Consonant phonemic errors associated with pure-tone configurations and certain kinds of hearing impairment.

Abstract

In the course of developing multiple-choice items for speech discrimination testing, phonemic errors were observed as they relate to pure-tone configurations and to certain types of hearing impairment. For the phonemes involved, the following observations were made: (1) The /s, ∫, t∫, d 3 / and the initial /t/ and /θ/ were easily identified by patients with flat pure-tone configurations, but were difficult for patients with sharply falling slopes, 500 to 4000 Hz. (2) Identification of the /s/ and the initial /t/ and /θ/ was highly dependent upon energy in the frequency range above 2000 Hz, whereas identification of the /∫, t∫, d 3 / was highly dependent upon the range between 1000 and 2000 Hz. (3) Over all the items testing a given stimulus phoneme, the total number of phonemes employed as alternate responses ranged from three to eight, averaging five. The actual erroneous responses for any given stimulus phoneme, however, were usually limited to two or three phonemes, and these were generally the same regardless of pure-tone configuration. (4) Although the error-response phonemes were usually the same as the stimulus phoneme in manner of production, some error phonemes were produced in a different manner, but in the same place, as the stimulus phoneme. Probability of error for individual phonemes seemed to be more closely related to pure-tone configurations than to kinds of hearing impairment.