My actual dataset includes my quantities of interest (A, B, C, and D) obtained from 1000 bootstrap iterations by estimation type (unweighted vs. weighted) as well as profile (W and B). The below is a fraction of my actual dataset by showing the first 10 iterations by estimation and profile variables.
I would like to obtain and store the 95% confidence interval overlap by estimation and profile variables separately for my quantities of interest (i.e., A, B, C, D). The below centile command returns me the 95% confidence intervals for A, B, C, and D but I wonder if there is a way to obtain and store the 95% confidence interval overlaps as well.
Thanks,
NM
bysort estimation profile_new: centile TLE HLE ULE PLE, centile(2.5 97.5)
Code:
* Example generated by -dataex-. For more info, type help dataex clear input str26 estimation float iteration str10 profile_new float(A B C D) "Unweighted" 1 "0.W" 31.22547 23.49367 7.731804 .7523879 "Unweighted" 2 "0.W" 31.419674 23.715366 7.704307 .7547935 "Unweighted" 3 "0.W" 31.432106 23.62115 7.810956 .7514975 "Unweighted" 4 "0.W" 31.53055 23.81336 7.717191 .7552472 "Unweighted" 5 "0.W" 30.975897 23.23714 7.738756 .7501684 "Unweighted" 6 "0.W" 30.924076 23.21421 7.709865 .750684 "Unweighted" 7 "0.W" 31.097807 23.40726 7.690547 .7526981 "Unweighted" 8 "0.W" 31.188095 23.28139 7.906707 .7464831 "Unweighted" 9 "0.W" 31.04022 23.28384 7.756382 .7501184 "Unweighted" 10 "0.W" 31.259636 23.34705 7.912586 .7468753 "Unweighted" 1 "1.B" 28.60185 18.081507 10.520342 .6321797 "Unweighted" 2 "1.B" 28.84507 18.180395 10.664673 .6302774 "Unweighted" 3 "1.B" 28.5759 17.774813 10.801084 .6220211 "Unweighted" 4 "1.B" 29.137196 17.764761 11.372436 .6096936 "Unweighted" 5 "1.B" 28.944084 17.77063 11.173453 .6139641 "Unweighted" 6 "1.B" 28.396053 17.762722 10.63333 .6255349 "Unweighted" 7 "1.B" 28.90983 17.92432 10.985507 .6200079 "Unweighted" 8 "1.B" 27.772045 17.330206 10.44184 .6240162 "Unweighted" 9 "1.B" 28.76616 18.037151 10.729006 .6270268 "Unweighted" 10 "1.B" 28.88433 18.255674 10.628655 .6320269 "Weighted" 1 "0.W" 31.242056 23.867405 7.374651 .7639512 "Weighted" 2 "0.W" 31.61572 23.85055 7.765174 .7543889 "Weighted" 3 "0.W" 31.22626 23.84484 7.381419 .763615 "Weighted" 4 "0.W" 31.38188 23.941223 7.440657 .7628996 "Weighted" 5 "0.W" 31.76251 24.47861 7.283905 .770676 "Weighted" 6 "0.W" 31.60152 23.86003 7.741492 .7550279 "Weighted" 7 "0.W" 31.68685 24.23607 7.450778 .7648621 "Weighted" 8 "0.W" 31.084545 23.783575 7.300969 .7651254 "Weighted" 9 "0.W" 31.06961 23.824186 7.245427 .7668002 "Weighted" 10 "0.W" 31.539604 24.01042 7.529183 .7612785 "Weighted" 1 "1.B" 28.2727 18.17689 10.09581 .6429131 "Weighted" 2 "1.B" 28.70915 18.328226 10.380924 .6384106 "Weighted" 3 "1.B" 27.879877 18.14567 9.734207 .6508518 "Weighted" 4 "1.B" 29.038834 18.873323 10.16551 .6499339 "Weighted" 5 "1.B" 28.3958 18.071054 10.324748 .6363988 "Weighted" 6 "1.B" 29.26605 19.116856 10.149196 .6532092 "Weighted" 7 "1.B" 28.50781 18.73163 9.776183 .65707 "Weighted" 8 "1.B" 28.23492 18.348585 9.886334 .6498543 "Weighted" 9 "1.B" 28.42887 18.459541 9.96933 .6493238 "Weighted" 10 "1.B" 28.82799 18.732927 10.09506 .6498173 end
0 Response to Obtaining the 95% confidence interval overlap
Post a Comment