Modeling Item-Level Data With Item Response Theory

Psychology can be separated into two camps, a substantive camp that is primarily interested in understanding important aspects of human behavior and thought and a methodological camp that is primarily interested in developing tools that will be used by the sub-stantive researchers to answer dif-ficult questions. Ideally, the two camps should have much cross-fertilization, and the lines between them should be blurred. However, complex theories are often con-structed before appropriate meth-ods are developed for testing the hypotheses. Additionally, quanti-tative psychologists often develop techniques that might be difficult and impractical to apply initially to real-data problems. These novel techniques need to be refined through robustness studies and analytical work before they can be used with real-world data. A psy-chometric framework called item response theory (IRT) has under-gone such prudent testing and re-