ietoolkit
ietoolkit copied to clipboard
worldbank
[iebaltab]: headers for groups of variables
Iebaltab is a great command. Thank you to the developers!
Is it possible to group variables in categories that get their own header (e.g. household characteristics), to create a balance table that is more structured - in my case for LaTex? For instance, through manually using the varlabels option? Otherwise, this would be a great functionality to add in the future.
Thank you for your feedback! We are happy that you find the command useful!
We are in the middle of a complete re-write of iebaltab. The number of features that were added since this command was first written is large, and it had turned into a patchwork where adding new features became a pain.
The feature you are suggesting well be included in the new version. We are trying to base new syntax as much as possible on solutions on how commonly used command already solves this. I guess you are aware of estout. They solve this using the options prehead() , posthead() , prefoot() and postfoot() to solve this. Is that a solution that would work for you?
Also, in the new version we will provide a matrix in ereturn list where each row is a balance variable, and each column is an estimate or a number related to this variable. Then users can implement their own code turning that to LaTeX. And hopefully they will share that and with permission we are happy to then implement great custom solutions into future versions of the command.
That sounds great, I am looking forward to the new version then!
Yes I am aware of estout. Regarding estout/esttab, I have used the refcat functionality to include headers in tables. I don't think I have ever used the other options you mentioned, I guess these are a bit more advanced. I was currently mostly looking into iebaltab due to its user-friendly functionality to include fixed effects and covariates, but if the header functionality is not there yet I might look into estout for now, or just make the iebaltab table without the headers. Thanks!
estout advantage is that it has a very wide usage, but you have to calculate the estimates to output yourself. So you would have to include the fixed effects and covariates yourself in your own balance regression if you were to use estout. But since estout is the most popular Stata package for outputting already calculated estimates, we are emulating their output best-practices in our command that both calculates and outputs estimates.
Before I wrote the first version of this command I ran my own balance regression with different specifications and then used estout to output them. iebaltab replaces that with one step, but obviously only work for balance tables.