search

worldbank / Stata-IE-Visual-Library

This is a repository maintained by DIME Analytics and containing example graphs on how to explore data sets and display results of Impact Evaluations using Stata. For information on how to contribute to the library and download codes and data sets, click on the link to GitHub below.
MIT License
48 stars 23 forks source link

Density plot with averages by gender #41

Closed MRuzzante closed 5 years ago

MRuzzante commented 5 years ago

Kernel density plots of an outcome variable over treatment group and gender with average line and text box.

The color palette used is inspired by the Telegraph, which, in turn, was "inspired by the 'Votes for Women' campaign in the UK as part of the initial suffrage movement in the early 20th century".

See https://blog.datawrapper.de/gendercolor/#fn1 for a discussion on colors and gender data and https://www.telegraph.co.uk/women/business/women-mean-business-interactive/ for the Telegraph case (h/t @luizaandrade).

luizaandrade commented 5 years ago

Thanks, Matteo! I'll put it up next week when I'm back to the office.

MRuzzante commented 5 years ago

No worries, Luiza! I will also add more graphs in the next days.

I did not include any tag, but I guess you can use the ones, which already exist, such as stata, graph, plots, impact-evaluations, and perhaps also density, average, gender, text-box.

luizaandrade commented 5 years ago

Thanks! The description is actually great, and the search should work on it.

MRuzzante commented 5 years ago

I actually ended up committing two new graphs in the same pull request:

  1. Quantile treatment effect estimates with fixed effects and clustered standard errors using qreg2 and coefplot
  2. Quantile treatment effect estimates by gender with fixed effects and clustered standard errors using qreg2 and coefplot

You could use as tags: stata, graph, plots, impact-evaluations, quantile-regression, coefplot, and gender (only for the second one).

Perhaps, it also makes sense to make a folder only for quantile regressions and impact distributional analysis, instead of keeping it under "Regression coefficients". Your choice! ;)

For useful resources/links, you could attach this paper by Lupe to the graph comment: http://documents.worldbank.org/curated/en/292901499351272899/Distributional-impact-analysis-toolkit-and-illustrations-of-impacts-beyond-the-average-treatment-effect