School of Statistics and Planning (SSP) Collections

Permanent URI for this collection

http://10.29.1.53:4000/handle/10570/174

Browse

Browsing School of Statistics and Planning (SSP) Collections by Title

  • Item
    Abortion in Uganda : the neglected dangers
    ( 2020-05-22) Godfrey, Asiimwe
    Abortion can be defined as an ending of a pregnancy by removal or expulsion of an embryo or fetus before it can survive outside the uterus. An abortion that occurs without intervention is known as a miscarriage or spontaneous abortion while deliberate methods done to end a pregnancy is called an induced abortion. In this paper, emphasis was put on induced abortion where globally 56% of unintended pregnancies end up in induced abortion. (Singh S et al., 2016). As of 2010-2014, an estimated 55.9 million abortions occur each year with majority from developing regions at 49.3 million (with a rate of 36 per 1,000) while in developed regions it stands at 6.6 million (with a rate of 27 per 1,000) (Sedgh G et al., 2016). Abortion in Uganda is illegal however the Ugandan Ministry of Health's 2006 National Policy Guidelines and Service Standards for Sexual and Reproductive Health and Rights gives a number of specific cases in which women have the right to seek an abortion, that includes rape, sexual violence, or incest, or when the woman has pre-existing conditions such as HIV or cervical cancer. It can also be performed by a licensed medical doctor in a situation where the woman's life is deemed to be at risk. 52% of all pregnancies in Uganda are unintended, and about a quarter of these unintended pregnancies end in abortion each year. It is estimated that 314,300 abortions took place in 2013 alone translating to 14% or at a rate of 39 per 1,000 of all pregnancies of women aged 15-49 in the country according to a 2013 brief from Guttmacher Institute. Unsafe abortion practices in Uganda have consequently led to around 5 million Ugandan women annually getting injured or getting disabled due to abortion-related consequences and 5.3% of maternal deaths occurring as a result of abortion complications. Abortion has also led to increased emotional risks which have a lasting negative impact on a woman. It is recommended that a clarification on Uganda’s abortion law and policies be done at all levels in all languages since this will raise awareness on the contents and the scope of Uganda’s abortion law among the medical community, the judicial system and women.
  • Item
    Academic achievement of Ugandan sixth grade students: influence of parents’education levels
    (Contemporary Issues in Education Research, 2013) Wamala, Robert ; Omala, Saint Kizito ; Jjemba, Evans
    The study investigates the influence of a father and mother’s education on the academic achievement of their child. The investigation is based on data sourced from the 2009 Southern African Consortium for Monitoring Education Quality survey comprising 5,148 records of sixth grade students enrolled in Ugandan primary schools. Students’ percentage scores in the health sciences, reading, and numeracy tests were adopted as a measure of academic achievement. The analysis was carried out using summary statistics and a multiple linear regression clustered by six geographical regions in Uganda: central, eastern, western, northern, southwestern, and North Eastern. In addition to father and mothers’ education, students’ test scores in the various disciplines were analysed by the characteristics of age, sex, rural-urban residence, grade repetition status (any grade), and length of pre-primary education. The results showed that the level of a father’s education required to predict whether the child will achieve better scores in all disciplines was primary education. However, a mother required secondary and post-secondary education to enable the child to obtain better scores in reading and numeracy, respectively. Much of the previous literature has suggested that children born to educated parents have higher academic achievement; the results of this study support this finding but also reveal a difference in the levels of a father and mother’s education required to predict their child’s achievement of better scores in formal education.
  • Item
    Adherence to antiretroviral treatment of HIV positive patients in Lira District, Northern Uganda
    (Makerere University, 2017-01) Opio, Bosco
    The main objective of the study was to determine the factors that affect adherence to ART among AIDS patients in Lira district, northern Uganda. The cross sectional studies was carried out on people living with HIV and were receiving free ART services in Lira District at AMACH Health Center (IV), BARR Health Center (III), and OGUR Health Center (IV). Secondary data were collected by reviewing 208 ART files for patients above 11 years from these Health Centers respectively. Adherence was measured by self- report, and patients were identified as having poor (sub-optimal) adherence if they took less than 95% of their pills during either the previous two days or four days. Data were presented using proportions, frequency tables, percentages, standard deviations and histograms. Pearson chi square statistics was used to determine significant associations between the dependent and independent variables. Logistic regression was used to further identify the factors associated with poor adherence. The final logistic model identified the following factors significantly associated with : Forgetfulness (OR=0.06, p-value=0.000), lack of Food (p-value=0.008; OR=0.15) and alcohol (OR=0.104; p-value=0.022) were the patients factors affecting adherence, Stigma (OR=0.06; p-value=0.01) and lack of partner support (OR=0.06; p-value=0.000) were the community factors affecting adherence, Distance (OR=0.07, p-value=0.000) and Drug stock- out (OR=0.045, p-value =0.037) were the health facility factors affecting adherence; Forgetfulness was the most common reason for non-adherence. Therefore, the ART counselors need to give emphasis on memory aids. Interventions to improve and sustain adherence should focus within the clinic to improve on interactions between the clients and the ART in-charges and outside the ART clinic to improve on patients’ social support, drug availability, campaign against stigma and others may be required.
