divide or the other outmoded images are major issues for future study and action.

Throughout this volume we have stressed the need for more intensive direct observation of what workers actually do in their jobs today. Changing the images of work and going beyond abstract arguments about trends in skills requires detailed and rich description and data reported from direct experiences of workers. Thus the sociological and anthropological traditions of observing and participating in real work settings and producing detailed narratives describing the actual experiences of workers need to be encouraged, with the objective of updating perspectives on work. But to be representative, these studies must examine the full array of occupations and workers found in the labor force today. Researchers are especially limited in their ability to describe what managers do at work today because it is difficult to measure. Furthermore, sociologists, industrial relations experts, anthropologists, and others continue to focus their efforts on more easily quantifiable jobs in lower-level occupational groups. It is also important to examine ways of integrating data describing what workers do from other disciplines, such as industrial and organizational psychology and human factors.

Direct observation and in-depth description of what workers do is a necessary but not sufficient input to update and continue to monitor changes in the aggregate structures of work and the content of jobs. To do this requires a national sample representative of the labor force. This type of data collection is needed both to complete the data collection and analysis needed to make O*NET™ operational and to realize its potential and to track systematically the changes in work and their consequences for organizations, individuals, and society.