Hi,

I'm using -coefplot- to create a graph showing the balance between covariates for a treatment and control group. I separately regress each covariate on the treatment variable, and store the estimate to plot. Important to my solution is that I can control the names along the y-axis like in my example below.

Note: I standardize the data so coefficients are comparable across regressions.

Here is a simple example:

Code:
// In this example, I check whether random covariates are balanced across foreign vs. domestic cars

sysuse auto, clear

eststo clear

local variables price mpg headroom trunk

foreach var of local variables {
    egen `var'_std = std(`var')
    eststo: reg `var'_std foreign
}

coefplot ///
    (est1, rename(foreign = "outcome1"))    ///
    (est2, rename(foreign = "outcome2"))    ///
    (est3, rename(foreign = "outcome3"))    ///
    (est4, rename(foreign = "outcome4")),   ///
    drop(_cons) xline(0) legend(off)             ///
    coeflabel(outcome1 = "Some name" outcome2 = "A different name" outcome3 = "Third name" outcome4 = "Final name")
As you can see from the output pasted below, the y-axis labels and ticks do not quite line up with the estimates and confidence intervals. Why is that?

I thought this would be something easy to solve by searching Statalist but I was unsuccessful in finding a solution. Any assistance would be appreciated. Thanks!