ABSTRACT
Many research questions and good research programs start with a conflict of ideas, views, and approaches. In the 1970s I, as a student, had the following conversation with my mentor LJC (Lee J. Cronbach):
LFH: I am thinking that within a computer-based training environment it will be possible to base items strictly on their content and cognitive structure so that at least some day they may be computer generated.
LJC: Item writing is an art!
LFH: That shouldn’t have to be true, because we should be able to incorporate all of our sound psychological knowledge into designing an item. Items should be based on theory and constructs derived from this theory. Finally any tested behavior will have to be traced back to exactly these constructs and theories!
