-
Derivation of Two Associate PBIBD from Necessary Properties of BIBD
Troon John Benedict,
Onyango Fredrick,
Karanjah Anthony
Issue:
Volume 12, Issue 4, July 2023
Pages:
66-71
Received:
23 April 2023
Accepted:
13 May 2023
Published:
6 July 2023
Abstract: One of the main shortfall of Balanced Incomplete Block Designs is that the design is not available for all parameter sets. Thus, some of the parameter sets that satisfy the necessary conditions for BIBD cannot be constructed as such. This means that the parameter sets for the designs is not possible to be constructed with a single association scheme (λ). Given that the structural difference that exist between BIBD and PBIBD is that the later do not have a single association scheme but rather m association schemes, the study believed that if the BIBD parameter sets could not be constructed as a single associations scheme then maybe if the design is broken down to more than one association scheme in form of a PBIBD then maybe such design might later be constructed using other methods as PBIBDs. The study aimed at determining whether two association scheme PBIBDs could be derived from necessary properties of BIBDs. The main reason for maiking the transformation is that for BIBD some of the designs that satisfy the necessary properties have been determined not to exist meaning that the designs could not be constructed using a single association scheme λ. Therefore, the study felt that if the designs could be broken down into 2 association scheme (λ1 and λ2) then such a design might be constructed as a PBIBD. The study related the necessary properties of BIBD and two association scheme PBIBD. Using the properties, the study created eight sets of linear equations and using the Gauss Jordan Elimination method, the study was able to solve for the eight unknown parameters of the PBIBD association scheme. The study was able in the end to convert a BIBD that satisfy necessary properties of BIBD into a two association scheme PBIBD that satisfy all the necessary properties of PBIBD.
Abstract: One of the main shortfall of Balanced Incomplete Block Designs is that the design is not available for all parameter sets. Thus, some of the parameter sets that satisfy the necessary conditions for BIBD cannot be constructed as such. This means that the parameter sets for the designs is not possible to be constructed with a single association schem...
Show More
-
Multi-Level Analysis of Risk Factors for Intimate Partner Violence Against Males
Douglas Andabati Candia,
Patrick Guge Oloo Weke,
Moses Mwangi Manene
Issue:
Volume 12, Issue 4, July 2023
Pages:
72-81
Received:
10 June 2023
Accepted:
30 June 2023
Published:
11 July 2023
Abstract: Globally, there is a rise in reporting of intimate partner violence against men though limited attention has been directed toward addressing the practice problem. Therefore, this study aimed to identify the risk factors associated with intimate partner violence against males. This study used data from Demographic and Health Surveys. We compared the two-level mixed-effects and one-level regression models with logit, probit, and clog log link functions. The two-level mixed-effects regression model fitted the data best. From the study sample, 44.2 percent of males had experienced IPV. The factors that increased the likelihood of experiencing IPV included belonging to the Catholic or Pentecostal religious denominations; being divorced or separated; fathering children with multiple partners and one’s partner exhibiting jealousy and other controlling behaviors. Males can also experience IPV with differences across regions of Uganda, hence a need for policies and interventions tailored specifically to the country's different regions. Additionally, there is a need to engage religious institutions and other stakeholders in sensitizing people on issues relating to IPV and multiple-partner fertility. The results showed that the multilevel models reported the lowest AIC values and fitted the data better than the ordinary regression models. Users of DHS datasets need to consider using multilevel models since the data is hierarchical in nature with respondents nested with geographical locations such as residence (rural/urban), districts, regions, etc., and the samples are obtained using multistage sampling which involves clustering of respondents.
Abstract: Globally, there is a rise in reporting of intimate partner violence against men though limited attention has been directed toward addressing the practice problem. Therefore, this study aimed to identify the risk factors associated with intimate partner violence against males. This study used data from Demographic and Health Surveys. We compared the...
Show More
-
Statistical Analysis of the Risk Factors Associated with Visceral Leishmaniasis Patients at Marsabit County Referral Hospital
Elias Elema Guyo,
Tum Isaac Kipkosgei,
Mathew Kosgei
Issue:
Volume 12, Issue 4, July 2023
Pages:
82-86
Received:
19 June 2023
Accepted:
6 July 2023
Published:
17 July 2023
Abstract: Visceral Leishmaniasis also known as Kala-azar, is a tropical infectious disease caused by female sandflies. It affects the internal organs, usually the spleen, liver and bone marrow. Globally, an estimated 700 000 to 1 million new cases of visceral leishmaniasis occur annually. In Kenya, 4000 cases occur while 5 million people are at risk of infection. The purpose of the study was to evaluate risk factors for visceral Leishmaniasis. The study adopted a retrospective cohort design. The study used secondary data from 2890 visceral leishmaniasis patients enrolled at Marsabit County referral hospital from September 2015 to September 2019. Cox proportional hazard model was used to establish the relationship between the survival time of visceral leishmaniasis patients and predictor variables. Data analysis was carried out using R statistical software. The risk factors which were significant predictors for survival time of visceral Leishmaniasis patients included; household design (cracked walls and thatched roof) [β =.435, p=.0001], living near anthills [β =.320, p=.0012], using bed nets [β= -.151, p=.0080], contact with infected dogs [β =.200, p=.0006], forest surroundings [β=.151, p=.0340] and sleeping outside at night [β =.169, p=.0260]. In conclusion, there was an increased case of visceral Leishmaniasis among patients who are not using bed nets, those living in cracked mud walls, those living near the forest, residing near ant hills, sleeping outside, and those in contact with infected dogs. The study recommended adopting appropriate practices such as avoiding contact with infected dogs, using bed nets at night, clearing forests surrounding homesteads, avoiding sleeping in the open at night, and reducing house proximity to ant hills and termite mounds to reduce the transmission from Visceral Leishmaniasis.
