Computerized decision models are increasingly being used in the public and private sector for decisions that affect everyone. Pension funds are invested on the basis of computer models; medical student interns are matched with hospitals on the basis of a computer model; air traffic is controlled on the basis of a computer model; income tax laws are changed on the basis of models; and so on. Yet, often those who make decisions on the basis of a computer model do not understand the algorithms used, or the assumptions made, in the model. They cannot critically assess the model's output. Moreover, decisions made using computer models are often not accessible to the public that is affected by them. Hence, increasing use of, and dependence upon, computer models raises important ethical questions. The most serious issues center on the effects of the use of models on the democratic process, the responsibilities of the model designers to their clients and to the public, and the treatment of risk that is often buried in a model. While a good deal of research has been done in recent years on the social impact of computers and on the ethical issues surrounding computers, only rarely have the ethical issues arising from the use of computer decision models been addressed. The investigator will collaborate in identifying and analyzing the ethical issues surrounding the use of computer models. The analysis will lead to recommendations for change in the way models are designed and used, and in the way modelers function in their professional roles. This project furthers VPW program objectives to provide opportunities for women to advance their careers in science or engineering through research, and to encourage other women to pursue careers in these areas through the investigator's enhanced visibility as a role model on the host campus. The proposed activities which contribute to the second objective include: presenting seminars; arranging informal discussion groups for women faculty and students; and teaching/co-teaching a course.
