Statistics and Experimental Design

gestions that humane patient care may be either uninfluenced or perhaps even enhanced by scientific medicine and the accompanying new morality. However, a warning by Babbie should be emphasized: It may be that "the indicators of humane patient care used in these analyses are not adequate to the research problem." Inferences about faculty performance based on faculty views reported by and often about themselves must remain tentative at best. At worst, they can be very misleading.