does sensitivity analysis show that imputed statistics closely align with survey results?
notes
https://github.com/impunation/sensitivity
create your main synthetic data file. then create it again with every data set except for one and show that it matches the data set that’s been excluded.