My Input: a type 2 association (RC model) with husband’s education and wife’s education:

ass2, with model 5a. Note that model 5e (which is supposed to yield unconstrained scores for row and col) yields exactly the same answer. I don’t know why.

 

Input:

 

man 2

dim 4 4

lab H W

mod {H,W,ass2(H,W,5a)}

dat [32016 33374 8407 988

28370 137876 43783 8446

7051 48766 61633 18195

984 13794 28635 51224]

 

 

Output:

 

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 4 4

  lab H W

  mod {H,W,ass2(H,W,5a)}

  dat [32016 33374 8407 988

  28370 137876 43783 8446

  7051 48766 61633 18195

  984 13794 28635 51224]

 

 

 

*** STATISTICS ***

 

  Number of iterations = 62

  Converge criterion   = 0.0000009227

 

  X-squared            = 11374.7967 (0.0000)

  L-squared            = 10528.7995 (0.0000)

  Cressie-Read         = 11029.4300 (0.0000)

  Dissimilarity index  = 0.0622

  Degrees of freedom   = 4

  Log-likelihood       = -1235422.80534

  Number of parameters = 11 (+1)

  Sample size          = 523542.0

  BIC(L-squared)       = 10476.1260

  AIC(L-squared)       = 10520.7995

  BIC(log-likelihood)  = 2470990.4628

  AIC(log-likelihood)  = 2470867.6107

 

WARNING: no information is provided on identification of parameters

 

 

 

*** FREQUENCIES ***

 

  H W     observed  estimated  std. res.

  1 1   32016.000  29323.716     15.722

  1 2   33374.000  38493.237    -26.092

  1 3    8407.000   6653.414     21.498

  1 4     988.000    314.697     37.955

  2 1   28370.000  33184.593    -26.430

  2 2  137876.000 125988.448     33.491

  2 3   43783.000  51752.316    -35.031

  2 4    8446.000   7549.708     10.315

  3 1    7051.000   5547.488     20.186

  3 2   48766.000  56545.669    -32.716

  3 3   61633.000  51950.996     42.478

  3 4   18195.000  21600.815    -23.173

  4 1     984.000    365.203     32.380

  4 2   13794.000  12782.646      8.945

  4 3   28635.000  32101.274    -19.346

  4 4   51224.000  49387.781      8.263

 

 

 

*** LOG-LINEAR PARAMETERS ***

 

* TABLE HW [or P(HW)] *

 

  effect           beta   exp(beta)

  main           9.6424  1.54E+0004

  H

   1            -0.7927      0.4526

   2             0.8419      2.3207

   3             0.4582      1.5812

   4            -0.5073      0.6021

  W

   1            -0.8381      0.4326

   2             1.0328      2.8088

   3             0.5805      1.7870

   4            -0.7752      0.4606

 

  type 2 association (row=H column=W)

  association    5.1277

  row           -0.6590  -0.2212   0.1859   0.6944

  adj row       -1.4923  -0.5010   0.4209   1.5724

  column        -0.6731  -0.2000   0.1856   0.6874

  adj column    -1.5241  -0.4528   0.4204   1.5566

 

 

 

*** (CONDITIONAL) PROBABILITIES ***

 

* P(HW) *

 

  1 1            0.0560

  1 2            0.0735

  1 3            0.0127

  1 4            0.0006

  2 1            0.0634

  2 2            0.2406

  2 3            0.0989

  2 4            0.0144

  3 1            0.0106

  3 2            0.1080

  3 3            0.0992

  3 4            0.0413

  4 1            0.0007

  4 2            0.0244

  4 3            0.0613

  4 4            0.0943

 

-----------------------------------------------------------------------

 

*And now, something different, using LEM to estimate the uniform association, or linear by linear association model. Notice that the residual df and the LRT chisquare are exactly as I calculated them from STATA, but the association parameter here is listed differently, so it is scaled differently in LEM.

ass2, with model type 2a.

 

 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 4 4

  lab H W

  mod {H,W,ass2(H,W,2a)}

  dat [32016 33374 8407 988

  28370 137876 43783 8446

  7051 48766 61633 18195

  984 13794 28635 51224]

 

 

*** STATISTICS ***

 

  Number of iterations = 38

  Converge criterion   = 0.0000006657

 

  X-squared            = 13083.9241 (0.0000)

  L-squared            = 12561.3499 (0.0000)

  Cressie-Read         = 12863.9121 (0.0000)

  Dissimilarity index  = 0.0649

  Degrees of freedom   = 8

  Log-likelihood       = -1236439.08052

  Number of parameters = 7 (+1)

  Sample size          = 523542.0

  BIC(L-squared)       = 12456.0029

  AIC(L-squared)       = 12545.3499

  BIC(log-likelihood)  = 2472970.3396

  AIC(log-likelihood)  = 2472892.1610

 

WARNING: no information is provided on identification of parameters

 

 

 

*** FREQUENCIES ***

 

  H W     observed  estimated  std. res.

  1 1   32016.000  28854.715     18.610

  1 2   33374.000  40075.847    -33.478

  1 3    8407.000   5506.474     39.088

  1 4     988.000    347.959     34.312

  2 1   28370.000  33991.074    -30.489

  2 2  137876.000 128324.718     26.663

  2 3   43783.000  47927.051    -18.929

  2 4    8446.000   8232.188      2.357

  3 1    7051.000   5110.281     27.148

  3 2   48766.000  52440.848    -16.047

  3 3   61633.000  53237.739     36.385

  3 4   18195.000  24856.140    -42.250

  4 1     984.000    464.931     24.073

  4 2   13794.000  12968.588      7.248

  4 3   28635.000  35786.736    -37.805

  4 4   51224.000  45416.712     27.250

 

 

 

*** LOG-LINEAR PARAMETERS ***

 

* TABLE HW [or P(HW)] *

 

  effect           beta   exp(beta)

  main           9.6597  1.56E+0004

  H

   1            -0.8261      0.4378

   2             0.8377      2.3110

   3             0.4428      1.5570

   4            -0.4544      0.6348

  W

   1            -0.8135      0.4433

   2             1.0150      2.7593

   3             0.5301      1.6991

   4            -0.7316      0.4812

 

  type 2 association (row=H column=W)

  association    4.9998

 

 

 

*** (CONDITIONAL) PROBABILITIES ***

 

* P(HW) *

 

  1 1            0.0551

  1 2            0.0765

  1 3            0.0105

  1 4            0.0007

  2 1            0.0649

  2 2            0.2451

  2 3            0.0915

  2 4            0.0157

  3 1            0.0098

  3 2            0.1002

  3 3            0.1017

  3 4            0.0475

  4 1            0.0009

  4 2            0.0248

  4 3            0.0684

  4 4            0.0867