The low grades that students score in some statistical units in Kenyan universities is of great concern and has evoked research interest in the teaching of some of the units and the students’ learning of the statistical content. The aim of the study was to investigate the difficulties undergraduate students experience in the learning of bivariate normal distribution in a Kenyan university. The research also aimed to answer the following research questions on the difficulties undergraduate students encounter in the learning of bivariate normal distribution. The first research question was based on the reasons why students find learning of bivariate normal distribution difficult and the second research question was to find the reasons why students experience such difficulties in learning bivariate normal distribution. The target population for this study included lecturers teaching statistics in the university, and second- and third- year students enrolled or who have previously completed the probability and statistics III unit, where the bivariate normal distribution content is covered. In selecting students for the study, the simple random sampling technique was employed while convenient sampling was used to select lecturers who participated in the study. A mixed methods design was adopted for this study where both quantitative and qualitative data was collected. A total of 175 students and six lecturers participated in this research study. All students who participated in the study did a bivariate normal distribution test (Appendix 1) designed by the researcher and then filled in a questionnaire (Appendix 2). The lecturers who participated in the study filled in an open-ended questionnaire (Appendix 3). The results showed that undergraduate students have difficulties in learning bivariate normal distribution. This is because most of them could neither state the bivariate normal distribution nor solve any of the application questions on the content. The students find it difficult to learn and comprehend the bivariate normal distribution equation with its many parameters and constants of the two random independent variables. The results also showed that students could not state the normal distribution equation nor could they solve questions on the normal distribution, which forms the foundational knowledge required for effective learning of the bivariate normal distribution content. ii Based on the results, the study recommended that emphasis should be placed on the basic and foundational knowledge of the normal distribution content and its applications before teaching bivariate normal distribution in probability and statistics III. In addition, it is recommended that all students should be involved in the learning of basic content to enable them to understand all parameters and constants in the equations and their applications. The study also recommends that lecturers revise the foundational knowledge and content related to the bivariate normal distribution before introducing and teaching the bivariate normal distribution content. This study also recommends that the university should consider a change of curriculum by teaching the bivariate normal distribution, as an introductory course to the unit under the multivariate distributions in statistics, in third year of the students’ studies. ; ; / Mathematics Education / M. Sc. (Mathematics Education)

This study explored language practices of trilingual undergraduate students of mathematics as they made sense of an algebraic task. Specifically, the study set out to explore whether, how and why trilingual undergraduate students used language(s) to make sense of mathematics. In this study a trilingual speaker is viewed as an individual proficient in three languages and whose proficiency in the languages is not necessarily equal. The speaker uses the three languages either separately or by switching between any two in ways that are determined by his/her communication needs. Exploring language practices helped me to understand how students position themselves as they engage with a mathematics task using mathematical Discourses (capital D) in relation to their trilingual language facility. This facility involves the use of either the language of learning and teaching (LoLT) (English) or the switching between two or three of the languages they know. These languages were home languages, including Kiswahili of the students. In tertiary institutions, English is the LoLT while the home languages are neither taught nor used in the classroom. The study used a qualitative inquiry process, specifically a case study approach. It was conducted at a public university in Kenya with a focus on first-year engineering students with mathematics in their programme. Data were collected using a students‟ questionnaire, and clinical and reflective interviews. A structured questionnaire was used to gather the baseline data, which was used for the selection of 15 interview participants. The clinical interviews provided information on language use as the students engaged with the task, explaining each step of the process, while the aim of the reflective interviews was to identify, ascertain and confirm various actions and different languages and language practices that were not apparent during the clinical interview. The interviews were transcribed and 11 paired transcripts were selected for analysis. The data were analysed using the methods of Discourse analysis (Gee, 2005). This analysis explored how students used language in tandem with non-language “stuff” in a single language or when switching between any two languages and how and why each was used. The focus was on the activities and identities they enacted through their interpretation of the given task and in part of the solution process. The findings revealed that when students engaged with mathematics, they drew on the LoLT only, or switched between the LoLT and their home languages or between the LoLT, home languages and Kiswahili. Those who switched did so when they were faced with interpretation challenges, when there was need to emphasise a point and due to habitual practices of switching. They commonly switched silently and communicated verbally in the LoLT. The purpose for code switching was to gain understanding of the task. On the other hand, a trilingual student is likely to remain in the LoLT because content has been taught and tasks presented in the LoLT. The key contribution of this study is its focus on the trilingual language context of undergraduate students of mathematics, an area that has not been researched up to now. Furthermore, this study has added to scholarly work in this discipline by establishing that code switching is not the preserve of students who are learning the LoLT; rather, it is a reality for trilingual students who are competent in the LoLT when they engage with mathematics. / Mathematics Education / D. Ed. (Mathematics Education)

This study explored language practices of trilingual undergraduate students of mathematics as they made sense of an algebraic task. Specifically, the study set out to explore whether, how and why trilingual undergraduate students used language(s) to make sense of mathematics. In this study a trilingual speaker is viewed as an individual proficient in three languages and whose proficiency in the languages is not necessarily equal. The speaker uses the three languages either separately or by switching between any two in ways that are determined by his/her communication needs. Exploring language practices helped me to understand how students position themselves as they engage with a mathematics task using mathematical Discourses (capital D) in relation to their trilingual language facility. This facility involves the use of either the language of learning and teaching (LoLT) (English) or the switching between two or three of the languages they know. These languages were home languages, including Kiswahili of the students. In tertiary institutions, English is the LoLT while the home languages are neither taught nor used in the classroom. The study used a qualitative inquiry process, specifically a case study approach. It was conducted at a public university in Kenya with a focus on first-year engineering students with mathematics in their programme. Data were collected using a students‟ questionnaire, and clinical and reflective interviews. A structured questionnaire was used to gather the baseline data, which was used for the selection of 15 interview participants. The clinical interviews provided information on language use as the students engaged with the task, explaining each step of the process, while the aim of the reflective interviews was to identify, ascertain and confirm various actions and different languages and language practices that were not apparent during the clinical interview. The interviews were transcribed and 11 paired transcripts were selected for analysis. The data were analysed using the methods of Discourse analysis (Gee, 2005). This analysis explored how students used language in tandem with non-language “stuff” in a single language or when switching between any two languages and how and why each was used. The focus was on the activities and identities they enacted through their interpretation of the given task and in part of the solution process. The findings revealed that when students engaged with mathematics, they drew on the LoLT only, or switched between the LoLT and their home languages or between the LoLT, home languages and Kiswahili. Those who switched did so when they were faced with interpretation challenges, when there was need to emphasise a point and due to habitual practices of switching. They commonly switched silently and communicated verbally in the LoLT. The purpose for code switching was to gain understanding of the task. On the other hand, a trilingual student is likely to remain in the LoLT because content has been taught and tasks presented in the LoLT. The key contribution of this study is its focus on the trilingual language context of undergraduate students of mathematics, an area that has not been researched up to now. Furthermore, this study has added to scholarly work in this discipline by establishing that code switching is not the preserve of students who are learning the LoLT; rather, it is a reality for trilingual students who are competent in the LoLT when they engage with mathematics. / Mathematics Education / D. Phil. (Mathematics, Science and Technology Education (Mathematics Education))