ABSTRACT
This is the first of three chapters on item response theory (IRT), perhaps the most significant practical innovation in large-scale testing in the second half of the 20th century. IRT’s item-level modeling enabled developments such as adaptive testing, matrix sampling of items, and constructing tests to targeted levels of accuracy. This chapter sketches the historical roots and fundamental models of IRT—tools and concepts that continue to influence practice and thinking. Differing views on the philosophical status of the IRT person variables θ are presented. The following chapters address, in turn, an explication of IRT from a sociocognitive perspective, and whether, and what sense, IRT models might constitute measures of psychological attributes.
