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

log:  C:\AAA Miker Files\newer web pages\soc_388_notes\soc_388_2007\class_nine_log.log

log type:  text

opened on:  23 Oct 2007, 11:08:28

. use "C:\AAA Miker Files\newer web pages\soc_388_notes\LA_intermar.dta", clear

. clear

. use "C:\AAA Miker Files\newer web pages\soc_388_notes\LA intermar class 5.dta", clear

. *Let me just show you something quickly about how to test parameters, using the HW2 dataset, LA intermarriage

.  set linesize 79

. desmat: poisson count  hed wed interfull

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

Poisson regression

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

Dependent variable                                                    count

Optimization:                                                            ml

Number of observations:                                                  16

Initial log likelihood:                                         -221501.223

Log likelihood:                                                  -24059.274

LR chi square:                                                   394883.898

Model degrees of freedom:                                                10

Pseudo R-squared:                                                     0.891

Prob:                                                                 0.000

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

nr Effect                                                    Coeff        s.e.

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

count

hed

1      2                                                     1.134**     0.007

2      3                                                     0.819**     0.006

3      4                                                    -0.017*      0.007

wed

4      2                                                     1.372**     0.007

5      3                                                     1.020**     0.007

6      4                                                    -0.278**     0.008

interfull

7      1                                                     1.722**     0.009

8      2                                                     0.676**     0.007

9      3                                                     0.537**     0.008

10     4                                                     2.487**     0.009

11   _cons                                                   8.652**     0.008

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

*  p < .05

** p < .01

. *How to test whether one coefficient is significantly different from another.

> .

. lincom _x_8-_x_9

( 1)  [count]_x_8 - [count]_x_9 = 0

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

count |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

(1) |    .139055   .0122419    11.36   0.000     .1150614    .1630486

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

. *significant difference between intermarriage for category 2 and 3

. test _x_8-_x_9=0

( 1)  [count]_x_8 - [count]_x_9 = 0

chi2(  1) =  129.03

Prob > chi2 =    0.0000

. *Are these tests the same?

. display 11.36^2

129.0496

. *something we will talk about later in the class, is why the chisquare (1) is really just the square of a standard Normal distribution. But here in any event, you can see that the two tests (lincom and test) provide the same information.

. clear

. *now on to use the Qian dataset.

. use "C:\AAA Miker Files\newer web pages\soc_388_notes\Qian 80-90 intermar.dta ", clear

. desmat: poisson count  mfulleth med4 ffulleth fed4 year

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

Poisson regression

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

Dependent variable                                                    count

Optimization:                                                            ml

Number of observations:                                                 512

Initial log likelihood:                                        -1402408.286

Log likelihood:                                                 -289165.150

LR chi square:                                                  2226486.273

Model degrees of freedom:                                                13

Pseudo R-squared:                                                     0.794

Prob:                                                                 0.000

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

nr Effect                                                    Coeff        s.e.

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

count

mfulleth

1      Hisp                                                  3.431**     0.037

2      black                                                 3.894**     0.037

3      white                                                 6.437**     0.037

med4

4      2                                                     1.072**     0.004

5      3                                                     0.595**     0.005

6      4                                                     0.235**     0.005

ffulleth

7      Hisp                                                  3.262**     0.035

8      black                                                 3.695**     0.034

9      white                                                 6.281**     0.034

fed4

10     2                                                     1.229**     0.004

11     3                                                     0.733**     0.005

12     4                                                     0.142**     0.005

year

13     90                                                   -0.415**     0.003

14   _cons                                                  -4.279**     0.050

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

*  p < .05

** p < .01

. poisgof

Goodness-of-fit chi2  =  576133.3

Prob > chi2(498)      =    0.0000

. table mfulleth ffulleth, contents (mean  r_endog)

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

|          ffulleth

mfulleth | Asian   Hisp  black  white

----------+---------------------------

Asian |     1      0      0      0

Hisp |     0      2      0      0

black |     0      0      3      0

white |     0      0      0      4

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

. desmat: poisson count  mfulleth*med4 ffulleth*fed4 year  r_endog

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

Poisson regression

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

Dependent variable                                                    count

Optimization:                                                            ml

Number of observations:                                                 512

Initial log likelihood:                                        -1402408.286

Log likelihood:                                                 -123146.804

LR chi square:                                                  2558522.965

Model degrees of freedom:                                                35

Pseudo R-squared:                                                     0.912

Prob:                                                                 0.000

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

nr Effect                                                    Coeff        s.e.