  • Item
    Airline delay time series differentials: Autoregressive integrated moving average model
    (IGI Global, 2014) Wesonga, Ronald ; Nabugoomu, Fabian ; Masimbi, Brian
    Flight delays affect passenger travel satisfaction and increase airline costs. The authors explore airline differences with a focus on their delays based on autoregressive integrated moving averages. Aviation daily data were used in the analysis and model development. Time series modelling for six airlines was done to predict delays as a function of airport's timeliness performance. Findings show differences in the time series prediction models by airline. Differential analysis in the time series prediction models for airline delay suggests variations in airline efficiencies though at the same airport. The differences could be attributed to different management styles in the countries where the airlines originate. Thus, to improve airport timeliness performance, the study recommends airline disaggregated studies to explore the dynamics attributable to determinants of airline unique characteristics.
  • Item
    Alcohol Attrition Rate and Associated Factors in Uganda: A Structural Equation Modelling Approach
    (Makerere University, 2019-11) Mpungu, Amos
    Alcohol and drug use in the Ugandan population has been cited among the major drivers behind the spread of HIV/AIDS; road accidents; domestic violence, to mention but a few. This study focused on assessing the rates of alcohol attrition and associated factors in Uganda clustering by region and residence. The assessment was done using socio-demographic, health, economic and behavioural factors. The study utilised secondary data sourced from the Uganda Non-communicable Diseases Risk Factor Survey consisting of 3987 participants selected countrywide and drawn using a three-stage sampling design. Structural equation modelling was used to identify both the direct and indirect determinants of alcohol attrition rates. Only 1890 out of the 3987 participants had consumed alcohol before and either stopped or reduced their consumption. The highest proportion (57.57%) of these 1890, had an alcohol attrition rate of less than a month. The rates of alcohol attrition were found to be significantly affected directly by sex and smoking status while indirectly affected by sex, age, residence, region and work status. The study recommends that initiatives and programmes geared towards addressing alcohol consumption and abuse should collectively involve addressing other forms of drug abuse especially smoking since it contributes directly to the rate of attrition. These should be also be tailored based on sex with more emphasis given to males since women tend to have higher rates of attrition compared to males. Additionally, different measures should be developed based on target age groups; rural/urban residences and the calibre of work the target group is engaged in.
  • Item
    Alcohol use among adults in Uganda: Findings from the countrywide non-communicable diseases risk factor cross-sectional survey
    (Taylor & Francis, 2016) Kabwama, Steven Ndugwa ; Ndyanabangi, Sheila ; Mutungi, Gerald ; Wesonga, Ronald ; Bahendeka, Silver K. ; Guwatudde, David
    Background: There are limited data on levels of alcohol use in most sub-Saharan African countries. Objective: We analyzed data from Uganda’s non-communicable diseases risk factor survey conducted in 2014, to identify alcohol use prevalence and associated factors. Design: The survey used the World Health Organization STEPS tool to collect data, including the history of alcohol use. Alcohol users were categorized into low-, medium-, and high-end users. Participants were also classified as having an alcohol-use-related disorder if, over the past 12 months, they were unable to stop drinking alcohol once they had started drinking, and/or failed to do what was normally expected of them because of drinking alcohol, and/or needed an alcoholic drink first in the morning to get going after a heavy drinking session the night before. Weighted logistic regression analysis was used to identify factors associated with medium- to high-end alcohol use. Results: Of the 3,956 participants, 1,062 (26.8%) were current alcohol users, including 314 (7.9%) low-end, 246 (6.2%) medium-end, and 502 (12.7%) high-end users. A total of 386 (9.8%) were classified as having an alcohol-use-related disorder. Male participants were more likely to be medium- to high-end alcohol users compared to females; adjusted odds ratio (AOR)=2.34 [95% confidence interval (CI)=1.88–2.91]. Compared to residents in eastern Uganda, participants in central and western Uganda were more likely to be medium- to high-end users; AOR=1.47 (95% CI=1.01–2.12) and AOR=1.89 (95% CI=1.31–2.72), respectively. Participants aged 30–49 years and those aged 50–69 years were more likely to be medium- to high-end alcohol users, compared to those aged 18–29 years, AOR=1.49 (95% CI=1.16–1.91) and AOR=2.08 (95% CI=1.52–2.84), respectively. Conclusions: The level of alcohol use among adults in Uganda is high, and 9.8% of the adult population has an alcohol-use-related disorder.
