Sign language iconicity and test construction theory for deaf individuals.

摘要

The role of iconicity and the extent to which it exists in American Sign Language (ASL) are unresolved issues. A practical concern is the validity of translating English based tests into signs without taking into account the possible effects of sign iconicity on test performance. It becomes difficult to differentiate those items that were correctly answered because the individual understood the linguistic symbol from those that had iconic cues that led the child to the correct answer. This study investigated the effects of manipulating distractor items of the Peabody Picture Vocabulary Test-Revised (PPVT-R) (Dunn, 1981), on the test response accuracy of non-signing, hearing college students and their signing deaf college peers. Thirty-two (32) deaf signers and 34 hearing non-signers served as subjects. A congenitally deaf, native signer administered the test to all subjects. The test was comprised of 40 items, 20 of which were as they appear in the original PPVT-R test format and 20 items manipulated to provide 3 new distractor pictures that had a similar appearance to the sign presented. There was a significant main effect for hearing status indicating that deaf signers demonstrate better overall performance than hearing non-signers. There was a significant hearing status by distractor type interaction (p =.001), as predicted. Results indicated that iconicity can be an important factor in the development of limited-choice format, receptive vocabulary tests using manual signs. The most striking implication of the present study is the fact that, for purposes of validity, the use of non-signing subjects in test development would serve to minimize the possibility that items would be measures of perceptual matching skills and maximize the possibility that the test would truly be a measure of receptive sign vocabulary. Recommendations are provided.