SPECTRAL DISPLAYS OF SPEECH AS AN AID TO READING

THE PROPOSED PROJECT IS DESIGNED TO TEST THE EFFECTIVENESS OF REAL-TIME SPECTRAL DISPLAYS OF SPEECH IN TEACHING READINGTO FIRST-GRADE LANGUAGE-IMPAIRED (LI) CHILDREN. YOUNG LI CHILDREN ARE DELAYED IN DEVELOPMENT OF RAPID-RATE AUDITORY PROCESSING AND SPEECH PERCEPTION SKILLS. THEY ALSO HAVE DIFFICULTY WITH PHONICS IN READING. SPECTRAL DISPLAYS OF SPEECH MAY HELP THEM TO SEGMENT SPEECH INTO SYLLABLES AND PHONEMES, AND TO MATCH SPOKEN WORDS WITH WRITTEN FORMS AND PHONEMES WITH GRAPHEMES, THUS FACILITATING EARLY READING ACQUISITION. IN PHASE I, 12 LI CHILDREN OF 51/2 TO 7 YEARS WILL BE TAUGHTPHONICS WITH AND WITHOUT THE AID OF SPECTRAL DISPLAYS. THEIR PERFORMANCE ON TESTS OF READING MONOSYLLABLES AND MATCHING MONOSYLLABIC WORDS WITH PICTURES WILL BE COMPARED BEFORE AND AFTER TRAINING, AND UNDER THE TWO TRAINING CONDITIONS. SCHEMATIZED SPECTROGRAMS WILL ALSO BE GENERATED, IN COLOR, USING AN IBM-COMPATIBLE PERSONAL COMPUTER. IN PHASE II, AMPLITUDE DISPLAYS WILL ALSO BE GENERATED AND DISPLAYED SIMULTANEOUSLY WITH SPECTROGRAPHIC DISPLAYS AND WITH PRINTED GRAPHEMES. THE SYSTEM WILL BE FIELD TESTED AND MODIFIED TO INCREASE EFFECTIVENESS. THE GOAL WILL BE TO MAKE THE SYSTEM COMMERCIALLY AVAILABLE ALONGWITH APPROPRIATE TRAINING PACKAGES.