A comparative study of the relative incidence of various vocal parameters among Black and white college females

Abstract

The purpose of this study was to make a comparative investigation of the voice quality of black and white women. A review of the literature revealed considerable research has been conducted in the area of black-white language differences in the United States, but 2) Can ethnic group identity be ascertained according to perception of voice quality more often than expected by chance?Indexical information on the subjects was collected for the following variables: 1) age, 2) length of longest sustained phonation, 3) optimum pitch, 4) modal pitch, 5) relationship of modal pitch to optimum pitch, 6) lowest possible pitch, 7) highest possible pitch, and 8) pitch range in semitones. The data analysis revealed no statistically significant differences between black and white women regarding these variables.In the experiment involving the JHVP, fifty samples of women's voices {twenty-five Blacks and twenty-five Whites) were submitted to listener analyses of voice quality by ten women judges, two of whom were trained speech pathologists and eight were in training at the graduate level in speech pathology. The voice samples were heard by reverse audio tape playback to eliminate dialectical differences. The validity of the ratings of the J1IVP is questionable, based on the poor inter-judge reliability. Recognizing this qualification, the ratings were nevertheless submitted to data analysis with the following results: Statistically significant differences observed were that black women were more often rated as too low-pitched during speech, hyponasal, and more "severe," while white women were more often rated as too soft and too breathy. No difference in the incidence of perceived hoarseness was fom1d.In Uw second experiment, brief segments of each of the fifty women's voices were presented by reversed playback in a randomized order for a forced listener choice of ethnic identity: Black or White.The results indicated such identity could not be made at a greater than chance level.