  • Item
    Analysis of access to health insurance coverage and associated factors among women of reproductive age 15-49 years
    (Makerere University, 2024-12) Keinerugaba, Nicholas
    The objective of the study was to analyze the association between socio-demographic, media and economic factors affecting the access to health insurance coverage among women of reproductive age 15-49 years. The study is based on secondary data from Individual Recode dataset which is an extract from the national representative demographic and health survey (UDHS 2016). A total of 18506 respondents were analyzed. Sample survey weighted results were used, Scot-Rao Adjusted Chi-square tests were used at bivariate analysis and complementary log-log regression model was used to examine the association of various study variables on access to health insurance. The study findings revealed that out of 18,506 respondents, 23.5% were aware of health insurance, of these 18.8% ever considered joining a health insurance scheme and about 1.4% of the respondents reported to be on a health insurance scheme. Study findings revealed that respondents’ level of education, religious affiliation and place of residence and access to information through media related factors significantly influenced access to health insurance at bivariate analysis. Further findings revealed that age, region and religion significantly influenced access to health insurance at multivariate level. That is those aged 26- 40 years had twice the odds of being covered by health insurance as compared people below 25 years (AOR= 2.167, CI=(1.520,3.089)). Similarly, those aged 40 years and above were more than two times the odds of being on health insurance (AOR=2.522, CI=(1.499,4.239)). Those coming from the western region (AOR=2.277, CI=(1.245,4.161)), being Anglican (AOR=1.816, CI=(1.191,2.766)) and being Roman Catholic (AOR=1.986, CI=(1.327,2.970)) had twice the odds of being covered by health insurance. Conclusions from the study showed that older persons had higher chances of joining health insurance and having a health insurance scheme. Further the studies concluded the role of religion in promoting access to health insurance. Finally, the study recommends that insurance companies and Ministry of Health can target older persons aged above 26 years, wealthy individuals and places of worship to ensure mass awareness of health insurance and access to health insurance.
  • Item
    Analysis of attrition of business establishments in Uganda panel surveys
    ( 2019-07-29) Muhanguzi, Hillary
    This study set-out to analyze attrition in economic panel surveys with specific emphasis on the factors associated with attrition in the Producer Price Index(PPI) Survey in Uganda. This was done using the PPI and Uganda Business Inquiry (2002, and 2008) datasets with key variables being time to attrition and status of attrition as failure variable. The dataset comprised 262 records of characteristics of business establishments, their management and time to attrition. Survival analysis techniques for interval censored data in R Software were adopted. The data were treated as right censored with the final time of observation as June 2015. The analysis was made for these ‘mixed case’ interval-censored survival data using nonparametric estimation of the survival function or Turnbull’s algorithm and finally a Cox proportional hazards model was fitted. It was established that the median survival time for a business establishment in the survey was 22 quarters (range:21-22 quarters)from the time of initiation. Attrition varied significantly by industry sector, location of the establishment and time of entry into the PPI survey (p<0.05). Particularly the business establishments in Greater Kampala had a 44% lower risk of attrition than those located up-country (HR= 0.56, p-value= 0.03), those that joined the PPI survey after 2004 had a 10%lower risk of attrition than those that were initiated in 2004 (HR= 0.90, p-value= 0.006), and those in the manufacturing sector have a 65% lower risk of attrition than those in the construction sector (HR=0.35, p-value= 0.006). However, the total number of staff employed and citizenship of managers of the business establishment were found insignificant (p>0.05). The main findings of this research effort suggest that the reasons for attrition of business establishments in the PPI survey lie within their own control. However, it is recommended that the Uganda Bureau of Statistics should rebase the producer price indices at least every five years, strengthen data collection in up-country business establishments and undertake in-depth qualitative investigation of the determinants of attrition.
  • Item
    Analysis of dichotomous outcomes: a case study of constituency characteristics’ data in the 2006 general elections in Uganda
    ( 2012-12) Ochama, A. Ahmed
    Using election data on the 2006 Presidential poll following re-introduction of multi-party system and various surveys’ data compiled around 2006, this study explored characteristics that could have guaranteed election victory to the incumbent. The study considered scores for Constituency ratings – derived or otherwise - on income and poverty levels, participation, incumbency, literacy levels, party dominance or allegiance of the Constituency to a party, effect of regional location and the nature of the Constituency. The study population comprised 214 Parliamentary Constituencies covering the whole of Uganda. A systematic approach utilising Hosmer-Lemeshow logistic regression technique for analysis of dichotomous outcomes was used. On analyzing the variables, all other quantitative predictors, except participation of the Constituency voters, average monthly per Capita incomes of voters and dominance by a political party, were eliminated. These were found to be statistically reliable for the restricted model in correctly predicting 70.56% of the data. The policy implications for multiparty systems suggest that political and coalition actors could optimise their decisions or vote chances by articulating alternative policy matters in their campaign strategies. Unobservable income and other forms of inducements or coercion on which data are seldom collected or lacking may influence polls’ outcomes. Future studies could focus on the influence of these factors on election results. Technically, it remains incumbent upon the Electoral Commission and government to be held accountable for the entire electoral environment during the poll’s cycle.
