Computer-Based Assessment of Word Knowledge in Teens With Learning Disabilities

• Published in: Language, speech & hearing services in schools Vol. 27; no. 1; pp. 21 - 28
• Main Authors: Wiig, Elisabeth H, Jones, Stephanie S, Wiig, Erik D
• Format: Journal Article
• Language: English
• Published: ASHA 01.01.1996
• Subjects: Adolescents; Computer Applications; Computer Assisted Testing; Expressive Language; Learning Disabilities; Lexicon; Norm Referenced Tests; Receptive Language; Secondary Education; Standardized Tests; Test Norms; Test of Word Knowledge; Test Validity; Vocabulary Skills
• ISSN: 0161-1461; 1558-9129
• DOI: 10.1044/0161-1461.2701.21

Computer-based and standard administrations of the Test of Word Knowledge (TOWK)-Level 2 core subtests (Word Definitions, Synonyms, Figurative Usage, Multiple Contexts) (Wiig & Secord, 1992) were compared for 30 subjects with learning disabilities. Half completed the computer-based version first and half the standard version first. Three weeks later, subjects were given either the standard or computer-based version in a counterbalanced design. The total, receptive, and expressive composites and three subtest means were highest for the standard administration ( p <.05). Correlation coefficients ( r ) ranged from .81 to .88 for composites ( p <.01). Confidence intervals (at 90% level) overlapped in 93% for the total, 97% for the receptive, and 93% for the expressive composite. Paired total scores were within the same diagnostic category (below or at/above normal) for 87%, receptive for 83%, and expressive for 90% of the subjects. Differenes in task formats influenced the results, indicating the need for independent norms for computer-based adaptations of standardized language tests.