Handbook of Executive Functioning by Sam Goldstein & Jack A. Naglieri

Author:Sam Goldstein & Jack A. Naglieri
Language: eng
Format: epub
Publisher: Springer New York, New York, NY

(b)Reliability of the TVCF scales

Tests can be either speed or power tests: with speeded tests the individuals’ scores depend on how quickly an individual can complete the task. With power tests, individuals are not timed as they complete the task. The TVCF contains both speed and power tests. The Classification subtest of the TVCF is a power test as the score depends on the number of right and wrong answers but the subtest is not timed. The Category Fluency and Letter Naming subtests include elements of both power and speed tests.

The Trails C subtest is a speed test. Coefficient alpha, perhaps the best measure of test score reliability, was computed for each TVCF subtest for selected subgroups in the TVCF normative sample. These groups include both genders and the largest ethnic groups in the USA as well as clinical groups of learning disabled and brain-injured subjects. See the TVCF Examiner’s Manual for the data (Reynolds & Horton, 2006).

The majority of the TVCF subtest reliability scores reach at least .70 or higher for all of the selected subgroups. For each gender and for the four major ethnic groups included in the sampling plan (European-Americans, African-Americans, Hispanic-Americans, and Asian-Americans) the majority of the coefficient alphas were excellent with almost all subtests but Classification-Number Correct showing values of .90 or better. For the Trails C scores, values of .70 or higher were obtained for all of the gender and ethnic subsamples.

The standard errors of measurement for the Category Fluency, Classification and Letter Naming TVCF subtests are grouped by the age intervals used to stratify the TVCF normative sample. See the TVCF Examiner’s Manual for the data (Reynolds & Horton, 2006).

For the Trails subtest, the standard error of measurement values was computed from the earlier standardization of the CTMT (Reynolds, 2003). The overall standard error of measurement for Trails C was 6 and further information can be found in the CTMT manual (Reynolds, 2003). Smaller SEMs indicate better reliability.

Reliability estimates may vary by subgroups within a population. The test scores need to be reliable for each tested subgroup that might experience test bias because of racial, ethnic, disability, or linguistic differences (Reynolds, 2000). The TVCF subtest scores are essentially equally reliable for all the subgroups investigated and suggest minimal bias relative to those subgroups. Using the Feldt (see Reynolds, 2000) technique, no significant differences among the reliability values of the groups were found. TVCF subtest score reliabilities were similar for males and females. When European-Americans, African-Americans, Hispanic-Americans, and Asian-Americans were considered, reliability estimates suggested comparable reliability across racial-ethnic groups.

The stability of the Trails C scores over time was investigated using the test–retest method. Thirty adults (i.e., ages 20–57 years) in the Southwestern USA were tested twice, with a 1-week period between testings. Age-corrected T-scores were calculated to control for any effects of age and the values were of sufficient magnitude to allow confidence in the Trail C test scores’ stability over time.

Test–retest studies also allow for assessment of practice effects. Novel tasks and nonverbal tasks such as Trails C tend to show the greatest practice effect (Reynolds & Horton, 2006).

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.