  • Item
    Analysis of epidemiological and clinical characteristics of Acute Flaccid Paralyses (AFP) cases in Uganda (1997 – 2004).
    (Makerere University, 2013-05) Beinomugisha, Geofrey
    This study was intended to analyze the epidemiological and clinical characteristics of AFP cases reported in Uganda for the period 1997 to 2004; and to determine whether the establishment of National Polio Expert Committee (NPEC) made a statistical difference on the epidemiological and clinical characteristics of Acute Flaccid Paralyses (AFP) cases reported in Uganda. The study was also intended to establish the incidence of AFP cases for the period 2001 to 2004 under the current surveillance system. The data used was obtained from the Uganda National Expanded Program on Immunization (UNEPI) with consent from the Uganda Virus Research Institute (UVRI), Entebbe. Using frequency tables, the distribution of AFP cases according to their epidemiological and clinical characteristics was established with respect to NPEC establishment. Selected surveillance indicators were presented in form of averages according to NPEC. These were the time laps for different time periods in AFP surveillance. These included time (lapse1) between date of onset of paralysis and date of investigation, time (lapse2) between the two dates of stool collection, time (lapse3) between date of notification and date of follow–up to examination and time (lapse4) between date of receiving specimens at national level after date of collection. The study further investigated whether associations existed between selected covariates and fitted a model to determine asymmetrical paralysis. A total of 1,229 AFP cases were reported from 1997 to 2004, in which 39% (474) of the cases were reported before the NPEC establishment and 61% (755) of the cases were reported thereafter. Average time lags for various steps in the process of AFP surveillance and surveillance systems generally improved at each step. For instance, the time (lapse1) between date of onset of paralysis and date of investigation significantly decreased from 21.6 days to just 9.5 days after NPEC establishment and the time (lapse3) between date of notification and date of follow–up to examination significantly decreased from 113.2 days to 78.9 days after NPEC establishment. AFP cases with asymmetrical paralysis slightly decreased by 7.2% under the current surveillance system (NPEC establishment). More still, cases of under dose OPV increased by 4.6%, while cases of over dose OPV decreased by 8.5%. Further, the proportion of cases in which paralysis progressing in three days of onset increased by 18.8%, while cases of residual paralysis at follow-up increased by 1.9%. Poliovirus strains 1, 2, and 3 were respectively less than 4% before or after NPEC establishment; poliovirus strain 3 had the highest percentage of 3.8% before NPEC establishment. Vaccine-derived strains 1, 2 and 3 respectively were between 0.5% and 3.0% in either era of NPEC. There were no wild poliovirus strains isolated between 1997 and 2004. Non-polio enteroviruses were found to be less than 20% (17.9% before and 18.9% after NPEC establishment) and that conformed to the studies conducted in Australia that showed that non-polio enteroviruses ranged from 15-45% of the AFP cases. Asymmetrical paralysis was found to have a significant association with region and possession of fever after NPEC establishment. Paralysis progression in three days was also found to have a significant association with possession of fever during NPEC era. Presence of any poliovirus type was found to have a significant association with age group and OPV dose administered. Presence of any vaccine virus type was found to have no significant association with any of the selected variables. Presence of non-polio enteroviruses was found to have a significant association with region and age group, while site of paralysis was found to have a significant association with region, age group, possession of fever and OPV dose. There is a great need to improve on vaccination culture by routinely sensitizing childcare takers and mothers about the importance of completing OPV dose to their children. AFP follow-ups should be launched immediately after 60 days of onset of paralysis as required by the standards in AFP surveillance and reporting. Avoid using the staff already at the facility; these already have an overload of patients with other illnesses and would not be effective in AFP surveillance and reporting.
