Learning more by crossing levels: evidence from airplanes, hospitals, and orchestras

Scholars generally conduct research at a single level of analysis (such as the individual, the group, or the organization level), although they often turn to the next-lower level for explanatory mechanisms. I suggest that robust understanding of social and organizational dynamics requires attention to higher as well as lower levels of analysis. The benefits of research and theory that 'brackets' one's focal phenomenon by attending to constructs at both higher and lower levels of analyses are illustrated with findings from research on aircraft cockpit crews, hospital patient care teams, and professional musical ensembles. Copyright (C) 2003 John Wiley Sons, Ltd.