School of Statistics and Planning (SSP) Collections
Permanent URI for this collection
Browse
Recent Submissions
1 - 5 of 426
-
ItemDeterminants of attitude towards female genital mutilation among women in the Gambia(Makerere University, 2025-12)Female Genital Mutilation/Cutting (FGM/C), also known as female circumcision, is a harmful traditional practice (HTP) with profound health and social implications for women and young girls. The high prevalence of FGM in The Gambia is a pressing issue. This study investigated the determinants of women's attitudes towards FGM in The Gambia. The study used secondary data from the Gambia Demographic and Health Survey (GDHS) 2019/2020. A weighted sample of 5,632 women aged 15–49 was analyzed at univariate, bivariate, and multivariate levels. A binary logistic regression model was fitted at the multivariate level. About half (50%) of women reported that FGM should be continued. The study found that women with secondary (AOR=0.73, CI=0.62–0.87) or higher education (AOR=0.34, CI=0.25–0.49) had lower odds of supporting FGM compared to those with no education. Women engaged in agriculture (AOR=1.21, CI=1.01–1.43) and those from middle-income households (AOR=1.52, CI=1.23–1.87) were more likely to support FGM, while those from the richest households (AOR=0.64, CI=0.49–0.83) were less likely. Women from Christianity and other religions (AOR=0.13, CI=0.06–0.29) were less likely to support FGM continuation compared to those from Islam. Ethnicity and region also showed significant associations, with women from certain tribes and regions being more or less likely to support FGM. In conclusion, the comprehensive analysis of demographic, socio-economic, and cultural factors underscores the complex interplay influencing attitudes towards female genital mutilation. The practice is largely supported by women of low socio-economic status and remains strongly influenced by religion and culture, highlighting the need for multifaceted interventions. The study recommends strengthening education and awareness programs targeting both women and men to emphasize the health risks and human rights violations associated with FGM. Community-based advocacy involving religious and traditional leaders should be promoted to challenge cultural beliefs sustaining the practice. Furthermore, economic empowerment programs for women should be enhanced to reduce dependency on traditional practices as sources of identity or income, while policy enforcement and monitoring should be intensified to ensure effective implementation of the anti-FGM law. Subject keywords; Female genital mutilation, Gambia, Women
-
ItemFactors associated with the healing time from obstetric fistula at Kitovu Mission Hospital, Uganda(Makerere University, 2025)In Uganda, about 2% of women who have ever given birth have experienced obstetric fistula. Patients with obstetric fistula can heal within 10-14 days but can take longer depending on several factors which increases the burden on the patient and the healthcare system. In Kitovu hospital, the time to obstetric fistula healing and the associated factors were not known. The study determined the mean and median time to obstetric fistula healing and its associated factors among patients managed at Kitovu mission hospital in Uganda. The study used a retrospective design. Relevant data were collected on 3793 study participants from the Kitovu hospital obstetric fistula database. The study used measures of central tendencies to summarize the characteristics of the study participants. The Kaplan-meier analysis was used to determine the average time to obstetric healing. The log-rank test and the discrete time logit model were used to determine the factors that are significantly associated with time to obstetric fistula healing. The study revealed that the mean and median time to obstetric fistula healing were 56 and 17 days respectively. Economically independent women were more likely to heal faster than economically dependent women (OR=1.352; 95% CI=1.240-1.473; p=0.000). Women with a fistula located > 3cm away from the vagina were more likely to heal faster than women with a fistula located < 3 cm away from the vagina (OR=1.390; 95% CI= 1.240-1.559; p=0.008). Women with one fistula are more likely to heal faster than those with multiple fistulae (OR=2.174; 95% CI 1.797-2.629; p=0.000). Overweight women were less likely to heal faster that those with normal weight and underweight (OR=0.864; 95% CI=0.800-0.934; p=0.000). Women with a fistula ≥ 3cm in diameter were less likely to heal faster than women with a fistula <3cm in diameter (OR=0,614; 95% CI= 0.540-0.699; p=0.000). Women with a large amount of fistula output were less likely to heal faster than women with moderate output or no output (OR=0.217; 95% CI=0.195-0.241; p=0.000). The study made the following recommendations. Kitovu Mission Hospital should strengthen targeted support programs for economically vulnerable obstetric fistula patients which should include financial assistance. The Ministry of Health, should review and strengthen existing obstetric fistula policies and clinical guidelines to provide for early initiation of treatment and standards for management of complex fistula types. Research institutions should conduct more research to triangulate the study findings and generate more context specific evidence on obstetric fistula in Uganda.
