California Verbal Learning Test-Second Edition

Archives of Clinical Neuropsychology 17 (2002) 509 – 512 Book review California Verbal Learning Test-Second Edition. Edited by D. Delis, J. Kramer, E. Kaplan, B. Ober. San Antonio, TX. The Psychological Corporation, 2000 The CVLT-II is a revision and restandardization of one of the most well-known verbal learning tests in neuropsychology. Although a popular test, the original CVLT was normed on a ‘‘reference sample’’ of 273 nonclinical subjects. This sample has often been criticized as being biased toward individuals of higher education and functioning, as well as reflecting a narrow range of memory performance (e.g., Elwood, 1995; Randolph, Gold, Kozora, Cullum, & Hermana, 1994; Weins, Tindall, & Crossen, 1994). Concerns have also been raised with regard to the ethnic minority makeup of that sample (Norman, Evans, Miller, & Heaton, 2000). A common conclusion has been that although the CVLT provides valuable qualitative information, its use in the quantification of memory functioning has been limited by inadequate normative data. These criticisms were clearly a driving force in the development of the Second Edition of the CVLT (CVLT-II). Although the CVLT-II is most noteworthy for its large-scale restandardization, a number of modifications have been made to the test itself, including a new word list, optional forced choice discrimination task, alternate and short forms, new indices, and revisions to old indices. The most obvious change in the test itself is the use of a new word list that is intended to be easier, with less geographic, cultural, and socioeconomic bias. The grocery shopping theme has been dropped in favor of an empirically derived word list comprised of words from four unrelated semantic categories. Another prominent change is the addition of an optional 16-item forced choice discrimination task that follows the traditional yes–no recognition task after a brief delay. The task is intended to assess effort, and the manual details a number of within-test comparisons that can aid in doing so. An alternate test form and a nine-word short form have also been introduced to improve the utility of the test in serial assessments and aid in the assessment of examinees with severe cognitive dysfunction. Less obvious modifications to the test include the introduction of several new process variables (e.g., a measure of idiosyncratic organizational strategies, two new slope indices, and variables characterizing intrusion errors), as well as modifications to existing variables (e.g., new method for calculating semantic clustering, recognition discriminability expressed as a d0 measure, and savings scores calculated from z scores instead of raw scores). All of these modifications are supported by well-presented theoretical and empirical rationales in the test manual, which is much more comprehensive than the original and contains ample technical data. 0887-6177/02/$ – see front matter D 2002 National Academy of Neuropsychology. PII: S 0 8 8 7 - 6 1 7 7 ( 0 1 ) 0 0 1 2 5 - 1 510 Book review / Archives of Clinical Neuropsychology 17 (2002) 509–512 Perhaps the most anticipated feature of the CVLT-II is its restandardization. The CVLT-II underwent nationwide standardization with a final normative reference sample consisting of 1087 individuals, ages 16–89, from 47 states. Stratification variables were based on US Census figures, and include gender, race/ethnicity (White, African American, Hispanic, and other), and education level. From the data in the manual comparing sample stratification to Census statistics for each age group, it is clear that the stratification efforts were highly successful. It is important to note that although education level is included as a stratification variable, normative data is not provided by education level. The normative data for the alternate and short forms are more difficult to evaluate. The data were derived by administering each to a small sample of individuals (288 for the alternate form, 278 for the short form) counterbalanced with the standard form, then using statistical methodology to calibrate scores on each. Demographic characteristics of the smaller calibration samples are not fully presented in the test manual, raising questions about the equivalence of the alternate and short form norms, particularly at extremes of age and education. Reliability data for the CVLT-II are good overall, and are adequately reported in the manual. Multiple methods were used (i.e., three internal consistency estimates, test–retest stability). Reliability was also examined in subgroups representing extremes of performance, and a small mixed neuropsychiatric sample (n = 124). Internal consistency estimates range from 0.78 to 0.89 in the normative sample, and from 0.80 to 0.96 in the mixed neuropsychiatric sample. Test–retest stability in a nonclinical sample of 78 individuals is adequate, and is reported for all major CVLT-II variables. Reliability of the CVLT-II Alternate Form is examined in terms of correlations with the Standard Form. Although this may speak more to the issue of concurrent validity rather than reliability, there is no reason to suspect that the Alternate Form would differ substantially in reliability. Reliability data for the Short Form, however, are conspicuously absent. Validity data for the CVLT-II consists primarily of analyses with nonclinical groups. The authors acknowledge that their strategy is to demonstrate the close association between the CVLT-II and the original edition, and thus build on vast existing clinical validity data on the CVLT. They do so by presenting data from a small nonclinical group (n = 62) that was administered both editions, demonstrating comparable mean scores and standard deviations, as well as adequate correlations between the tests. The authors conclude that it is valid to compare raw scores between the formats, although this practice may be questionable in groups most likely to benefit from the new word list (e.g., low education, low SES, rural populations). The factor structure of the CVLT-II is examined in both the normative sample and a mixed neuropsychiatric sample, with results that are comparable between groups, and comparable to the factor structure presented in the original CVLT manual. Data comparing the CVLT-II to the Wechsler Abbreviated Scale of Intelligence (WASI) are comparable to similar data reported for the original CVLT. However, given the sample size (n = 72) and correlational nature of the data, they are of negligible clinical utility. Overall, adequate validity is demonstrated for the CVLT-II. However, minimal validity data is provided for the new optional forced choice discrimination task. The task is intended to assess examinee effort, and is clearly well designed. It includes both concrete words as distractors, as well as Book review / Archives of Clinical Neuropsychology 17 (2002) 509–512 511 abstract words (presumably less likely to be mistakenly chosen (...truncated)