ABSTRACT
Item generation has long been appealing to test developers. Around 1970, Wells Plively and colleagues (Plively, Patterson, & Page, 1968) proposed a system based on item forms to generate an infinitely large domain of mathematics problems. Each item form contained fixed text for a mathematics word problem, along with variable elements and replacement rules for these elements. For example, the specific numbers in the word problem were variable elements. Similarly, facet theory (see Shye, Elizur, & Hoffman, 1994), developed from Guttman (1969), postulates that rules can be specified to map specific features into an item. The other chapters in this book support the great potential of item generation in several contexts and need not be reviewed here.
