Heuristics for generating and validating surmise relations across, between and within sets/tests

Abstract

Generating and validating surmise relations across, between and within sets has been strongly stimulated for practical reasons. For instance, the number of psychological tests (sets of items) used simultaneously in diagnostic settings continues to grow. Therefore it is useful to develop new methods that economize assessment procedures and that allow for restructuring of tests. For this purpose, we replace the term ’sets’ and apply in this chapter the Surmise Relations across, between and within Tests (SRbT) approach. Applying the SRbT approach is however not restricted to tests (sets of items); it can be used e.g. in Educational Science for developing and restructuring curricula in order to obtain among other things a meaningful sequence of courses (Hockemeyer, Albert, & Brandt, 1998). Another example in Computer Science is (re-)structuring hypertext documents respecting prerequisites (Albert and Hockemeyer, 1997).