[David Radwin]

Hi Ayesha. I don't know of any research that has addressed this specific situation, but you might start with the "match first, then regress on matched sample" approach for complex weighted data described above in #2.