---------------------------------------------------------------------------------------------------------------------------
name: <unnamed>
log: C:\Users\dupontwd.VANDERBILT\Documents\Dropbox\OFFICE_desktop\ChoiFisherTest\Stata Program\ado\ForSubmission\c
> hoi_lr_test\test.log
log type: text
opened on: 31 May 2016, 14:08:46
. * test.log
.
. * This program tests choi_lr_test.ado and choi_lr_testi.ado
. * Ancillary files that are needed by these programs are
. * nl_choi_support_interval.ado, choi_lr_support_interval.ado
. * choi_lr_hypergeom.ado and choi_lr_hyperg_prog.ado.
.
. * These programs calculate statistics described in
. * Choi et al. 2015. Elucidating the Foundations of Statistical Inference with 2 x 2 Tables.
. * PLoS ONE 10(4): e0121263. doi:10.1371/journal.pone.0121263
.
. * Output from this program may be checked against the Shiny app posted at
. * http://statcomp2.vanderbilt.edu:37212/ProfileLikelihood/
. * or against the R program posted at
. * https://cran.r-project.org/web/packages/ProfileLikelihood/index.html
.
. set more off
. use testdata.dta
.
. * Test choi_lr_test program
. choi_lr_test case exposed [fweight=patients]
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
.
. * Check scalars returned by the cc program
. di r(p)
.13057002
. di r(or)
5
. di r(lb_or)
.40865859
. di r(ub_or)
75.971185
. di r(afe)
.8
. di r(lb_afe)
-1.4470304
. di r(ub_afe)
.98683711
. di r(afp)
.6
. di r(chi2)
2.2857143
.
. * Check scalars calculated by the choi_lr_test program
. di r(or_cond)
4.4742438
. di r(clr)
2.9944571
. di r(chi2_clr)
2.1935259
. di r(p_choi)
.13859174
. di r(lr6pt8lsi_lb)
.62397643
. di r(lr6pt8lsi_ub)
42.991698
.
. choi_lr_test case exposed [fweight=patients], exact k(32)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
1-sided Fisher's exact P = 0.1573
2-sided Fisher's exact P = 0.3147
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
1/32 LSI for the OR = [.32148657, 109.60039]
. * Check scalars returned by the cc program
. di r(p1_exact)
.15734266
. di r(p_exact)
.31468531
.
. * Check that extra LSIs are returned
. di r(lrklsi_lb)
.32148657
. di r(lrklsi_ub)
109.60039
.
. choi_lr_test case exposed [fweight=patients], corn
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .6433784 38.07362 (Cornfield)
Attr. frac. ex. | .8 | -.5542952 .9737351 (Cornfield)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
. choi_lr_test case exposed [fweight=patients], woolf
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .5841514 42.79712 (Woolf)
Attr. frac. ex. | .8 | -.7118849 .9766339 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
.
. * Check behavior for tables with larger sample sizes
. replace patients = patients*12
(4 real changes made)
. choi_lr_test case exposed [fweight=patients], woolf k(16)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 72 24 | 96 0.7500
Controls | 36 60 | 96 0.3750
-----------------+------------------------+------------------------
Total | 108 84 | 192 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 2.690274 9.292735 (Woolf)
Attr. frac. ex. | .8 | .6282906 .8923891 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 27.43 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9539211
Likelihood ratio under the null hypothesis = 1213489.1
Likelihood ratio chi-squared statistic = 28.018021
P value for the null hypothesis that the OR equals 1 = 1.202e-07
1/6.825936 LSI for the OR = [2.7020791, 9.3208557]
1/16 LSI for the OR = [2.3971936, 10.628488]
.
. use testdata.dta, clear
. replace patients = patients*30
(4 real changes made)
. choi_lr_test case exposed [fweight=patients], woolf k(32)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 180 60 | 240 0.7500
Controls | 90 150 | 240 0.3750
-----------------+------------------------+------------------------
Total | 270 210 | 480 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 3.378545 7.399635 (Woolf)
Attr. frac. ex. | .8 | .7040146 .8648582 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 68.57 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9815164
Likelihood ratio under the null hypothesis = 1.821e+15
Likelihood ratio chi-squared statistic = 70.276267
P value for the null hypothesis that the OR equals 1 = 5.155e-17
1/6.825936 LSI for the OR = [3.384192, 7.409481]
1/32 LSI for the OR = [2.9695546, 8.5152491]
.
