Recent methods to improve generalizations from nonrandom samples typically invoke assumptions such as the strong ignorability of sample selection, which is challenging to meet in practice. Although researchers acknowledge the difficulty in meeting this assumption, point estimates are still provided and used without considering alternative…