Vol. 14, No. 3, July 1996 | BATYA FRIEDMAN, HELEN NISSENBAUM
The article by Batya Friedman and Helen Nissenbaum explores the concept of bias in computer systems, categorizing it into three main types: preexisting, technical, and emergent. Preexisting bias arises from social institutions, practices, and attitudes, while technical bias stems from technical constraints or considerations. Emergent bias occurs in the context of use, often due to changing societal knowledge, population, or cultural values. The authors provide a framework for understanding and addressing these biases, using case studies from various fields such as banking, commerce, education, and law. They argue that freedom from bias should be a key criterion for judging the quality of computer systems, similar to criteria like reliability, accuracy, and efficiency. The article concludes by discussing practical methods for minimizing bias in system design and the ethical responsibilities of designers in the face of biased client demands.The article by Batya Friedman and Helen Nissenbaum explores the concept of bias in computer systems, categorizing it into three main types: preexisting, technical, and emergent. Preexisting bias arises from social institutions, practices, and attitudes, while technical bias stems from technical constraints or considerations. Emergent bias occurs in the context of use, often due to changing societal knowledge, population, or cultural values. The authors provide a framework for understanding and addressing these biases, using case studies from various fields such as banking, commerce, education, and law. They argue that freedom from bias should be a key criterion for judging the quality of computer systems, similar to criteria like reliability, accuracy, and efficiency. The article concludes by discussing practical methods for minimizing bias in system design and the ethical responsibilities of designers in the face of biased client demands.