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

count

mfulleth

1      Hisp                                                  5.526**     0.222

2      black                                                 3.649**     0.223

3      white                                                 6.035**     0.224

med4

4      2                                                     1.846**     0.235

5      3                                                     2.461**     0.227

6      4                                                     2.790**     0.225

mfulleth.med4

7      Hisp.2                                               -1.506**     0.235

8      Hisp.3                                               -2.624**     0.228

9      Hisp.4                                               -4.150**     0.226

10     black.2                                              -0.972**     0.235

11     black.3                                              -2.149**     0.228

12     black.4                                              -3.553**     0.226

13     white.2                                              -0.701**     0.235

14     white.3                                              -1.786**     0.227

15     white.4                                              -2.413**     0.225

ffulleth

16     Hisp                                                  4.831**     0.175

17     black                                                 1.810**     0.178

18     white                                                 5.377**     0.177

fed4

19     2                                                     1.705**     0.186

20     3                                                     2.105**     0.182

21     4                                                     2.387**     0.179

ffulleth.fed4

22     Hisp.2                                               -1.235**     0.187

23     Hisp.3                                               -2.235**     0.183

24     Hisp.4                                               -3.838**     0.182

25     black.2                                              -0.708**     0.187

26     black.3                                              -1.550**     0.182

27     black.4                                              -2.896**     0.180

28     white.2                                              -0.400*      0.186

29     white.3                                              -1.297**     0.182

30     white.4                                              -2.123**     0.179

year

31     90                                                   -0.415**     0.003

r_endog

32     1                                                     4.497**     0.089

33     2                                                     2.107**     0.042

34     3                                                     6.914**     0.050

35     4                                                     2.871**     0.042

36   _cons                                                  -5.892**     0.282

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

*  p < .05

** p < .01

. poisgof

Goodness-of-fit chi2  =  244096.6

Prob > chi2(476)      =    0.0000

. *For this class I am not bothering with deviation coding for the variables because I don't care about the coefficients today, I just want to look at goodness of fit and number of terms.

. desmat: poisson count  mfulleth*med4*fed4 ffulleth*med4*fed4 year  r_endog

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

Poisson regression

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

Dependent variable                                                    count

Optimization:                                                            ml

Number of observations:                                                 512

Initial log likelihood:                                        -1402408.286

Log likelihood:                                                  -10672.055

LR chi square:                                                  2783472.463

Model degrees of freedom:                                               116

Pseudo R-squared:                                                     0.992

Prob:                                                                 0.000

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

nr Effect                                                    Coeff        s.e.

