given the increasing reliance on teams and alternative employment contracts. However, the committee is unaware of any systematic establishment-based occupational surveys that gather detailed information about work organization.
If work is changing (due to different uses of technology, a greater diversity of employment contracts, demographic changes of workers, etc.), then adequately representing change in occupational information systems requires an adequate sampling plan for measuring change. At a minimum, this would require repeated measurements of a representative sample of the U.S. workforce. A cross-sectional plan repeated periodically would be adequate to identify changing trends and patterns of work. A longitudinal sampling plan including the same respondents over time would be needed to track the changing career paths of workers.
Since the nature of the sample determines the type of data collected and its usefulness for various applications, and an adequate sampling plan is critical to the next stage of O*NET™ development, in the committee's judgment, high priority should be given to evaluating the appropriateness of alternative sampling strategies.
Issues of the usefulness of an occupational information system like O*NET™ include the quality level of the data, enabling potential for coping with change, cross-walks among related systems, system maintenance and control, and further technological developments.
Initial validity and reliability studies were undertaken during the first three years of O*NET™ prototype development (Peterson et al., 1996,1999). Such research is important to establish the measurement integrity of the instruments and the reliability of the data gathering procedures. In the committee's view, these studies should continue and should be expanded to include laboratory and field studies comparing various conditions of data collection, such as differing demographic makeup of job experts,