The failure rate in mathematics at school is high, not only in South Africa, but also internationally. Furthermore, learners with an apparently high general ability or aptitude for mathematics sometimes underachieve in the subject, while some learners with an apparently low general intellectual ability or aptitude for mathematics sometimes achieve well in the subject. Little attention is nonetheless given to learners' study orientation in mathematics, in spite of the fact that research has indicated that school mathematics is one of the best predictors of success in tertiary studies. An investigation into some epistemological approaches to the learning process in mathematics confirms that learners' achievement in mathematics is significantly affected by their study orientation in mathematics. The primary aim of this thesis was the development and evaluation of a study orientation questionnaire (SOM) in mathematics. Data processing procedures especially referred to the following two primary aims with the study: (a) Standardisation of the questionnaire. Steps carried out to evaluate the questionnaire psychometrically, include factor and item analysis. In the case of Grade 8 and 9, three fields (Study habits in mathematics, Mathematics anxiety and Study attitudes in mathematics) were identified. A fourth field, Locus of control regarding mathematics, was identified only in the case of learners in Grade 1 0 and ll. It was established that the SOM apparently has criterion related validity, as well as content and construct validity for the three language groups as a whole. Reliability coefficients for the SOM can in most cases be regarded as satisfactory. (b) Comparative studies to determine the applicability of the SOM. Analysis of variance techniques were used to determine where significant differences between groups (including grade-, mother tongue and sex groups) lay. Where MANOV AS showed significant differences, further investigation was carried out to determine in respect of which individual fields (single variables) groups differed significantly. By means of LSM it was determined which groups differed significantly in regard of the separate fields. Some of the findings include: * It seems that African language speakers in both grade groups are really trying to achieve in mathematics, but that their best efforts are not successful. * It would appear that girfs' level of Mathematics anxiety drops in Grade l 0 and ll. * Learners in Grade l 0 and ll show lower levels of Mathematics anxiety and more sufficient Study habits in mathematics than their counterparts in Grade 8 and9. * Perhaps the most significant finding is the phenomenon that African language learners in Grade l 0 and l l achieved much worse in regard of Locus of control than Afrikaans and English-speaking learners. A number of factors probably contribute towards this state of affairs, including language problems, teachers who are underqualified, African language learners' less than optimal socioeconomic status (SES) in general, a lack of facilities and text books and disruption which is still being experienced in rnnny traditionally black schools. It is recommended that these matters ·are attended to in an effort to create circumstances for more sufficient achievement in mathematics by learners from all language groups. / Thesis (PhD)--University of Pretoria, 1997. / Human Sciences Research Council (South Africa) / Psychology / PhD / Unrestricted