. use testdata.dta, clear
. replace patients = patients*100
(4 real changes made)
. choi_lr_test case exposed [fweight=patients], woolf k(8)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 600 200 | 800 0.7500
Controls | 300 500 | 800 0.3750
-----------------+------------------------+------------------------
Total | 900 700 | 1600 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 4.033904 6.19747 (Woolf)
Attr. frac. ex. | .8 | .7521012 .8386438 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 228.57 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9944476
Likelihood ratio under the null hypothesis = 8.826e+50
Likelihood ratio chi-squared statistic = 234.6139
P value for the null hypothesis that the OR equals 1 = 5.877e-53
1/6.825936 LSI for the OR = [4.035843, 6.2001042]
1/8 LSI for the OR = [4.0014063, 6.2550805]
.
. use testdata.dta, clear
. replace patients = patients*200
(4 real changes made)
. choi_lr_test case exposed [fweight=patients], woolf k(8)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 1200 400 | 1600 0.7500
Controls | 600 1000 | 1600 0.3750
-----------------+------------------------+------------------------
Total | 1800 1400 | 3200 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 4.295722 5.819743 (Woolf)
Attr. frac. ex. | .8 | .7672103 .8281711 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 457.14 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9972227
Likelihood ratio under the null hypothesis = 8.41e+101
Likelihood ratio chi-squared statistic = 469.38195
P value for the null hypothesis that the OR equals 1 = 4.37e-104
1/6.825936 LSI for the OR = [4.2967386, 5.8210048]
1/8 LSI for the OR = [4.2706733, 5.8572896]
.
. * Check if program handles 1 record per patient properly
. use testdata.dta, clear
. expand patients
(12 observations created)
. choi_lr_test case exposed
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
.
. * Check behavior for tables with 0 patients in some cells
. use testdata.dta,clear
. list
+---------------------------+
| case exposed patients |
|---------------------------|
1. | 0 0 5 |
2. | 0 1 3 |
3. | 1 0 2 |
4. | 1 1 6 |
+---------------------------+
. replace patients =0 if _n==3
(1 real change made)
. replace patients =0 if _n==4
(1 real change made)
. choi_lr_test case exposed [fweight=patients]
Too many empty cells to estimate OR
Bad data input
. replace patients =2 if _n==3
(1 real change made)
. choi_lr_test case exposed [fweight=patients] , k(16)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 0 2 | 2 0.0000
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 3 7 | 10 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 0 | 0 4.841502 (Cornfield)
Prev. frac. ex. | 1 | -3.841502 1 (Cornfield)
Prev. frac. pop | . |
+-------------------------------------------------
chi2(1) = 1.07 Pr>chi2 = 0.3006
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 0
Likelihood ratio under the null hypothesis = 2.1428571
Likelihood ratio chi-squared statistic = 1.5242801
P value for the null hypothesis that the OR equals 1 = .21697293
1/6.825936 LSI for the OR = [0, 3.7822765]
1/16 LSI for the OR = [0, 7.3282039]
. replace patients =2 if _n==1
(1 real change made)
. replace patients =0 if _n==2
(1 real change made)
. replace patients =5 if _n==3
(1 real change made)
. replace patients =3 if _n==4
(1 real change made)
. choi_lr_test case exposed [fweight=patients] , k(16)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 3 5 | 8 0.3750
Controls | 0 2 | 2 0.0000
-----------------+------------------------+------------------------
Total | 3 7 | 10 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | . | .2065475 . (Cornfield)
Attr. frac. ex. | 1 | -3.841502 . (Cornfield)
Attr. frac. pop | .375 |
+-------------------------------------------------
chi2(1) = 1.07 Pr>chi2 = 0.3006
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = infinity
Likelihood ratio under the null hypothesis = 2.1428571
Likelihood ratio chi-squared statistic = 1.5242801
P value for the null hypothesis that the OR equals 1 = .21697293
1/6.825936 LSI for the OR = [.26439104, infinity]
1/16 LSI for the OR = [.13645908, infinity]
. replace patients = patients*100
(3 real changes made)
. choi_lr_test case exposed [fweight=patients] , k(8)
| exposed | Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 300 500 | 800 0.3750
Controls | 0 200 | 200 0.0000
-----------------+------------------------+------------------------
Total | 300 700 | 1000 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | . | 31.2292 . (Cornfield)
Attr. frac. ex. | 1 | .9679787 . (Cornfield)
Attr. frac. pop | .375 |
+-------------------------------------------------
chi2(1) = 107.14 Pr>chi2 = 0.0000
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = infinity
Likelihood ratio under the null hypothesis = 2.629e+35
Likelihood ratio chi-squared statistic = 163.11415
P value for the null hypothesis that the OR equals 1 = 2.362e-37
1/6.825936 LSI for the OR = [61.74006, infinity]
1/8 LSI for the OR = [56.981507, infinity]
. * Test choi_lr_testi program
. choi_lr_testi 6 2 3 5
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
.