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

count

mfulleth

1      Hisp                                                  5.374**     0.382

2      black                                                 3.227**     0.388

3      white                                                 5.893**     0.383

med4

4      2                                                     0.639       0.566

5      3                                                     0.651       0.620

6      4                                                    -1.406       1.305

mfulleth.med4

7      Hisp.2                                               -1.193*      0.470

8      Hisp.3                                               -1.907**     0.521

9      Hisp.4                                               -2.073       1.135

10     black.2                                              -0.650       0.479

11     black.3                                              -1.098*      0.538

12     black.4                                              -2.042       1.203

13     white.2                                              -0.818       0.469

14     white.3                                              -1.637**     0.519

15     white.4                                              -1.278       1.125

fed4

16     2                                                     0.343       0.584

17     3                                                    -1.335       1.119

18     4                                                    -0.738       1.116

mfulleth.fed4

19     Hisp.2                                               -0.881       0.482

20     Hisp.3                                                0.524       1.080

21     Hisp.4                                               -1.202       1.149

22     black.2                                              -0.600       0.491

23     black.3                                               0.987       1.088

24     black.4                                              -0.922       1.205

25     white.2                                              -0.568       0.481

26     white.3                                               0.754       1.079

27     white.4                                              -0.811       1.141

med4.fed4

28     2.2                                                   1.901**     0.699

29     2.3                                                   3.225**     1.189

30     2.4                                                   0.520       1.283

31     3.2                                                   1.712*      0.749

32     3.3                                                   4.484**     1.207

33     3.4                                                   3.269**     1.208

34     4.2                                                   2.680       1.394

35     4.3                                                   6.282**     1.667

36     4.4                                                   6.928**     1.661

mfulleth.med4.fed4

37     Hisp.2.2                                              0.506       0.567

38     Hisp.2.3                                             -1.314       1.127

39     Hisp.2.4                                              0.755       1.294

40     Hisp.3.2                                              0.613       0.616

41     Hisp.3.3                                             -1.314       1.141

42     Hisp.3.4                                              0.316       1.215

43     Hisp.4.2                                             -0.458       1.191

44     Hisp.4.3                                             -1.875       1.525

45     Hisp.4.4                                             -0.390       1.572

46     black.2.2                                             0.361       0.579

47     black.2.3                                            -1.549       1.138

48     black.2.4                                             0.944       1.351

49     black.3.2                                             0.335       0.635

50     black.3.3                                            -1.734       1.155

51     black.3.4                                             0.284       1.275

52     black.4.2                                             0.189       1.264

53     black.4.3                                            -1.352       1.583

54     black.4.4                                             0.055       1.662

55     white.2.2                                             0.625       0.566

56     white.2.3                                            -1.074       1.125

57     white.2.4                                             1.284       1.284

58     white.3.2                                             0.746       0.613

59     white.3.3                                            -1.024       1.139

60     white.3.4                                             0.648       1.205

61     white.4.2                                            -0.325       1.180

62     white.4.3                                            -1.472       1.516

63     white.4.4                                            -0.034       1.558

ffulleth

64     Hisp                                                  4.401**     0.271

65     black                                                 1.628**     0.280

66     white                                                 5.213**     0.271

ffulleth.med4

67     Hisp.2                                               -0.039       0.401

68     Hisp.3                                               -0.558       0.485

69     Hisp.4                                               -0.780       1.101

70     black.2                                              -0.173       0.411

71     black.3                                              -1.091*      0.503

72     black.4                                              -0.072       1.173

73     white.2                                               0.115       0.400

74     white.3                                              -0.493       0.482

75     white.4                                              -0.721       1.092

ffulleth.fed4

76     Hisp.2                                               -0.016       0.409

77     Hisp.3                                               -1.071*      0.470

78     Hisp.4                                               -2.292**     0.695

79     black.2                                               0.278       0.419

80     black.3                                              -0.660       0.488

81     black.4                                              -1.463       0.781

82     white.2                                               0.330       0.408

83     white.3                                              -0.742       0.468

84     white.4                                              -1.903**     0.679

ffulleth.med4.fed4

85     Hisp.2.2                                             -1.019*      0.514

86     Hisp.2.3                                             -0.410       0.574

87     Hisp.2.4                                              0.143       0.817

88     Hisp.3.2                                             -0.710       0.591

89     Hisp.3.3                                             -0.003       0.627

90     Hisp.3.4                                              0.140       0.816

91     Hisp.4.2                                             -0.028       1.186

92     Hisp.4.3                                             -0.313       1.174

93     Hisp.4.4                                             -0.076       1.276

94     black.2.2                                            -0.871       0.528

95     black.2.3                                            -0.213       0.595

96     black.2.4                                             0.357       0.903

97     black.3.2                                            -0.507       0.612

98     black.3.3                                             0.253       0.653

99     black.3.4                                             0.253       0.902

100    black.4.2                                            -1.177       1.260

101    black.4.3                                            -1.412       1.249

102    black.4.4                                            -0.946       1.385

103    white.2.2                                            -0.989       0.513

104    white.2.3                                            -0.348       0.570

105    white.2.4                                             0.519       0.799

106    white.3.2                                            -0.536       0.588

107    white.3.3                                             0.108       0.623

108    white.3.4                                             0.588       0.800

109    white.4.2                                             0.279       1.175

110    white.4.3                                            -0.019       1.164

111    white.4.4                                             0.625       1.259

year

112    90                                                   -0.415**     0.003

r_endog

113    1                                                     4.139**     0.095

114    2                                                     2.003**     0.043

115    3                                                     6.921**     0.050

116    4                                                     2.878**     0.042

117  _cons                                                  -4.370**     0.438

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

*  p < .05

** p < .01

. *Important note: desmat generates all the lower order interactions for you, automatically, when you use the * to put variables together. Also note that these models are summarized in my comprehensive Excel file.

