Nonresponse bias in public leadership research: an empirical assessment