Hi everyone,

May be a silly question, but I am having trouble finding clear information on how to interpret findings from teffects/tebalance output.

I am using teffects ipwra to look at whether completing a treatment had an impact on the likelihood of an outcome.

The data are highly confidential, so I cannot include data here. I'm also not too sure how to mock up another version using other data? However, I have attached a screenshot without variable names. I'm concerned because some of the variance levels and standardised differences appear quite large across some variables (i.e. weighted differences of .19 and weighted variance ratios of .64). From the output attached, are the data suitably balanced?

I have tried to change the format of the variables of those with the largest differences by changing between continuous to categorical, etc. but I have not had much luck!

Any advice would be greatly appreciated

Marlee


. tebalance overid, nolog

Overidentification test for covariate balance
H0: Covariates are balanced:

chi2(21) = 20.7296
Prob > chi2 = 0.4756