  • Item
    An analysis of factors determining voter turnout : a case study of Kampala Central Division
    (Makerere University, 2009) Namata, Ssimbwa Gladys
    The determinants of voter turnout in Uganda have not been researched on like it has been done in other countries like South Africa, USA and others; yet voting is a crucial exercise in every country. For instance, since South Africa's society is segmented on ethnic lines, ethnicity therefore plays a crucial role in affecting voting patterns and political choices, followed by socio-economic status and age of the voter. In USA, American citizens with more education and income, employed individuals, home owners, those who have stayed longer at current residence, professionals, women, older individuals, married individuals and blacks were more likely to register and vote. In Uganda, there was need to research on the determinants of voter turnout and the magnitude of their contributions. The major aim of the study was to analyse factors determining voter turnout. The data used in the research was mainly primary data, collected by use of questionnaires in Kampala Central Division. However, some secondary data was obtained; like the list of registered voters which was used as the sampling frame. This was provided by the Electoral Commission of Uganda, a body in charge of elections. Systematic sampling procedure was applied at all stages in sample selection. The data was analyzed using logistic regression model which showed the length of time at current residence as the most significant predictor, followed by education, then age, eligibility of household members to vote and lastly not knowing whom to vote. The findings were that; for each additional year a person resides in a particular area, the odds in favour of voting increases by 4.6%, the less educated vote more than the higher educated, the odds in favor of a person to vote increases by 6.2% each additional year of age, the odds of not voting for persons from households where some members are below the voting age are one and a half times more than for those where all members are eligible voters and many voters not know who to vote. The other variables, which are; sex, marital status, region of origin, economic status, lack of interest, elections rigging and other forms of malpractices and being very busy or process is time wasting were not significant. It was therefore concluded that age, length of time at current residence, education, eligibility of household members to vote and not knowing whom to vote affect voting. There is need therefore, for the Electoral Commission to carry out more elaborate sensitization programmes about voting targeting all registered and unregistered voters.
  • Item
    Analysis of financial sector national accounts: Emphasis on financial intermediation (1995-2009)
    (Makerere University, 2011-06) Echoku, Samuel Okorom
    The study attempts to obtain good estimated measure of the output of financial services activities in the economy by first computing the gross output and value added of the financial intermediation services following international recommendations in the System of National Accounts 1993 (SNA 93). In this study the trends of the output and value added have also been computed. The period of coverage was 1995-1 to 2009-4. A couple of methods were applied to first of all compute the dependent variable (Value Added) and then to analyse the data sets and produce the forecasting equation. The current prices gross output was deflated to obtain constant prices gross output. Constant prices gross output was then multiplied by a constant base period input/output ration to obtain the constant prices value added (the dependent variable). In order to determine the characteristics of each variable, stationarity tests including; graphical, Auto Correlation Function (ACF) and correlogram test, and Augmented Dickey-Fuller Unitroot Test were done. Differencing and lagging methods were applied to the non stationary variables wherever necessary to eliminate none compliance problems from the variables. The correlogram was used to determine the existence of multicollinearity and four variables (Bank Deposits, M1 Currency in circulation, M2 money supply and the loan stocks) were found to be correlated. The study sought to determine the factors contributing to the growth of the financial intermediation activities/services and the financial sector in totality. In this study emphasis was on financial intermediation because it contributes to well over eighty percent of the output of the financial activities. In the reduced model three variables were found to be significant in the projection of the Financial Intermediation value added. The variables include; bank deposits, interest rates and bank rate and they all positively contribute to the forecast values. Forecasts for the 2010 and 2011 quarters, value added was done based on the empirical model and projections of the independent variables. There were no significant relationships between the loan stocks, money supply M2, currency in circulation M3 and price index to the trends of value added. Also inflation rates registered during the study period did not have a significant impact on the financial intermediation services performance. On the basis of the findings of the study, the last section gives recommendations for the policy makers to take on if the growth of financial activities are to be stimulated and encouraged. The recommendations are based on the three variables namely; Bank deposits, bank rate and interest rates that were found to be significant in determining the performance of the financial intermediation services value added. The three variables were found to positively influenced trends of the value added. The author recommends observance of stringent monetary and fiscal policies that will encourage savings to increase on the bank deposits, and monetary policies that will lead to high but not too high bank rates and therefore interest rates.
  • Item
    An analysis of pooled data for monitoring primary school pupils’ proficiency
    ( 2009-11-12) Opaman, Amos N. C.
    In a bid to improve on the quality and relevance of education, Uganda government commissioned the Education Policy Review Commission (EPRC) in 1987 to “appraise the entire system of education and recommend measures and strategies to address inadequacies hitherto identified.” The EPRC submitted its report in 1989 which was followed by the Government White Paper on Education in 1992. The Education Policy Review Commission (1989) report and the Government White Paper (1992) identified flaws in the educational assessment system, especially with regards to monitoring progress in education. The overall objective of this study was to analyze pooled data with a view to monitoring primary school pupils’ proficiency for quality assurance in order to aid assessment and monitoring of basic education, particularly in Uganda. Principal Component analysis was carried out to reduce on the dimensionality of the data set. Five principal components (uncorrelated variables) were extracted and then subjected to linear regression and logistic regression analyses. The study found out that in general, private schools were 3.0 times better positioned to improving the proficiencies of primary pupils compared to government schools. Overall, however, government-aided schools were noted to be generally good in raising the proficiency of primary school pupils. The availability of safe water in schools is a basic requirement for the development of primary pupils’ proficiency. This is 3.9 times more likely to enhance the proficiency of primary pupils in the Adequate Category proficiency level than those in the Basic Category proficiency level compared with schools without safe water sources. The study recommends that the learning environment (availability of scholastic materials, trained and motivated teachers and motivated learners) in rural schools should be enhanced, since urban schools were 0.79 times more likely to better the proficiency of pupils, basing on these factors. In conclusion, pupils’ proficiencies are highly affected by the surrounding environment of study and need to be addressed if better achievement is to be enhanced.
