Reliability and practice effects on the Luria-Nebraska neuropsychological battery and the Revised Wechsler memory scale

This study was designed to examine the reliability of equivalent and parallel forms of two representative neuropsychological tests: the Revised Wechsler Memory Scale (RWMS) created by Elbert W. Russell (1975) and the Luria-Nebraska Neuropsychological Battery (LN) developed by Charles J. Golden and his colleagues (Golden, Hammeke, & Purisch, 1978; Golden, Purisch, & Hammeke, 1978, 1979; Lewis, Golden, Moses, Osmon, Purisch, & Hammeke, 1979; Osmon, Golden, Purisch, Hammeke, & Blume, 1979; Purisch, Golden, & Hammeke, 1978). Although studies have been conducted which explore the reliability of Golden's scoring procedure, the split-half reliability of the original form of the battery, and the test-retest reliabilities of the 14 clinical scales, no research has been conducted to cross-validate these results or to examine the reliability of using parallel forms. Information concerning the internal consistency of the Luria-Nebraska, as well as test reliability, standard error(s) of measurement, and practice effects are vitally important for clinicians interested in employing neuropsychological instruments on a repeated basis to evaluate a patient's process of recovery from brain damage. The examination of the reliability of parallel test forms is particularly valuable since this reliability procedure is "the most rigorous" test of the accuracy of an assessment device (Thorndike & Hagen, 1961, p. 178).