. poisgof

Goodness-of-fit chi2  =  19147.11

Prob > chi2(395)      =    0.0000

. *This Model still fits terribly, however, what do we think is missing?

. *What's missing? take the previous model, and add every other variable interacted with year...

. desmat: poisson count  mfulleth*med4*fed4 ffulleth*med4*fed4  mfulleth*year ffulleth*year med4*year fed4*year  r_endog

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

Poisson regression

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

Dependent variable                                                    count

Optimization:                                                            ml

Number of observations:                                                 512

Initial log likelihood:                                        -1402408.286

Log likelihood:                                                   -2532.653

LR chi square:                                                  2799751.267

Model degrees of freedom:                                               128

Pseudo R-squared:                                                     0.998

Prob:                                                                 0.000

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

nr Effect                                                    Coeff        s.e.

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

count

mfulleth

1      Hisp                                                  5.304**     0.384

2      black                                                 3.147**     0.389

3      white                                                 5.796**     0.384

med4

4      2                                                     0.672       0.566

5      3                                                     0.646       0.620

6      4                                                    -1.305       1.306

mfulleth.med4

7      Hisp.2                                               -1.186*      0.470

8      Hisp.3                                               -1.908**     0.521

9      Hisp.4                                               -2.053       1.135

10     black.2                                              -0.642       0.479

11     black.3                                              -1.099*      0.538

12     black.4                                              -2.018       1.203

13     white.2                                              -0.809       0.469

14     white.3                                              -1.639**     0.519

15     white.4                                              -1.249       1.125

fed4

16     2                                                     0.359       0.584

17     3                                                    -1.551       1.119

18     4                                                    -0.930       1.116

mfulleth.fed4

19     Hisp.2                                               -0.878       0.482

20     Hisp.3                                                0.492       1.080

21     Hisp.4                                               -1.231       1.149

22     black.2                                              -0.596       0.491

23     black.3                                               0.949       1.088

24     black.4                                              -0.956       1.205

25     white.2                                              -0.563       0.481

26     white.3                                               0.709       1.079

27     white.4                                              -0.852       1.141

med4.fed4

28     2.2                                                   1.900**     0.699

29     2.3                                                   3.246**     1.189

30     2.4                                                   0.539       1.283

31     3.2                                                   1.713*      0.749

32     3.3                                                   4.481**     1.207

33     3.4                                                   3.266**     1.208

34     4.2                                                   2.675       1.395

35     4.3                                                   6.350**     1.668

36     4.4                                                   6.989**     1.662

mfulleth.med4.fed4

37     Hisp.2.2                                              0.506       0.567

38     Hisp.2.3                                             -1.313       1.127

39     Hisp.2.4                                              0.756       1.294

40     Hisp.3.2                                              0.613       0.616

41     Hisp.3.3                                             -1.314       1.141

42     Hisp.3.4                                              0.316       1.215

43     Hisp.4.2                                             -0.459       1.191

44     Hisp.4.3                                             -1.870       1.525

45     Hisp.4.4                                             -0.386       1.571

46     black.2.2                                             0.361       0.579

47     black.2.3                                            -1.548       1.138

48     black.2.4                                             0.945       1.351

49     black.3.2                                             0.335       0.635

50     black.3.3                                            -1.734       1.155

51     black.3.4                                             0.284       1.275

52     black.4.2                                             0.188       1.263

53     black.4.3                                            -1.347       1.583

54     black.4.4                                             0.060       1.661

55     white.2.2                                             0.624       0.566

56     white.2.3                                            -1.073       1.125

57     white.2.4                                             1.285       1.284

58     white.3.2                                             0.746       0.613

59     white.3.3                                            -1.025       1.139

60     white.3.4                                             0.648       1.206

61     white.4.2                                            -0.325       1.180

62     white.4.3                                            -1.466       1.516

