UCLA Academic Technology Services HomeServicesClassesContactJobs
Search

Stata Textbook Examples
An Introduction to Categorical Analysis by Alan Agresti
Chapter 3: Three-Way Contingency Tables

Table 3.1, page 54. The command tab3way can be downloaded from the internet by typing findit tab3way in the command line (see How can I use the findit command to search for programs and get additional help? for more information about using findit).
use http://www.ats.ucla.edu/stat/stata/examples/icda/deathp, clear

tab3way victim deathp defendant [fw=count], coltot rowp

Frequency weights are based on the expression: count
Table entries are cell frequencies and row percentages
Missing categories ignored

--------------------------------------------
          |       defendant and deathp      
          | ------ 1 -----    ------ 2 -----
   victim |      0       1         0       1
----------+---------------------------------
        1 |    414      53        37      11
          |  88.65   11.35     77.08   22.92
          | 
        2 |     16               139       4
          | 100.00             97.20    2.80
          | 
    TOTAL |    430      53       176      15
          |  89.03   10.97     92.15    7.85
--------------------------------------------
Table 3.2 on page 58 and calculation on page 59.
use http://www.ats.ucla.edu/stat/stata/examples/icda/clinic, clear

tab3way treat  clinic response [fw=count] , rowtot

Frequency weights are based on the expression: count
Table entries are cell frequencies
Missing categories ignored
------------------------------------------------------
          |            response and clinic            
          | -------- 0 --------    -------- 1 --------
    treat |     0      1  TOTAL        0      1  TOTAL
----------+-------------------------------------------
        0 |    18      2     20       12      8     20
        1 |    12      8     20        8     32     40
------------------------------------------------------

cc response treat [fw=count], by(clinic)

          clinic |       OR      [95% Conf. Interval]    M-H Weight
-----------------+-------------------------------------------------
               0 |          1      .266931    3.66141          2.88 (exact)
               1 |          1     .0872055   6.599538          1.28 (exact)
-----------------+-------------------------------------------------
           Crude |          2     .8134474   4.921018               (exact)
    M-H combined |          1     .3825288   2.614182               
-------------------------------------------------------------------
Test of homogeneity (M-H)      chi2(1) =     0.00  Pr>chi2 = 1.0000
                   Test that combined OR = 1:
                                Mantel-Haenszel chi2(1) =      0.00
                                                Pr>chi2 =    1.0000
                                                
cc treat clinic [fw=count], by(response)

        response |       OR      [95% Conf. Interval]    M-H Weight
-----------------+-------------------------------------------------
               0 |          6     .9222592   64.61686            .6 (exact)
               1 |          6     1.587285   23.15472      1.066667 (exact)
-----------------+-------------------------------------------------
           Crude |          6     2.264456   16.37396               (exact)
    M-H combined |          6     2.259288   15.93423               
-------------------------------------------------------------------
Test of homogeneity (M-H)      chi2(1) =     0.00  Pr>chi2 = 1.0000
                   Test that combined OR = 1:
                                Mantel-Haenszel chi2(1) =     14.07
                                                Pr>chi2 =    0.0002
                                                
use http://www.ats.ucla.edu/stat/stata/examples/icda/deathp, clear

cc defendant deathp [fw=count], by(victim)

          victim |       OR      [95% Conf. Interval]    M-H Weight
-----------------+-------------------------------------------------
                 |   .4306105     .2003285   .9955573      8.842718 (exact)
           white |          0            0   8.894517      .4025157 (exact)
-----------------+-------------------------------------------------
           Crude |   1.446202     .7776757   2.836998               (exact)
    M-H combined |   .4118627     .1991173   .8519141               
-------------------------------------------------------------------
Test of homogeneity (B-D)      chi2(1) =     0.20  Pr>chi2 = 0.6517
                   Test that combined OR = 1:
                                Mantel-Haenszel chi2(1) =      5.80
                                                Pr>chi2 =    0.0161
Section 3.2, page 60. Cochran-Mantel-Haenszel Methods using Chinese smoking and lung cancer data set.
use http://www.ats.ucla.edu/stat/stata/examples/icda/cancer, clear

cc cancer smoke [fw=count], by(center) 

          center |       OR      [95% Conf. Interval]    M-H Weight
-----------------+-------------------------------------------------
               1 |      2.196     1.307112   3.710289      10.86957 (exact)
               2 |   2.142962     1.840433   2.495332       117.909 (exact)
               3 |   2.175265     1.838162   2.574387      96.75867 (exact)
               4 |   2.850341     1.939615   4.204899      17.02389 (exact)
               5 |   2.319148     1.759581   3.060508      35.62906 (exact)
               6 |   1.587963     1.076653   2.344985      22.11024 (exact)
               7 |   2.369146     1.092397   5.476443      5.112676 (exact)
               8 |    2.00321      1.04818   3.883171         7.476 (exact)
-----------------+-------------------------------------------------
           Crude |   2.135552     1.948418   2.340687               (exact)
    M-H combined |   2.174482     1.984002   2.383249               
-------------------------------------------------------------------
Test of homogeneity (M-H)      chi2(7) =     5.19  Pr>chi2 = 0.6372
                   Test that combined OR = 1:
                                Mantel-Haenszel chi2(1) =    280.14
                                                Pr>chi2 =    0.0000

How to cite this page

Report an error on this page

UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services


The content of this web site should not be construed as an endorsement of any particular web site, book, or software product by the University of California