Another notable effort to better synchronize tasks, scoring guides, and measurement models with explicit 1cit models of cognition is evidence-centered assessment design (Mislevy and 2006, Mislevy et al, 2002a). While the evidence-centered approach is more Ricon scente often associated with task analysis and the identifi cation of item facets and/or production rules (Mislevy 1996), we note in passing that this approach is compatible with a construct evy modeling formulation (Mislevy et al. 2002b). Indeed, one of the benefits of construct modeling as a ling as an assessment practice is its flexibility. is described above, th ed above, the construct modeling framework is comprehensive and allows In addition to the multidimensional for specification ecification of a of a variety of models given varied task formats and scoring assignments.