Rubric scoring of a clinical test of executive functioning

摘要

Executive functions (EF) are complex abilities that allow one to successfully complete independent, deliberate, and novel goal-directed activities (Lezak, Howieson, & Loring, 2004). EF tests require solving problems with minimal direction from the examiner (Baddeley, 1992; Shallice & Burgess, 1991). Because EF skills tend to show up globally, Lezak et al. (2004) suggested clinicians will learn more about one’s EF abilities by observing how he or she goes about solving a problem than from a test score. If this is the case, a “rubric” score that took into consideration “how” a problem was solved may provide the clinician with better information for treatment planning than a test score, as long as it did not greatly affect test sensitivity or specificity. Sensitivity and specificity are important factors in determining the usefulness of EF tests. Sensitivity refers to the probability of identifying abnormal functioning in an impaired individual or “hit rate” of a test, whereas specificity reflects the probability of correctly identifying healthy individual with the test (Cartoni & Lincoln, 2005; Kiel & Kaszniak, 2002).Rubrics are useful scoring tools that divide tasks into component elements and provide a description of levels of performance for each element (Goodrich, 2005; Hanna & Smith, 1998). Rubrics have been widely used to assess student performance (Andrade, 2000; Falchikov, 1986; Goodrich, 1997), but have not been used to score EF tests. The aim of this study was to examine sensitivity and specificity for a clinical test of EF, the Rapid Assessment of Problem Solving test (RAPS) when scored with a rubric that allowed the examiner to describe the quality of performance using a standard that was developed from a large body of normative research. This differs from using the three traditional test scores from the RAPS that require time intensive calculations.