  • Item
    Analysis of repeated measurements of HIV viral load as a continuous variable while accounting for missing values
    (Makerere University, 2023-12) Sendege, Susan Habert
    Recent HIV research has predominantly employed Single Measure Frameworks (SMF), relying on the latest viral load data while ignoring missing values, despite criticisms for information loss and neglecting correlations. This study aimed at assessing repeated measures framework and single measures framework by examining factors affecting viral load reduction while accounting for missing values. The analysis was conducted on data from Mukono health facilities providing HIV treatment and care. A total of 1670 records of ART patients, was used in the analysis and generalized linear (mixed) models (GLMM) was adopted at the modeling stage. All variables included in this analysis were recorded upon patients’ enrollment on treatment, with the exception of change in the treatment regimen. A GLMM was applied to the data both before and after imputation under the RMF framework. The best-fitting model, selected for evaluating factors influencing viral load copies, was the GLMM fitted to multiply imputed data. The analysis indicated that gender and adherence rating did not have a significant effect on viral load copies. Additionally, patient age, marital status, duration on treatment, WHO clinical stages, and ownership of the facility were included in the analysis. The results showed that marital status, duration on treatment, and the type of health facility had a significant effect on viral load copies. Specifically, viral load copies were higher among those who were currently or formerly married (𝛽 = 0 49 0 30; 𝑆𝐸 = 0 042 0 052; 𝑝 = 0 0000). However, viral load copies were found to be lower among patients who had a longer duration on treatment (𝛽 = 􀀀0 01; 𝑆𝐸 = 0 001; 𝑝 = 0 0000) and were receiving treatment at a private facility (𝛽 = 􀀀0 196; 𝑆𝐸 = 0 077; 𝑝 = 0 0000). The study highlights the significance of recognizing repeated data patterns in longitudinal settings and addressing missing values in health research. It proposes a similar investigation in controlled environments to evaluate SMF and RMF in presence of missing values.
  • Item
    Analysis of the factors that influence outcome of tuberculosis treatment using first line drugs in Uganda - Case study: Mbarara University Teaching Hospital and Mulago Hospital
    ( 2008-11) Nansamba, Bitiyali
    The increase in lack of adherence to tuberculosis treatment has been attributed to patients’ failure to complete treatment whereby they expose themselves to the risk of reactivation of the disease and on the other hand health services have failed to evaluate treatment outcome because they lose contact with the patients before completion of treatment. This study aimed at analyzing the factors that influence treatment outcome after 8 months of short course chemotherapy. Data were summarized using frequency tables, the chi-square test was used to test for independence and binary regression was used to determine the factors that influence treatment outcome after 8 months of treatment. The study involved patients who were treated at Mulago and Mbarara University teaching hospitals under the Directly Observed Treatment short course (DOTS) and followed up for a period of 8 months for the year 2003. Data was obtained from hospital records of patients which included variables like age, sex, category of patient, treatment regimen given, category of disease and outcome of treatment (this was noted as cured, not cured, defaulted, referred and died) at the end of the period. The study involved 1437 patients who were treated for over 8 months with initial therapy of 2 months and a continuation phase of 6 months using different treatment regimens. It comprised of 625 females and 812 males of all ages. Of the patients involved, 69.03% tested positive for pulmonary tuberculosis and 30.97% had extra pulmonary tuberculosis. 29.44% of the patients were cured after 8 months of treatment, 20.11% were not cured though they completed treatment, 47.74% defaulted, 0.35% died and 2.37% transferred. Patients of all ages were treated with the youngest patient being one year old and the oldest was 86 years old and on average, patients were 33 years old. The study established that the only factors that influence treatment outcome were, pretreatment status, hospital visited and treatment regimen used in the first 2 months as observed from the complementary log log model. The age of the patient, treatment regimen used for the continuation phase of 6 months and patient category did not influence outcome. Patients that tested positive for pulmonary tuberculosis had a higher chance of getting cured at the end of the 8 months of treatment. From this study, the following recommendations were made, proper record management to assist in the follow up of patients, sensitization of the public concerning the disease and decentralization of health services to reduce on the transport costs.
