A dinâmica de crescimento populacional de uma espécie é permeada pela relação entre as desvantagens da competição intraespecífica e os benefícios da presença de conspecíficos. Para muitas espécies, os benefícios da cooperação podem superar as desvantagens da competição. A correlação positiva entre tamanho populacional e adaptabilidade em populações muito pequenas é conhecida como efeito Allee demográfico. Apesar de haver modelos matemáticos isolados para os diferentes tipos de efeitos Allee, não há um modelo simples que os abranja e os conecte a modelos de crescimento mais gerais (como o de Richards). Propomos unificar modelos de efeitos Allee e o de crescimento de Richards em um modelo que permita um novo ponto de vista sobre o efeito Allee demográfico. Um exemplo do aumento das possibilidades descritivas de tal generalização é a emergência de mais de uma transição cooperação-competição quando considerado um caso particular desse novo modelo (Allee-Gompertz). Apesar da importância do crescimento populacional, a maioria dos modelos básicos de transmissão de doenças infecciosas considera o tamanho populacional constante ou adota simplificações pouco plausíveis. Nesta tese, mostramos as deficiências de um modelo compartimental dinâmico de tuberculose já consagrado e propomos um novo modelo com crescimento populacional logístico. Quando comparados, nosso modelo apresenta previsões mais pessimistas para a erradicação da doença a longo prazo quando testado com parâmetros que definem políticas de controle pouco eficientes. Realizamos tais predições adotando estratégias de controle de países desenvolvidos e subdesenvolvidos. Visto que esses modelos compartimentais desprezam aspectos espaciais, desenvolvemos uma modelagem computacional de agentes, baseada no modelo proposto, com duas estruturas subjacentes: redes aleatórias e redes reais. A súbita emergência de tuberculose resistente a drogas como consequência de tratamentos ineficazes é também um resultado das implementações desses modelos em dois cenários distintos. Esses resultados são comparados com os do modelo compartimental e com os de um modelo de estrutura subjacente mais simples e, como novo resultado, surge nos dois modelos a possibilidade de erradicação da doença em menos de uma década após o início do tratamento. Esse resultado é possível desde que sejam adotadas estratégias eficientes de controle. / The one-species population growth dynamics is permeated by the relationship between the harms from the intraspecific competition and the benefits from the presence of conspecifics. For many species, the benefits from conspecific cooperation may outweigh the harms from competition. The positive correlation between population size and total fitness in very small population known as demographic Allee effect. Although there are isolated mathematical models for different types of Allee effects, there is not a simple model that covers and connects them to more general growth models (like Richards). We propose to unify models of Allee effects and the Richards growth one in a model that allows a new perspective on the demographic Allee effect. An example of the increased descriptive possibilities of such generalization is the emergence of more than one transition cooperation-competition when considering a particular case of this new model (Gompertz-Allee). Despite the importance of population growth, most basic models of infectious diseases transmission considers population size constant or adopts implausible simplifications. In this thesis, we show the shortcomings of a dynamic compartmental model of tuberculosis already established and we propose a new model with population logistic growth. When compared, our model provides more pessimistic forecasts for the eradication of the disease in the long term if it is tested with parameters that define inefficient control policies. We perform such predictions adopting control strategies from developed and underdeveloped countries. Since these compartmental model disregards spatial aspects, we developed a computational agent model, based on the proposed model, with two underlying structures: random networks and real networks. The sudden emergence of drug-resistant tuberculosis as a result of ineffective treatments is also a result from the implementations of these models in two distinct scenarios. These results are compared with the ones from a compartimental model and with the ones from a model with simpler underlying structure and, as a new result, the possibility of eradicating the disease in less than a decade after beginning the treatment appears on the two models. This result is possible adopting effective control strategies.

The term aggregation refers to overdispersion and both are used interchangeably in this thesis. In addressing the problem of prevalence of infectious parasite species faced by most rural livestock farmers, we model the distribution of faecal egg counts of 15 parasite species (13 internal parasites and 2 ticks) common in sheep and goats. Aggregation and excess zeroes is addressed through the use of generalised linear models. The abundance of each species was modelled using six different distributions: the Poisson, negative binomial (NB), zero-inflated Poisson (ZIP), zero-inflated negative binomial (ZINB), zero-altered Poisson (ZAP) and zero-altered negative binomial (ZANB) and their fit was later compared. Excess zero models (ZIP, ZINB, ZAP and ZANB) were found to be a better fit compared to standard count models (Poisson and negative binomial) in all 15 cases. We further investigated how distributional assumption a↵ects aggregation and zero inflation. Aggregation and zero inflation (measured by the dispersion parameter k and the zero inflation probability) were found to vary greatly with distributional assumption; this in turn changed the fixed-effects structure. Serial autocorrelation between adjacent observations was later taken into account by fitting observation driven time series models to the data. Simultaneously taking into account autocorrelation, overdispersion and zero inflation proved to be successful as zero inflated autoregressive models performed better than zero inflated models in most cases. Apart from contribution to the knowledge of science, predictability of parasite burden will help farmers with effective disease management interventions. Researchers confronted with the task of analysing count data with excess zeroes can use the findings of this illustrative study as a guideline irrespective of their research discipline. Statistical methods from model selection, quantifying of zero inflation through to accounting for serial autocorrelation are described and illustrated. / Statistics / M.Sc. (Statistics)

The results of mathematics level 3 have always been a problem at TVET colleges as this hampers the certification rate and the progress of the students to level 4. Students who did not do well in the current subject are not allowed to register that subject in the following level. Even though the students are allowed to progress to level 4 they won’t be certificated for both levels until they pass the remaining subject. The above challenges made the researcher to check during the marking and moderation of November / December examination the course of poor results for mathematics level 3. In the process of checking the researcher discovered that rectangular hyperbola is one of the topics that the students of mathematics level 3 are struggling with. This study therefore focuses on exploring the challenges faced by TVET Level 3 NCV students in understanding the hyperbolic function in mathematics. In addition to the literature review, an empirical investigation based on a qualitative approach and involving semi-structured interviews with the students of a TVET college in North West was conducted to collect data. The analysis of documents relevant to the study was also used as the other method. The study used participatory action research, where the researcher, collaborators and students work alongside each other to collect data and to improve practice and follow the spiral pattern of reflection, analysing the results and adapting the action. The research design and methodology was qualitative. This helped the researcher to understand the challenges students faced in the learning of rectangular hyperbola and also came up with ways to minimise those challenges. The data collection methods used was interviewing using semi-structured questions, pre-test and post-tests. During data collection different interventions (IN1 –IN3) was used depending on the understanding of the students. For ethical consideration, ethical clearance was obtained from UNISA. DHET, the principal of the college, collaborators, parents and students will also give written consent on forms which will be sent out explaining what we envisage. Since research was voluntary, an explanation was given that this was not compulsory and that participation was completely voluntary and that they could withdraw at any time. In this study, various methods to empower students were recommended. Recommendations are also made on what was found in this study, as are recommendations for further study. / Mathematics Education / M. Ed. (Mathematics Education)