SYNTHETIC SPEECH AUDIOMETRIC SCREENING TEST

THIS RESEARCH WILL DETERMINE THE FEASIBILITY OF A NEW TEST OF AUDITORY FUNCTION BASED ON DISCRIMINATION OF PHONETIC CUES IN SYNTHETIC SPEECH. THE SPEECH WILL BE SYNTHESIZED WITH PRECISELY CONTROLLED PARAMETERS. AS PLANNED, A LISTENER WOULD BE PRESENTED WITH A SERIES OF SINGLE WORDS OF SYNTHETIC SPEECH AND ASKED TO IDENTIFY A PRINTED REPRESENTATION OF THE WORD WITHIN A SET OF SIMILAR WORDS. THE SYNTHESIZED WORD AND THE OTHER WORDS DIFFER BY A SINGLE PHONEME (E.G., SIT, SAT, FAT, FIT). THESE WORDS HAVE BEEN SYNTHEISZED; THEREFORE, THEIR SPECTRALAND TEMPORAL PROPERTIES ARE COMPLETELY KNOWN. BECAUSE CORRECT IDENTIFICATION OF THE TEST WORD DEPENDS ON THE AUDIBILITY OF THE ACOUSTIC FEATURE DISTINGUISHING THE TEST WORD, RESULTS OF THE TEST ALLOW INFERENCES TO BE DRAWN ABOUT A LISTENER'S HEARING SENSITIVITY IN THE FREQUENCY REGION CHARACTERISTIC OF THAT ACOUSTIC FEATURE.