. * Check scalars returned by the cci program
. di r(p)
.13057002
. di r(or)
5
. di r(lb_or)
.40865859
. di r(ub_or)
75.971185
. di r(afe)
.8
. di r(lb_afe)
-1.4470304
. di r(ub_afe)
.98683711
. di r(afp)
.6
. di r(chi2)
2.2857143
.
. * Check scalars calculated and returned by the choi_lr_testi program
. di r(or_cond)
4.4742438
. di r(clr)
2.9944571
. di r(chi2_clr)
2.1935259
. di r(p_choi)
.13859174
. di r(lr6pt8lsi_lb)
.62397643
. di r(lr6pt8lsi_ub)
42.991698
.
. choi_lr_testi 6 2 3 5, exact k(8)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
1-sided Fisher's exact P = 0.1573
2-sided Fisher's exact P = 0.3147
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
1/8 LSI for the OR = [.57712034, 47.72445]
.
. * Check scalars returned by the cci program
. di r(p1_exact)
.15734266
. di r(p_exact)
.31468531
.
. * Check that extra LSIs are returned
. di r(lrklsi_lb)
.57712034
. di r(lrklsi_ub)
47.72445
.
. choi_lr_testi 6 2 3 5, corn
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .6433784 38.07362 (Cornfield)
Attr. frac. ex. | .8 | -.5542952 .9737351 (Cornfield)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
. choi_lr_testi 6 2 3 5, woolf k(16)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .5841514 42.79712 (Woolf)
Attr. frac. ex. | .8 | -.7118849 .9766339 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 2.29 Pr>chi2 = 0.1306
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
1/16 LSI for the OR = [.42305043, 73.372846]
. choi_lr_testi 6 2 3 5, exact k(8)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 6 2 | 8 0.7500
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 9 7 | 16 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | .4086586 75.97119 (exact)
Attr. frac. ex. | .8 | -1.44703 .9868371 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
1-sided Fisher's exact P = 0.1573
2-sided Fisher's exact P = 0.3147
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.4742438
Likelihood ratio under the null hypothesis = 2.9944571
Likelihood ratio chi-squared statistic = 2.1935259
P value for the null hypothesis that the OR equals 1 = .13859174
1/6.825936 LSI for the OR = [.62397643, 42.991698]
1/8 LSI for the OR = [.57712034, 47.72445]
.
. * Check larger sample sizes
. choi_lr_testi 72 24 36 60, k(8)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 72 24 | 96 0.7500
Controls | 36 60 | 96 0.3750
-----------------+------------------------+------------------------
Total | 108 84 | 192 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 2.576879 9.768128 (exact)
Attr. frac. ex. | .8 | .6119337 .8976262 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 27.43 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9539211
Likelihood ratio under the null hypothesis = 1213489.1
Likelihood ratio chi-squared statistic = 28.018021
P value for the null hypothesis that the OR equals 1 = 1.202e-07
1/6.825936 LSI for the OR = [2.7020791, 9.3208557]
1/8 LSI for the OR = [2.6376818, 9.5689775]
. choi_lr_testi 180 60 90 150, k(16)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 180 60 | 240 0.7500
Controls | 90 150 | 240 0.3750
-----------------+------------------------+------------------------
Total | 270 210 | 480 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 3.317747 7.54922 (exact)
Attr. frac. ex. | .8 | .6985906 .867536 (exact)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 68.57 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9815164
Likelihood ratio under the null hypothesis = 1.821e+15
Likelihood ratio chi-squared statistic = 70.276267
P value for the null hypothesis that the OR equals 1 = 5.155e-17
1/6.825936 LSI for the OR = [3.384192, 7.409481]
1/16 LSI for the OR = [3.1339526, 8.0382116]
. choi_lr_testi 600 200 300 500, woolf k(32)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 600 200 | 800 0.7500
Controls | 300 500 | 800 0.3750
-----------------+------------------------+------------------------
Total | 900 700 | 1600 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 4.033904 6.19747 (Woolf)