63     white.4.4                                            -0.029       1.558

ffulleth

64     Hisp                                                  4.270**     0.272

65     black                                                 1.615**     0.281

66     white                                                 5.104**     0.272

ffulleth.med4

67     Hisp.2                                               -0.028       0.401

68     Hisp.3                                               -0.560       0.485

69     Hisp.4                                               -0.742       1.101

70     black.2                                              -0.172       0.411

71     black.3                                              -1.092*      0.503

72     black.4                                              -0.070       1.172

73     white.2                                               0.125       0.399

74     white.3                                              -0.495       0.482

75     white.4                                              -0.690       1.092

ffulleth.fed4

76     Hisp.2                                               -0.011       0.409

77     Hisp.3                                               -1.136*      0.470

78     Hisp.4                                               -2.350**     0.695

79     black.2                                               0.278       0.419

80     black.3                                              -0.667       0.488

81     black.4                                              -1.470       0.781

82     white.2                                               0.334       0.408

83     white.3                                              -0.796       0.468

84     white.4                                              -1.953**     0.679

ffulleth.med4.fed4

85     Hisp.2.2                                             -1.019*      0.514

86     Hisp.2.3                                             -0.408       0.574

87     Hisp.2.4                                              0.144       0.817

88     Hisp.3.2                                             -0.710       0.591

89     Hisp.3.3                                             -0.003       0.627

90     Hisp.3.4                                              0.140       0.816

91     Hisp.4.2                                             -0.030       1.185

92     Hisp.4.3                                             -0.304       1.174

93     Hisp.4.4                                             -0.068       1.276

94     black.2.2                                            -0.871       0.528

95     black.2.3                                            -0.213       0.595

96     black.2.4                                             0.357       0.903

97     black.3.2                                            -0.507       0.612

98     black.3.3                                             0.254       0.653

99     black.3.4                                             0.253       0.902

100    black.4.2                                            -1.178       1.260

101    black.4.3                                            -1.410       1.249

102    black.4.4                                            -0.944       1.385

103    white.2.2                                            -0.989       0.513

104    white.2.3                                            -0.346       0.570

105    white.2.4                                             0.520       0.799

106    white.3.2                                            -0.536       0.588

107    white.3.3                                             0.108       0.623

108    white.3.4                                             0.588       0.800

109    white.4.2                                             0.277       1.175

110    white.4.3                                            -0.011       1.163

111    white.4.4                                             0.632       1.259

year

112    90                                                   -1.120**     0.091

mfulleth.year

113    Hisp.90                                               0.204*      0.090

114    black.90                                              0.235*      0.097

115    white.90                                              0.281**     0.089

ffulleth.year

116    Hisp.90                                               0.388**     0.083

117    black.90                                              0.044       0.091

118    white.90                                              0.331**     0.082

med4.year

119    2.90                                                 -0.144**     0.009

120    3.90                                                  0.021*      0.010

121    4.90                                                 -0.504**     0.012

fed4.year

122    2.90                                                 -0.066**     0.010

123    3.90                                                  0.689**     0.011

124    4.90                                                  0.625**     0.013

r_endog

125    1                                                     4.120**     0.095

126    2                                                     2.005**     0.043

127    3                                                     6.920**     0.051

128    4                                                     2.876**     0.042

129  _cons                                                  -4.126**     0.439

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

*  p < .05

** p < .01

. poisgof

Goodness-of-fit chi2  =  2868.306

Prob > chi2(383)      =    0.0000

. *still poor fitting by LRT, but the BIC becomes negative here...

. tabulate med4 [fweight=count]

med4 |      Freq.     Percent        Cum.

------------+-----------------------------------

1 |     74,785       14.28       14.28

2 |    218,475       41.73       56.01

3 |    135,645       25.91       81.92

4 |     94,637       18.08      100.00

------------+-----------------------------------

Total |    523,542      100.00

. display ln(523542)

13.168373

. display 2868-(383*13.168)

-2175.344

. *One of the things we will talk more about on Thursday, is how to be systematic about whether you need a certain set of interactions, or not. That is, howto go about building models so that you start to have some confidence that you have the important stuff in there.

. exit, clear