Creating Concentration Curves for Binary Outcome using Survey Data

[Charles Ackah]

Dear all,
I am trying to create concentration curves using 3 waves of survey data collected in 2010, 2012, and 2014. The outcome variable (morbid) is a binary variable taking the value 1 when a child has ill health and 0 otherwise. The socioeconomic variable is log of per capita household income (pcy). I want to create the concentration curves for all the 3 waves and combine them in one graph. I have read the related posts but can't seem to succeed. Any help will be greatly appreciated. My example data is posted below:

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(wave morbid pcy)
3 0  3.819281
3 0 3.9515884
3 0 3.9515884
3 0  4.006615
3 0  4.177999
3 0 4.3120995
3 0   4.34096
3 0 4.3855033
3 0 4.3855033
3 0 4.3855033
3 0 4.3855033
3 0 4.3855033
3 0 4.3931103
3 0 4.3931103
3 0 4.4143877
3 0 4.4143877
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0 4.4430676
3 0  4.444446
3 0  4.444446
3 0  4.444446
3 0  4.444446
3 0 4.4544964
3 0 4.4544964
3 0 4.4544964
3 0 4.4544964
3 0 4.4544964
3 0 4.4544964
3 1 4.4544964
3 0 4.4544964
3 0  4.514471
3 0  4.520625
3 0  4.520625
3 0  4.520625
3 0 4.5207686
3 0 4.5207686
3 0 4.5207686
3 0 4.5722795
3 0 4.5975876
3 0 4.5975876
3 0 4.6061473
3 0  4.650611
3 0  4.650611
3 0  4.650611
3 0  4.650611
3 0  4.651075
3 0  4.651075
3 0  4.651075
3 0  4.665702
3 0  4.667589
3 0  4.667589
3 0  4.678933
3 0  4.678933
3 0  4.678933
3 0 4.6829553
3 0 4.6829553
3 0 4.7182803
3 0 4.7182803
3 0  4.726477
3 0  4.733525
3 0  4.733525
3 0  4.733525
3 0 4.7439804
3 0 4.7485523
3 0 4.7638173
3 0 4.7638173
3 0 4.7638173
3 0 4.7638173
3 0 4.7638173
3 0  4.768857
3 0  4.768858
3 0  4.778506
3 0  4.781909
3 0  4.781909
3 0   4.78492
3 0  4.790968
3 0  4.790971
3 0  4.790971
3 0  4.799129
3 0  4.799129
3 0  4.799129
3 0  4.799177
3 0  4.799177
3 0  4.809937
3 0  4.809937
3 0  4.809937
3 0  4.809937
3 0  4.836478
3 0  4.836478
3 0  4.836478
3 0  4.863235
3 0  4.869103
3 0  4.881122
3 0  4.881122
3 0  4.890372
3 0  4.890748
3 0  4.891485
3 0  4.891485
3 0  4.903506
3 0  4.903506
3 0  4.903506
3 0  4.903506
3 0   4.91504
3 0   4.91504
3 0   4.91504
3 0  4.919936
3 0    4.9245
3 0    4.9245
3 0  4.933956
3 0  4.938131
3 0  4.938131
3 0  4.938131
3 0  4.946213
3 0  4.949223
3 0  4.955271
3 0  4.955271
3 0  4.955271
3 0  4.964413
3 0  4.966456
3 0  4.966456
3 0  4.967242
3 0  4.969528
3 0  4.969528
3 0   4.97261
3 0  4.982432
3 0  4.982432
3 0  4.982432
3 0  4.990629
3 0  4.999382
3 0  4.999867
3 0  4.999867
3 0  5.003062
3 0  5.003062
3 0  5.003062
3 0  5.003062
3 0  5.003062
3 0  5.003062
3 0   5.01398
3 0  5.015727
3 0  5.015727
3 0  5.015727
3 0  5.015727
3 0  5.020254
3 0  5.022638
3 0  5.022638
3 0  5.022638
3 0  5.025297
3 0  5.025297
3 0  5.025297
3 0  5.028369
3 0  5.049753
3 0  5.049753
3 0  5.049753
3 0  5.051774
3 0  5.051774
3 1  5.051774
3 0  5.054117
3 0  5.070161
3 0  5.070161
3 0  5.070161
3 0  5.071578
3 0  5.071578
3 0  5.071578
3 0  5.080895
3 0  5.083105
3 0  5.083105
3 0  5.083105
3 0   5.09429
3 0   5.09429
3 0   5.09429
3 0  5.100102
3 0  5.100102
3 0  5.100443
3 0  5.100443
3 0  5.100443
3 0  5.100443
3 0  5.100443
3 0  5.105492
3 0  5.105492
3 0  5.109244
3 0  5.109244
3 0  5.112408
3 0  5.113124
3 0  5.113124
3 0  5.113688
3 0  5.122177
3 0  5.122177
3 0  5.122177
3 0  5.122177
3 0  5.122177
3 0  5.122177
3 0  5.135325
end

[Leonardo Guizzetti]

I’m not familiar with the term concentration curve. Can you describe what that is?

[Charles Ackah]

When I run the following code, I get the following output (with the confidence interval) for the concentration curve. How do I suppress the confidence interval in the concentration curve?

conindex morbid, rankvar(pcy) truezero svy graph

[Leonardo Guizzetti]

-conindex- is a user-contributed package. You are asked to describe where it is from.

My suggestion is to read the help file with -help conindex- to see if there is an option to do what you ask. If not, perhaps you can answer my earlier question so others might more productively help you.