Measurement of intrinsic deficiency in transmitted speech: The diagnostic discrimination test (DDT)

Conventional intelligibility tests do not normally permit a distinction to be made between intrinsic deficiencies of speech-processing or communication systems (in which irrevocable loss of useful information occurs) and cosmetic deficiencies (in which there is simply a failure to render the transmitted speech into a perceptually assimilable form). Because such tests require a recognition response from the listener, they are of little value for testing, for example, prosthetic devices for the congenitally deaf, who have no prior experience with, nor associated labels for, the elementary sounds of speech. Conceivably they may occasionally lead to unduly pessimistic characterizations of the effectiveness of certain types of speech coding systems. The Diagnostic Discrimination Test offers one remedy for this situation. It draws from the same speech materials as the Diagnostic Rhyme Test (1), but requires the listener simply to judge whether the members of pairs of test words are the same or different. A correct response thus indicates that information regarding the state of one of six distinctive features has been preserved, even where the test words are not recognizable as such.