participation, whereas those workers who have high control-high demand jobs have the highest. This result suggests that job characteristics have an important impact on the extent to which we are capable or willing to enjoy life and to contribute to society, thereby illustrating the pervasive repercussions of healthy--or better, unhealthy--work on society as a whole (cf. Reed, 1996).
The findings just presented take on an even more impressive status when we consider the following methodological points. First, the studies that support the demand-control model are large-scale efforts, with sample sizes usually over 1,000 and in one case comprising one fourth of the working population of Sweden (a sample size of 958,096). These large samples lend further credence to the reliability and generalizability of the findings obtained. Second, much of this research made an explicit attempt to control for other conventional risk factors, such as age, smoking patterns, weight, education, and cholesterol level. These other factors played a smaller role in predicting health than did the dimensions specified by the demand-control model ( Karasek & Theorell, 1990). Thus, psychological demands, and especially decision latitude, play a unique and significant role in determining worker health.
This body of research is of great importance for computer-based work, the focus of our book, because research has shown that information technology acts as an amplifier of the effects represented by the demand-control model ( Karasek & Theorell, 1990). That is, computerization combined with high psychological demands and low decision latitude leads to even more negative effects on worker health. Therefore, the body of research reviewed by Karasek and Theorell is of great value because it adds another constraint on a framework for analyzing human work. From the viewpoint of health, a work analysis methodology must strive to improve decision latitude by providing workers with the autonomy to make decisions and the opportunity to exercise and develop skill. The quantitative data that we presented earlier in this subsection indicate that enormous economic gains may be realized by adopting such an approach. Even more important, however, explicitly designing for healthy work will lead to benefits that cannot be quantified, such as humane jobs that allow people to participate fully in life and society.
What implications can we draw from this section? If a framework for work analysis is to lead to effective computer-based work, it must strive to: (a) support workers in adapting to, and coping with, events that are unfamiliar to them and that have not been anticipated by system designers, (b) identify the functionality that is required to accomplish intellectual tasks requiring discretionary decision making (i.e., usefulness issues), (c) be based on an understanding of human capabilities and limitations (i.e., usability issues), and (d) improve decision latitude by providing workers with the autonomy to make decisions and the opportunity to exercise and develop skill. The evidence reviewed previously suggests that a work analysis framework that satisfies these constraints will lead to the design of computer-based work that is safe, productive, and healthy. In the remainder of this book, we argue that this three-way objective can be achieved by deliberately and systematically designing information systems that support worker adaptation.