.0079 C 2 0.9921 D 1 0.0020 D 2 0.9980 E = 0.5302, lambda = 0.1069