  • Item
    Analysis of Viability of Core Inflation Components Based on Excluded Items in Uganda
    (Makerere University, 2018-12-19) Niyimpa, Edgar
    The purpose of the study was to assess the viability of core inflation components based on traditionally excluded items of the consumer price index in Uganda. Particularly, this was achieved through subjecting all Energy, Fuel and Utilities (EFU) and Food crops to volatility and persistence tests. The study used time series data obtained from the Uganda Bureau of Statistics (UBOS) for the period July 2010 to June 2017 and covered all the 10 baskets monitored by UBOS. The measurement variables included volatility and persistence. The standard deviation was employed to measure the volatility and the sum of auto-regressive coefficients (AR ) in a univariate inflation equation was used to measure persistence. In the study, Food and related items were noted to be more volatile than EFU items. Particularly tobacco leaves, peas, matooke and malewa recorded the highest volatility with a low persistence while cassava fresh, water melons, apples, ground nuts and french beans recorded the least price volatility during the study period. On the other hand, the EFU items namely petrol, diesel, paraffin, NWSC water, firewood and electricity recorded the least variation in prices with a high persistence during the study period. The study further indicated that inclusion of the EFU and food items found to be less volatile and highly persistent, did not yield significantly different results when comparing means of the recompiled and existing core index at 5% level of significance (p>0.05). However, the recompiled core index was more precise than the existing core index because the standard error reduced by 7%. In light of the findings, this study recommends periodical review of the core components based on both volatility and persistence in order to improve estimates.
  • Item
    Analyzing industrial electricity demand in Uganda from 1998 to 2022
    (Makerere University, 2024-08) Mukasa, Muhammed
    This study was conducted for the purpose of analyzing industrial electricity demand in Uganda from the year 1998 to 2022. To achieve this, descriptive statistical and econometric methodologies were utilized to analyze the influence of electricity price, industrial income (GDP) and Index of Industrial Production (IIP) on electricity demand. Using the Augmented Dickey-Fuller test, it was found out that variables are first difference stationary which is the necessary condition for a cointegrated system. A co integration test using the Johansen-Juselius method was used to determine if there is a long run relationship among electricity consumption, its price, index of production and the GDP of the industrial sector. The Error Correction Model was then applied to determine the short run elasticities and the Error Correction Term was interpreted to determine the speed at which electricity demand returns to equilibrium after a change in GDP, IIP and electricity price. The study revealed that in the short run, electricity price has no effect on quantity of electricity demanded. However, in the long run, electricity demand is price sensitive with a price elasticity of -0.14 in large industries. GDP has a positive influence on electricity demanded both in the short and long run. In the long run, the income elasticity for large and medium industry is 0.78 and 0.32 respectively and it is found to be 0.032 for the large industries in the short run. This makes electricity a normal good as well as a necessity for the industrial sector. IIP has a positive (0.06) and negative (-0.09) influence on electricity demand in large and medium industries respectively. This implies that large industries are more energy efficient as compared to medium industries. In both medium industry and large industry, there was convergence towards long run equilibrium at a rate of 0.4% and 0.11% respectively. Thus, there is a long run relationship among electricity consumption and electricity price, industrial income and index of industrial production. This study has revealed that electricity is a necessity and is thus a key input factor for the industrial sector in Uganda. Therefore, government regulatory bodies should take action by controlling the price of electricity and ensuring a steady supply of this indispensable input in Uganda’s industries. Also, it is recommended that new industries should consider purchasing and installing equipment with improved technologies to ensure efficient use of electricity in their production processes.
  • Item
    An application of Data Mining Classification Techniques to Electricity Fraud Detection; A Case of Umeme (U) Ltd
    ( 2021-11) Okuda, Boniface
    The main objective of this research was to develop a suitable classification model that will be able to identify and predict customers with fraudulent consumption. This research addresses the lack of effective methods for fraud detection in electricity consumption by focusing on four different data mining classification techniques; K-Nearest Neighbors, Logistics Regression, Gaussian Naïve Bayes Classifier and Linear Discriminant analysis. These techniques were applied to classify consumption data from and electricity Distribution Company into fraud and non-fraud consumption. The data used consisted of; customers‟ electricity consumptions for a period 45 months from January 2014 to Sept 2017 and the customer's electricity irregularities data (anomaly case) for the same period based on a random inspection undertaken by UMEME Ltd. The results gained in this thesis indicated that the linear Discriminant analysis classifier is more suited for anomaly detection using machine learning techniques, gaining an accuracy of 96%. The K-Nearest Neighbor classifier obtained an accuracy of 93%, Logistic Regression classifier obtained an accuracy of 79%, and finally the Gaussian Naïve Bayes classifier produced an accuracy of 27% and it was the model with the least performance. The weak performance of Naïve Bayes algorithm is because it tends to treat all the input dimensions as independent from each other which is not always true in a real world setting. In other words, if there is an existence of covariance between two or more input dimension, the Gaussian Naïve Bayes classifier does not model it. This research recommends that UMEME Ltd explores the possibility of using the Linear Discriminant Analysis classifier in detecting electricity fraud. Furthermore, the selected techniques work as a black box, without induced descriptive rules to show how the attributes indicate fraudulent behavior, so this research also recommends that rule based model could be explored. In addition, detection algorithms can be enhanced by introducing more real-world parameters and variables. Real world parameters and variables can be the economic situation of a person or illegal activity in the past.
