RCII

Powers and Xie(2000)のRCIIモデルをLEMで。Table 4.16とTable4.17より。

LEM: log-linear and event history analysis with missing data.

Developed by Jeroen Vermunt (c), Tilburg University, The Netherlands.

Version 1.0 (September 18, 1997).

*** INPUT ***

man 2

dim 5 4

lab A P

mod {A P ass2(A,P,5a) }

dat[

44 11 38 62

59 41 147 293

23 11 13 27

27 8 16 27

258 57 105 110

]

nfr

nco

*** STATISTICS ***

Number of iterations = 11

Converge criterion = 0.0000001579

X-squared = 5.4457 (0.4880)

L-squared = 5.5574 (0.4745)

Cressie-Read = 5.4746 (0.4845)

Dissimilarity index = 0.0165

Degrees of freedom = 6

Log-likelihood = -3448.92938

Number of parameters = 13 (+1)

Sample size = 1377.0

BIC(L-squared) = -37.8085

AIC(L-squared) = -6.4426

BIC(log-likelihood) = 6991.8184

AIC(log-likelihood) = 6923.8588

WARNING: no information is provided on identification of parameters

*** LOG-LINEAR PARAMETERS ***

* TABLE AP [or P(AP)] *

effect beta exp(beta)

main 3.7197 41.2520

A

1 -0.1679 0.8455

2 0.8917 2.4393

3 -0.8955 0.4084

4 -0.8426 0.4306

5 1.0143 2.7573

P

1 0.2428 1.2748

2 -0.7965 0.4509

3 0.0835 1.0871

4 0.4702 1.6003

type 2 association (row=A column=P)

association 1.3083

row 0.0753 0.7763 -0.0983 -0.1550 -0.5983

adj row 0.0861 0.8880 -0.1124 -0.1773 -0.6843

column -0.7432 -0.1265 0.2713 0.5984

adj column -0.8501 -0.1447 0.3103 0.6844