Attr. frac. ex. | .8 | .7521012 .8386438 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 228.57 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9944476
Likelihood ratio under the null hypothesis = 8.826e+50
Likelihood ratio chi-squared statistic = 234.6139
P value for the null hypothesis that the OR equals 1 = 5.877e-53
1/6.825936 LSI for the OR = [4.035843, 6.2001042]
1/32 LSI for the OR = [3.7536033, 6.6831072]
. choi_lr_testi 1200 400 600 1000, woolf k(32)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 1200 400 | 1600 0.7500
Controls | 600 1000 | 1600 0.3750
-----------------+------------------------+------------------------
Total | 1800 1400 | 3200 0.5625
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 5 | 4.295722 5.819743 (Woolf)
Attr. frac. ex. | .8 | .7672103 .8281711 (Woolf)
Attr. frac. pop | .6 |
+-------------------------------------------------
chi2(1) = 457.14 Pr>chi2 = 0.0000
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 4.9972227
Likelihood ratio under the null hypothesis = 8.41e+101
Likelihood ratio chi-squared statistic = 469.38195
P value for the null hypothesis that the OR equals 1 = 4.37e-104
1/6.825936 LSI for the OR = [4.2967386, 5.8210048]
1/32 LSI for the OR = [4.0810354, 6.1363946]
.
. * Check behavior with empty cells
. choi_lr_testi 0 2 3 5
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 0 2 | 2 0.0000
Controls | 3 5 | 8 0.3750
-----------------+------------------------+------------------------
Total | 3 7 | 10 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | 0 | 0 4.841502 (Cornfield)
Prev. frac. ex. | 1 | -3.841502 1 (Cornfield)
Prev. frac. pop | . |
+-------------------------------------------------
chi2(1) = 1.07 Pr>chi2 = 0.3006
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = 0
Likelihood ratio under the null hypothesis = 2.1428571
Likelihood ratio chi-squared statistic = 1.5242801
P value for the null hypothesis that the OR equals 1 = .21697293
1/6.825936 LSI for the OR = [0, 3.7822765]
. choi_lr_testi 3 5 0 2, k(8)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 3 5 | 8 0.3750
Controls | 0 2 | 2 0.0000
-----------------+------------------------+------------------------
Total | 3 7 | 10 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | . | .2065475 . (Cornfield)
Attr. frac. ex. | 1 | -3.841502 . (Cornfield)
Attr. frac. pop | .375 |
+-------------------------------------------------
chi2(1) = 1.07 Pr>chi2 = 0.3006
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = infinity
Likelihood ratio under the null hypothesis = 2.1428571
Likelihood ratio chi-squared statistic = 1.5242801
P value for the null hypothesis that the OR equals 1 = .21697293
1/6.825936 LSI for the OR = [.26439104, infinity]
1/8 LSI for the OR = [.23114771, infinity]
. choi_lr_testi 300 500 0 200, k(16)
Proportion
| Exposed Unexposed | Total Exposed
-----------------+------------------------+------------------------
Cases | 300 500 | 800 0.3750
Controls | 0 200 | 200 0.0000
-----------------+------------------------+------------------------
Total | 300 700 | 1000 0.3000
| |
| Point estimate | [95% Conf. Interval]
|------------------------+------------------------
Odds ratio | . | 31.2292 . (Cornfield)
Attr. frac. ex. | 1 | .9679787 . (Cornfield)
Attr. frac. pop | .375 |
+-------------------------------------------------
chi2(1) = 107.14 Pr>chi2 = 0.0000
Note: Exact confidence levels not possible with zero count cells.
Likelihood ratio statistics conditioned on the marginal exposure rate
Maximum likelihood estimate (MLE) of odds ratio = infinity
Likelihood ratio under the null hypothesis = 2.629e+35
Likelihood ratio chi-squared statistic = 163.11415
P value for the null hypothesis that the OR equals 1 = 2.362e-37
1/6.825936 LSI for the OR = [61.74006, infinity]
1/16 LSI for the OR = [42.585057, infinity]
.
. log close
name: <unnamed>
log: C:\Users\dupontwd.VANDERBILT\Documents\Dropbox\OFFICE_desktop\ChoiFisherTest\Stata Program\ado\ForSubmission\c
> hoi_lr_test\test.log
log type: text
closed on: 31 May 2016, 14:09:09
---------------------------------------------------------------------------------------------------------------------------