  • Item
    Assessing differences in secondary school attendance before and after the introduction of universal secondary education in Uganda
    (Makerere University, 2018-08) Ssessanga, Adrian
    The objective of the study was to investigate differences in secondary school attendance before and after the introduction of Universal Secondary Education (USE) in Uganda. The study focused on children of the secondary school going ages of 13 to 18 years as spelt out by the Ministry of Education and Sports. Having introduced the USE program in 2007, the study made this investigation using data of the 2002 Uganda Population and Housing Census (UPHC) conducted five years before the introduction of USE as well as data of the 2014 UPHC seven years after the introduction of USE. The 2002 and 2014 UPHC comprised of 347,701 and 513,045 children of secondary school going aged 13 to 18 respectively. Differentials in secondary school attendance were investigated by the child’s socio-demographic characteristics, household level and external environment factors using a nonlinear multivariate decomposition of the logistic regression. The differences in secondary school attendance were decomposed into a component attributable to variations in the characteristics of children and a component attributable to variations in the effects of predictors before and after the introduction of USE. In the results, the secondary school attendance in 2002 (25.8%; CI: 25.6-25.9) and 2014 (34.9%; CI: 34.7-35.0) demonstrated a significant gap in school attendance of 9.1%. Variations in effects of predictors between 2002 and 2014 contributed the highest percentage (81.8%) to the secondary school attendance gap (p<0.05). In the detailed decomposition, differences in the effects of predictors were noted in variables namely age, disability status, religion, parental survival, region, residence and source of household livelihood. Differences in characteristics were also noted in the aforementioned variables including sex (p<0.05). Although findings show that the orphaned, rural-based and disabled child was still less likely to attend secondary school after the introduction of USE, the period after USE should have had a notably higher likelihood of a less privileged child attending secondary school education which has been presented by literature as expensive and for a chosen few. Government should therefore conduct routine monitoring and evaluation of its programs to ensure that services are diligently delivered such that for the case of USE in secondary education, the targeted beneficiary children attend and do not simply purport to attend.
  • Item
    Assessing differentials in modern contraceptive use among women in Uganda between 2000/1 and 2011: An application of the logit-based decomposition
    (Makerere University, 2017-01) Bwambale, Ezra
    The objective of the study was to assess differentials in modern contraceptive use among sexually active women (15-49) in the period 2000/1-2011. The gap in modern contraceptive use was decomposed into components attributed to differences in characteristics of women and differences in effects of the predictors in two phases namely 2000/1-2006 and 2006-2011. The assessment was based on samples derived from 2000/1, 2006 and 2011 Uganda Demographic and Health Survey data consisting of 4905, 5513 and 5633 records, respectively. Analysis was done using frequency distributions and nonlinear Oaxaca-blinder decomposition of the logistic regression. In the results, rates in modern contraceptive use increased from 23.3% in 2000/1 to 31.0% in 2011. Overall, differentials in modern contraceptive use were attributed to the compositional variations in the characteristics of women and variations in effect of the predictors between 2006 and 2011 (p<0.01). Compositional variations of women accounted for 34.0% of the overall gap in modern contraceptive use among women between 2006 and 2011. The compositional variations in the study period were significantly noted in the variables namely type of residence, education level, number of living children and exposure to media FP messages (p<0.05). The variables contributed 3.6%, 17.1%, -0.7% and 9.2% to the gap in contraceptive use between 2006 and 2011, respectively. In relation to effects, 66.0% of the overall gap was accounted for by variations in the effects of the predictors. Variations in effect of wealth quintile contributed 18.0% to the gap in modern contraceptive use (p<0.05).Worth noting is that the overall gap in modern contraceptive use among women was not significant between 2000/1 and 2006 (p>0.05). In light of the findings, future efforts for enhancing utilization of modern contraceptive methods in the country should focus on promoting universal education, publicity of family planning programs and improved child health care. However, additional research is needed to assess the contribution of supply side factors that would have been also important to this study.