Integrating analyses of data bases into statistical instruction

Students in graduate-level applied statistical courses should be trained to manipulate realistic data bases which are sufficiently large and complex that they provide verisimilitude with respect to thesis studies and other real-world applications. At the University of Maryland, we have been integrating data base manipulation into our intermediate-level statistics instruction for several years. This presentation concentrates on several issues related to the use of data bases in statistical instruction: appropriate course level; desirable characteristics of data bases; the role of mainframe and microcomputer statistical software; integration of data base manipulation skills into instruction on statistical topics; and grading practices.