Generalizability theory (G-theory) allows us to estimate the magnitude of these sources of error simultaneously. Using G-theory, we could simulta examine the homogeneity of the items and the stability of the responses time, as well as the interaction of these two sources of error. The classical and common approach to G-theory "estimates sources of variation among scores within an analysis of variance framework' (Cardinet et al. 2010, p. viii). Using allows for the partitioning of variance into multiple sources, which can be crossed or nested. Although the topic of G-theory is incredibly important study of score reliability, providing a full treatment of the analytic technique s the scope of this book. Readers should consult Brennan (2001), Cardinet (2010), Meyer (2010), and Shavelson and Webb (1991), all of which provide horough introductions to the technique. We will now turn to the two major types of reliability evidence that are of most oncern for the affective instrument developer: internal consistency and stability