Obtaining a forest plot showing the standardised mean differences after teffects ipw and aipw

[Gaby Kassler]

Dear

Statalisters,
I have been trying to produce a forest plot-style of graph to show the mean in the treated and untreated, before and after IPTW/AIPTW. I have

managed to get tables by using calculated stabilised weights with

pbalchk
(these

IPTW weights were obtained after a

logit regression and then applied to

pbalchk
) and I can

obtain a similar table after

teffects

ipw with

tebalance summarize.

I looked up all the documentation for in the

te manual and looked around for code, but I cannot seem to find any program that allows me to produce a forest plot like the one in attachment.
Unfortunately, the paper that features this plot did not publish the code too.

Thanks as always for your help.
- Gaby

[Gaby Kassler]

Apologies for the text - I seem to have trouble formatting it.

[Andrew Musau]

The figure in #1 is just a simple twoway scatter plot with 2 groups. As long as you have the points (represented by the markers), it should be straightforward to reproduce. I do not know much about what output you get from teffects ipw/ aiptw.

[Gaby Kassler]

Thanks Andrew for pointing me in the right direction - I saw the same outputs in R and thought that the plot in my post might be something akin to a coefplot.

[Adam Alhassan]

Gaby Kassler this might be too late, but I came across this post while trying to figure out how to configure my pre- and post-ipw weighting forest plot. I think the pbalchk command has a graph option that produces this exact same graph in your attachment, Gaby.

[Gaby Kassler]

Hi Adam Alhassan yes, that was a while ago and I used scatter I think. I had also installed that command but did not notice the graph option. Thanks for the tip!