Open Access Original Research Article

Modeling Key Drivers of Under-Five Child Malnutrition in Marsabit County, Kenya: Application of the Logit Model

Benjamin Kiptoo Rop, J. Koech, A. Otieno

Asian Journal of Probability and Statistics, Page 1-12
DOI: 10.9734/ajpas/2021/v12i230281

Malnutrition remains one of the major problems in developing countries affecting both adults and children under 5 years. The use of binary logistic regression model was employed, and parameters of interest estimated. Results showed that 29.3 percent of the children were acutely malnourished. There was an insignificant difference between household food security and child malnutrition status  Factors such as the age of caregivers, household size, the gender of the child, and the level of education of caregivers, if the child was weighed at birth, source of income, the occupation status, and the distance to the water source remained insignificant at a multivariate level. However, factors such as full-term maternal pregnancy, the child being ill for the past two weeks, and the study site were strong significant factors affecting the status of childhood malnutrition. Moreover, mothers with full-term pregnancy up to the birth were 53 percent less likely to have malnourished infants when compared to their counterparts whose pregnancy was not term. Mothers/caregivers who traveled more than half a kilometer were twice more likely to have their children malnourished than those who had traveled less than half a kilometer. It recommended that the policymakers and the entire County government of Marsabit should build more social amenities that provide pregnant women with full-term maternal checkups for both antenatal and postnatal care. County government of Marsabit should lobby and mobilize resources for food aid or cash transfers to households.

Open Access Original Research Article

Factors of the Prevalence of COVID-19 for Females in Amhara Region, Northern Ethiopia

Yenew Alemu

Asian Journal of Probability and Statistics, Page 13-18
DOI: 10.9734/ajpas/2021/v12i230282

The COVID-19 pandemic in Ethiopia is a global epidemic of coronavirus disease 2019 caused by severe acute respiratory syndrome coronavirus 2. Amhara Region is a regional state in northern Ethiopia and the homeland of the Amhara people. The main objective of this study was to identify factors of the prevalence of COVID-19 for females in Amhara region. The data set was obtained from Amhara Public Health Institute 2020.  A negative Binomial regression model was conducted to find out the determinants of the prevalence of COVID-19 for females. Out of 5,627 confirmed cases, 96 patients have died and 1,483 confirmed cases were females from 138 daily reports.  Number of recovered COVID-19 cases ( = -0.4566332, 95%CI: (-0.7364772, -0.1767892), P-value < 0.05), severe cases ( = 1.038589, 95%CI: (0.7531619, 1.324017), P-value < 0.05), total deaths ( = 0.5164175, 95% CI: (0.1438362, 0.8889987), P-value <0.05) and the average age of patients per day ( = 1.511936, 95% C.I: (0.9220257, 2.101846), P-value < 0.05 ) were statistically significant factors for the confirmed case of COVID-19 for females in Amhara region. Above 30 average age of patients per day, below 19 number of recovered cases, above 12 number of severe cases, and above two number of deaths are optimally high-risk factors of confirmed cases of COVID-19 for females.

Open Access Original Research Article

Estimating Stress-Strength Model for Weighted Lomax Distribution

M. M. E. Abd El-Monsef, Ghareeb A. Marei, N. M. Kilany

Asian Journal of Probability and Statistics, Page 19-31
DOI: 10.9734/ajpas/2021/v12i230283

This paper aims to estimate the stress-strength reliability parameter  when  and  are follow the weighted Lomax (WL) distribution. The behavior of stress-strength parameters and reliability have been studied by using maximum likelihood and Bayesian estimators through the Monte Carlo simulation study which carried out showing satisfactory performance of the estimators obtained. Finally, two real data sets representing waiting times before service of the customers of two banks A and B are fitted using the WL distribution and used to estimate the stress-strength parameters and reliability function.

Open Access Original Research Article

Constraint and Unconstraint of Vector Autoregressive Model; Using GDP Growth Rate of Agriculture, Industries, Building/Construction, Whole-Sale/Retail and Services in Nigeria

Ekpenyong, Aniedi. Moses, Isaac, Didi Essi

Asian Journal of Probability and Statistics, Page 32-40
DOI: 10.9734/ajpas/2021/v12i230284

In this research, multivariate Time Series was adopted to model the Gross Domestics Product (GDP) growth rate of Nigeria on five (5) variables namely: Agriculture, Industries, Building/construction, Wholesales/Retails trade and Services The data was collected from National Bureau of Statistics, range quarterly from 1985 to 2017, a total of 33years. Real (R) software was used as a tool to analyze the model. The data were grouped into 10 pairs of 2 parameter variables, 10 pairs of 3 parameters variables, 5 pairs of 4 parameters variables, and the complete 5 parameter variables. In each group, the best model was selected and Lag's using Akaike Information Criteria, then the unconstrained (vector autoregressive) AIC of the model was compared with that of constrained (simplified vector autoregressive) AIC model. The unconstraint models with AIC values (-11.973, -17.1111, -22.1823, and 25.8996) at lag (5) was compared with that of constraint models with AIC values of (-12.5116, -17.5298, -22.2894 and -25.9916), the outcome showed that constraint models performed better than unconstraint models.

Open Access Original Research Article

A New Calibration Estimator of Population Mean for Small Area with Nonresponse

J. Iseh Matthew, J. Bassey Kufre

Asian Journal of Probability and Statistics, Page 41-51
DOI: 10.9734/ajpas/2021/v12i230286

This paper considered the challenges of population mean estimation in small area that is characterized by small or no sample size in the presence of unit nonresponse and presents a calibration estimator that produces reliable estimates under stratified random sampling from a class of synthetic estimators using calibration approach with alternative distance measure. Examining the proposed estimator relatively with existing ones under three distributional assumptions: normal, gamma, and exponential distributions with percent average absolute relative bias, percent average coefficient of variation, and average mean squared error as evaluation criteria using simulation analysis technique, the new estimator exhibited a more reliable estimate of the mean with less bias and greater gain in efficiency. Further evaluation using coefficient of variation under varying nonresponse rates to validate the results of variations suggests that the estimator is a suitable alternative for small area estimation. This finding has therefore contributed to the development of an ultimate estimator for small area estimation in the presence of unit nonresponse.