Thesis (MPhil)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: Little scholarly reflection has been published on the subject of medical research and translation. The aim of this study is to contribute to such literature by investigating the quality of original and retranslated medical questionnaires. The various steps medical researchers follow when translating their questionnaires are considered and discussed. Particular attention is given to questionnaires on AIDS-related topics in South Africa, as well as to the role of translation in ensuring the collection of valid data in medical research. Different translation approaches, which are followed when translating medical texts, and the impact they have on the quality of the research, are discussed. These approaches are the linguistic, text-linguistic and functional approaches. Attention is given to translators as communicators and mediators, as well as to the more general role of the translator. This study hypothesises that the quality of translations of medical research questions is largely inadequate in communicating effectively with the target culture for which they are intended. The retranslation hypothesis stating that retranslations are closer to the source text (ST) than original translations is supported. Afrikaans- and Xhosa-speaking adolescents from two secondary schools in the Cape Peninsula participated in a before-after study. These learners received self-administered medical questions on the two occasions. The first set comprised original translations, while the second set contained retranslations of the ST questions. Evaluation questions were included to assess the quality of these translations. The design, translation approach and quality of the original translations are explained, as is the development of the retranslation and evaluation questionnaires. Translations that do not consider their target audience lead to communication gaps, which have an adverse effect on the validity of data derived from questionnaires that are used in medical research. The results of most of these questions are compared for the two target cultures and are analysed qualitatively and quantitatively. The data are further explored to establish whether and how the translational quality of medical questionnaires can be improved. These aspects and the suggested translation process are discussed while bearing in mind the limitations of a study of this kind. Recommendations are made for possible improvement to the quality of translations of medical questionnaires. Projections for further studies in this direction complete this empirical investigation into translation and medical research. / AFRIKAANSE OPSOMMING: Relatief min akademiese nadenke is gepubliseer oor die onderwerp mediese navorsing en vertaling. Hierdie studie wil 'n bydrae maak tot sodanige literatuur deur 'n ondersoek na die kwaliteit van oorspronklik vertaalde en hervertaalde mediese vraelyste. Die onderskeie stappe wat mediese navorsers ten opsigte van die vertaling van hulle vraelyste volg, word bespreek. Aandag word spesifiek gerig op vraelyste oor vigsverwante temas in Suid-Afrika, asook op die rol van vertaling in die versekering dat geldige data in mediese navorsing ingesamel word. Verskeie benaderings wat gevolg word in die vertaling van mediese tekste en die impak wat hulle het op die kwaliteit van die navorsing word bespreek. Hierdie benaderings is die linguistiese, tekslinguistiese en funksionalistiese benaderings. Aandag word geskenk aan vertalers as kommunikeerders en tussengangers, asook die meer algemene rol van die vertaler. Hierdie studie veronderstel dat die kwaliteit van vertalings van mediese navorsingsvraelyste grootliks onvoldoende is om effektief met die betrokke teikengehoor te kommunikeer. Die hervertalingshipotese wat sê dat hervertalings nader aan die brontaal (Bn as oorspronklike vertalings is, word ondersteun. Afrikaans- en Xhosa-sprekende adolessente van twee sekondêre skole in die Skiereiland het deelgeneem aan 'n voor- en agtemastudie. Hierdie leerders het op beide geleenthede die vraelyste self voltooi. Die eerste stel het oorspronklike vertalings bevat terwyl die tweede stel hervertalings van die BT bevat het. Evalueringsvrae is ingesluit om die kwaliteit van hierdie vertalings te help bepaal. Die ontwerp, vertalingsbenadering en kwaliteit van die oorspronklike vertalings word verduidelik, so ook die ontwikkeling van die hervertaling- en evalueringsvraelyste. Vertalings wat nie die teikengehoor in ag neem nie, lei tot kommunikasiegapings wat die geldigheid van data afkomstig van vraelyste in mediese navorsing nadelig kan raak. Die resultate van die meeste van hierdie vrae word vir die twee teikenkulture vergelyk, en dit word kwalitatief en kwantitatief ontleed. Die data word verder ondersoek om vas te stelof en hoe die kwaliteit van die vertaling van mediese vraelyste verbeter sou kon word. Bogenoemde aspekte en die voorgestelde vertalingsproses word bespreek met inagneming van die beperkings van 'n studie van hierdie aard. Voorstelle word gemaak vir die moontlike verbetering van die vertaling van mediese vraelyste. Vooruitskattings vir verdere studie in hierdie rigting voltooi hierdie empiriese ondersoek na vertaling en mediese navorsing. Ek dra hierdie tesis op aan Mattheus vir sy liefde, geduld en al die tee-aandraery, asook aan Ettienne en Jeanelle wat met minder ondersteuning van my as andersins hulle skoolloopbane moes voltooi het.