ABSTRACT
Modem computerized data analysis builds on advances in com puter technology and statistical methodology. Computing environments specifically designed for data analysis have evolved since the 60’s on computing platforms that reflect the state of the art in computing tech nology: first mainframes, then personal computers, and now worksta tions. Such environments incorporate aspects of data analysis that in clude model identification, diagnostic checking, transformation and the usual inferential procedures (Thisted, 1986; Huber, 1987). Becker et al. (1987) specify requirements for computerized dynamic graphics meth ods that include identification, deletion, linking, brushing, scaling, and rotation. These techniques are briefly defined below and some of them are then illustrated in Figures 11. la-11.1c with data from Case Study 2.