-
ItemMale circumcision uptake in Uganda and it's associated factors: a structural equation modelling approach(Makerere University, 2025)The World Health Organization (WHO) and UNAIDS recommend male circumcision as a preventive measure against HIV, with evidence showing about a 60% reduction in the risk of male-to-female HIV transmission. This study aimed to identify factors influencing the uptake of male circumcision in Uganda. Data from 5,008 men aged 15–54 years were drawn from the Uganda Demographic and Health Survey. A Generalized Structural Equation Model (GSEM) was applied to simultaneously estimate both direct and indirect effects of potential predictors on circumcision uptake. Results from direct effect analysis show that, holding other factors constant, increased likelihood of having undergone male circumcision was associated with men with secondary education (coef = 0.153, p=0.000) compared to men with no education, men who are rich (coef= 0.076, p =0.000) compared to men who are poor, men living with a partner (coef= 0.087, p =0.000) compared to men never in union. However an increase in the age of respondents was associated with less likelihood of having undergone male circumcision. Indeed the likelihood of having undergone male circumcision was lower for age group 35-44(coef= -0.139, p=0.000) and age group 45-54(coef= -0.176, p= 0.000) compared to age group 15-24. ). Also rural residents were less likely to have undergone male circumcision (coef= -0.069, p =0.000) compared to urban residents. Region was also found to significantly affect male circumcision status. Indeed, men from northern (coef = -0.237, p = 0.000) and western (coef=-0.108, p=0.000) were less likely to have undergone male circumcision compared to men from Central. Indirect effect analysis showed no significant mediation through comprehensive HIV knowledge, indicating that observed associations were driven entirely by direct effects. Increasing male circumcision uptake in Uganda requires targeted health education for men with lower education levels, expanded free or subsidized services for poorer men, and couple‑focused communication to encourage partner support. Age‑specific interventions should address older men’s concerns about pain and healing, while rural access must be improved through mobile clinics, community health workers, and local leadership engagement. Finally, region‑specific strategies that respond to cultural beliefs and service gaps are essential to ensure equitable uptake across the country.
-
ItemPredictors of women’s participation in crop farming and livestock rearing in Uganda(Makerere University, 2025)This study examined the predictors of women’s participation in crop farming and livestock rearing in Uganda. Specifically, it assessed the effects of land size and tenure systems, access to extension services, availability of financial and input resources, and selected socio-demographic characteristics on women’s engagement in the two agricultural subsectors. Given that women constitute a substantial share of Uganda’s agricultural workforce, identifying these predictors is critical for informing targeted and inclusive agricultural policies. The study employed descriptive statistics, correlation analysis, and logistic regression techniques using data from the 2019 Annual Agricultural Survey. The results indicate that 57.7 percent of women participated in crop farming, while 62.9 percent were involved in livestock rearing. In crop farming, land tenure emerged as a key determinant, with Mailo and customary tenure reducing participation, while access to public land and larger land sizes increased women’s engagement. Access to extension services and storage facilities positively influenced participation, whereas reliance on radio or farmer-to-farmer information channels and being divorced or separated were associated with lower participation. In livestock rearing, Mailo tenure, access to extension services, farmer training, agricultural credit, transport, and storage facilities significantly enhanced women’s involvement, while leasehold tenure, higher levels of education, and information obtained through radio or informal sources reduced participation. Notable regional disparities were observed, with women in the Northern region being more likely to engage in livestock rearing than their counterparts in the Central region. The study recommends strengthening women’s access to and control over land, expanding agricultural extension and training services, and improving access to credit, transport, and storage infrastructure to enhance women’s participation in both crop farming and livestock rearing. In addition, policy interventions should account for socio-demographic differences by supporting single or divorced women, addressing regional inequalities, and promoting entrepreneurship and market linkages to encourage greater participation among educated women.
-
ItemAn approach to enhance the performance of the Xgboost Classifier(Makerere University, 2025)XGBoost is a dominant machine learning model for prediction and classification tasks. The XGBoost algorithm is an ensemble that often outperforms other machine learning models due to its enhanced predictive performance, efficiency, and regularization technique that prevent overfitting and underfitting. However, its heavy reliance on hyperparameter tuning creates computational weaknesses due to the intensive resource requirements of traditional methods like grid and random search. Furthermore, the raw features used for classification tasks may contain complex, non-linear relationships, not explicitly captured by XGBoost’ s base leaners. This study proposed an improved alternative by combining k-means clustering with Bayesian-optimized XGBoost. To validate this approach, the study utilised the red wine dataset from the UCI data repository. We first derived objective quality clusters from physicochemical attributes (like acidity, sugar, alcohol content) using k-means. Thereafter, two hyperparameter tuning approaches were then compared: (1) traditional hyperparameters, (2) Bayesian optimization. This study demonstrates that combining k-means clustering with Bayesian-optimized XGBoost significantly improves model classification accuracy compared to the use of traditional hyperparameters. When evaluated, the cluster-based model with Bayesian optimization achieved a 97.9% accuracy, F1-score of 97.4% and recall of 98.05%. On the other hand, the baseline model achieved 93.1% accuracy, 96.18% F1-score and 97.2% recall. This study demonstrates that the integration of k-means clustering with Bayesian optimization significantly enhances the performance of the XGBoost classifier. Consequently, we recommend deploying this validated model in real-world applications, such as automated wine quality grading, as well as in other industrial domains that require scalable and accurate classification solutions.