Die vertaling en ontwikkeling van sinsmateriaal vir die evaluasie van spraakpersepsie by Xhosa-sprekendes (Afrikaans)

Abstract

AFRIKAANS: Hierdie studie het die eerste fase van die ontwikkeling van toetsmateriaal vir die evaluasie van gehoorgestremde Xhosa-sprekendes behels. In die Iig van die Suid¬Afrikaanse klimaat van kulturele- en taaldiversiteit is 'n behoefte aan kultuurgeldige toetsmateriaal binne die veld van oudiologiese rehabilitasie geïdentifiseer. Vorige navorsing het uitsluitlik op die samestelling van woordelyste gefokus. Hierteenoor verskaf sinsmateriaal hoë-vlak kontekstuele Iinguïstiese inligting. Die geldigheid van die evaluasieproses word gevolglik verhoog aangesien die individu se vermoë om residuele gehoor te gebruik, binne 'n meer verteenwoordigende luisteromgewing geëvalueer word. Gedurende die ontwerpfase van die studie is die Bamford-Kowal-Bench (BKB) Sentence Lists deur 'n moedertaalspreker van Xhosa vertaal. Twee komitees was onderskeidelik vir die grammatikale en kulturele korreksie en die oorsig oor die ouderdomstoepaslikheid van die vertaalde sinslyste verantwoordelik. Aanpassings is op grond van die komitees se aanbevelings gemaak, waarna die sinslyste aan Studie A van die ontwikkelingsfase onderwerp is. Met die oog op die evaluasie van die kulturele-, dialektiese- en ouderdomstoepaslikheid van die woordeskat, is sestig proefpersone geselekteer. Daar is gepoog om 'n balans te verkry ten opsigte van geslag, ouderdom (8-15 jaar en 18-25 jaar) en gebied (stad en platteland). Analise van variansie ontledings het geen statisties beduidende veskille tussen die onderskeie sinslyste aangedui nie. Ten einde die betroubaarheid van metings te verhoog, is die volledige stel sinne vervolgens aan 'n rekenaarprogram onderwerp met die oog op die samestelling van 'n finale stel sinslyste wat gebalanseer is ten opsigte van foneeminhoud. Tien Iyste bestaande uit tien sinne elk is uiteindelik saamgestel en aan Studie B van die ontwikkelingsfase onderwerp. Met die oog op die kontrolering van die ekwivalensie van saamgestelde sinslyste is sestien proefpersone vir hierdie Studie geselekteer. Die finale stel van tien sinslyste is teen 'n vasgestelde sein-tot-ruis verhouding aan hierdie proefpersone aangebied en response is statisties ontleed. Die Tukey Groeperings het by 'n 95% vertrouensinterval bevestig dat sinslyste 6 en 10 beduidend van die ander Iyste verskil het ten opsigte van moeilikheidsgraad. Hierdie Iyste is uit die toetsmateriaal weggelaat. Die laaste hoofstukke van die studie behels 'n kritiese evaluering van die studie asook voorstelle vir verdere navorsing met die oog op standaardisasie van hierdie materiaal. ENGLISH: This study involved the first phase of the development of evaluation material for the assessment of hearing-impaired Xhosa speakers. In light of the present South African climate of cultural and language diversity, the need for culturally valid evaluation material in the field of audiological rehabilitation, was identified. Previous research focused exclusively on the compilation of wordlists. Sentences, however, provide high level contextual linguistic information. The validity of evaluation is thus increased as a result of evaluating the individual within a more representative listening environment. During the design phase of the study, the Bamford-Kowal-Bench (BKB) Sentence Lists were translated by a mothertongue Xhosa speaker. Two committees were responsible for the grammatical and cultural correction as well as an overview of the age appropriateness of vocabulary. Based on the committees' recommendations certain adaptations were made before subjecting the material to Study A of the development phase. In order to evaluate the cultural-, dialectal and age appropriateness of the vocabulary, sixty subjects were selected. Subjects were balanced in terms of gender, age (8-15 years and 18-25 years) and geographical area (urban and rural). Statistical analysis of variance indicated no significant difference amongst the sentence lists. In order to increase the reliability of measurements, the entire set of sentences were subject to a computerized programme to compile phonemically balanced lists. Ten lists of ten sentences each were compiled and implemented in Study B of the development phase. Sixteen subjects were selected to participate in this study in order to verify the equivalency of the ten lists. The final set of ten lists was presented at a fixed signal-tonoise ratio and subjects' responses were analysed statistically. The Tukey Grouping confirmed at a 95% confidence interval, that Lists 6 and 10 were significantly different in terms of difficulty. These lists were omitted from the final material. The last chapters of the dissertation included a critical evaluation of results as well as recommendations for future research aimed at standardisation of the material.