Abstract: Visceral Leishmaniasis also known as Kala-azar, is a tropical infectious disease caused by female sandflies. It affects the internal organs, usually the spleen, liver and bone marrow. Globally, an estimated 700 000 to 1 million new cases of visceral leishmaniasis occur annually. In Kenya, 4000 cases occur while 5 million people are at risk of infec...
Show More
-
A Comparative Study of Methods of Remedying Multicolinearity
Issue:
Volume 12, Issue 4, July 2023
Pages:
87-91
Received:
13 July 2023
Accepted:
3 August 2023
Published:
28 August 2023
Abstract: This study is aimed at investigate the impact of multicollinearity on a model's predictive accuracy and assess the effectiveness of different techniques in handling multicollinearity. The purpose of this study is to compare several methods of addressing multicollinearity in regression analysis and to determine their effectiveness in improving the accuracy and reliability of the results. The specific methods to be compared include OLS regression, Two-stage regression Ridge regression and Lasso regression. The study simulated six predictor variables with high levels of multicollinearity and compared the performance of four regression models: Ordinary Least Square (OLS), Two-Stage Least Squares (Two-Stage), Ridge regression, and Lasso regression. The models were evaluated using metrics such as the Variance Inflation Factor (VIF), root mean squared error (RMSE), Akaike information criterion (AIC), Bayesian information criterion (BIC), and adjusted R-squared. The results showed that Ridge and Lasso regression models were more effective in handling multicollinearity than OLS and Two-Stage regression models. Ridge regression had the lowest RMSE and best predictive performance among the models, and Ridge and Lasso regression had better model fit and were more effective in handling overfitting than OLS and Two-Stage regression models. The study concludes that using Ridge and Lasso regression models can improve a model's predictive accuracy and reduce the impact of multicollinearity on the model.
Abstract: This study is aimed at investigate the impact of multicollinearity on a model's predictive accuracy and assess the effectiveness of different techniques in handling multicollinearity. The purpose of this study is to compare several methods of addressing multicollinearity in regression analysis and to determine their effectiveness in improving the a...
Show More
-
A Polynomial-Based Mathematical Modelling of Age-Specific Fertility and Marital Fertility Rates of Assam
Sourav Jyoti Gogoi,
Manab Deka
Issue:
Volume 12, Issue 4, July 2023
Pages:
92-102
Received:
27 July 2023
Accepted:
14 August 2023
Published:
28 August 2023
Abstract: The emergence of population characteristics has become a significant focus within the field of demographic studies. Among these characteristics, fertility holds a pivotal role. In order to obtain a comprehensive comprehension of the complex dynamics of human fertility behaviour, mathematical models are a commonly used and appropriate tool. However, the scholarly discourse on fertility data modelling in the north-eastern Indian state of Assam has been relatively limited. The objective of this study is to investigate the fertility patterns among women in the reproductive age range by creating a polynomial-based mathematical model for Age-Specific Fertility Rates (ASFRs) and Age-Specific Marital Fertility Rates (ASMFRs) specific to the region of Assam. Moreover, we employed diverse statistical methodologies to ascertain the soundness and reliability of the formulated model. The utilization of a quartic polynomial model facilitates a more precise estimation of both ASFRs and ASMFRs. In addition, the Total Fertility Rate (TFR) and Total Marital Fertility Rate (TMFR) are estimated, as well as the age at which mothers experience elevated fertility rates. Velocity and Elasticity curves are fitted to the ASFRs and ASMFRs, unveiling noteworthy disparities between the velocity curves of 2020 and those of previous years (2015 and 2011). Specifically, the ASFRs in Assam exhibited their highest values among mothers aged approximately 27 years in 2020, while in 2015 and 2011, the peak ASFRs were estimated for mothers around 24 years old. Similarly, the ASMFRs in 2020 reached their zenith among mothers aged approximately 15 years, whereas in 2015 and 2011, the highest ASMFRs were estimated for mothers around 21 and 19 years old, respectively. Furthermore, the study sheds light on the decline in the number of women marrying at an illegal age group in Assam during 2020 compared to 2015 and 2011.
Abstract: The emergence of population characteristics has become a significant focus within the field of demographic studies. Among these characteristics, fertility holds a pivotal role. In order to obtain a comprehensive comprehension of the complex dynamics of human fertility behaviour, mathematical models are a commonly used and appropriate tool. However,...
Show More