SCOPUA Journal of Applied Statistical Research
Not a member yet
23 research outputs found
Sort by
The Exact Analysis of Augmented Incomplete Latin Square Design with One Missing Observation
Get PDFAs Federer\u27s augmented Latin square design (ALSD) is one of the most important augmented designs used in plant breeding programs, this paper aims to introduce an exact analysis for an ALSD with a missing check treatment. Namely, the novelty of this present work is evident in proposing and defining the augmented incomplete Latin square design (AILSD). In this light, the effects of rows, columns, checks, and new treatments were evaluated. All previous known studies were limited by focusing on dealing with complete datasets, paying less attention to the problem of the presence of a missing value, and this is the gap we are trying to fill here. Moreover, computing the regression sum of squares (RSS) for both full and reduced models was essential to figure out the other needed sum of squares. A numerical example and R simulation study were carried out to assess the performance of the proposed design. The importance of employing R comes from filling the calculation gaps involved in analysing AILSD
Shako Kumaraswamy distributions with properties and applications in different fields
Get PDFThe Shako Kumaraswamy distribution is a novel and adaptable extension of the Kumaraswamy distribution that is presented in this paper. By adding a shape parameter, the suggested model expands on the traditional Kumaraswamy distribution and improves its ability to present a variety of distributional shapes and tail behaviours. The probability density function, quantile function, cumulative distribution function, moments, and other reliability metrics are all derived as part of a thorough mathematical analysis. Additionally, limiting behaviours and special situations are analysed to demonstrate the generality of the suggested distribution. The behaviour of the estimators under various sample sizes and parameter settings is evaluated in a thorough Monte Carlo simulation study to evaluate the effectiveness of the parameter estimation techniques. Additionally, the Shako Kumaraswamy distribution\u27s practicality is illustrated using two real-world data sets from different domains.
A comprehensive review on the advancement of the Xgamma Distribution
Get PDFThe present article provides a comprehensive review of the development and applications of the Xgamma distribution (XGD). The Xgamma distribution, conceptually parallel to the Lindley distribution, emerges as a combination of the exponential and gamma distributions with an appropriate weighting coefficient. Owing to its flexibility and analytical tractability, the XGD has gained considerable attention in reliability and lifetime data modeling. In recent years, numerous researchers have proposed various extensions and generalizations of the Xgamma distribution by employing diverse transformation techniques and distributional families. These developments have further enhanced its modeling capability and adaptability to complex data structures. A particularly interesting aspect of this review is the discussion of practical applications, where several real data sets used in previous studies are examined to demonstrate the empirical relevance and versatility of the extended forms of the XGD across different scientific and engineering fields
Evaluating Nigeria Exchange Group All Share Index: Insights from Linear and GARCH Modeling Techniques
Get PDFThis study investigates the volatility of the Nigeria Exchange Group All Share Index (ASI) using linear regression and Generalized Autoregressive Conditional Heteroskedasticity (GARCH) modeling techniques. Despite prevalent public concerns regarding stock market instability, our analysis reveals that these perceptions are often exaggerated, driven largely by historical price levels and media representation. We employed a linear regression model to analyze monthly historical ASI data from 1985 to 2023, establishing a significant positive relationship between time and ASI values, with an R2 value of 0.7493, indicating that approximately 75% of the variance in ASI can be explained by the model. The Breusch-Godfrey test highlighted significant serial correlation in the residuals, necessitating further analysis using GARCH models to account for time-varying volatility. Our findings suggest that traditional asset pricing models may overlook alter- native risk measures that investors prioritize, emphasizing the need for a more nuanced understanding of market behavior. The adequacy of the model is achieved with a p-value 0.000017. Overall, this study contributes to the existing literature by offering insights into the dynamics of the Nigerian stock market and its volatility patterns, which are crucial for investors and policymakers alike
Assessment of the performance of the Bayesian Forecasting for Vector ARMA Processes
Get PDFMultivariate time series are widely observed in numerous domains. In economics, for example, you can monitor the annual savings in conjunction with the real interest rate. Such variables are jointly analyzed to understand the dynamic interactions that exist between them, thus improving the precision of forecasts. Enhanced forecasting is achievable when the series are examined together, especially when one series holds information about another one. An approximate Bayesian analytical method to estimate and forecast vector autoregressive moving average (Vector ARMA) processes was introduced by Shaarawy (1989). A basic goal for the research in hand is the numerical assessment of the proposed approach in tackling forecasting difficulties associated with Vector ARMA processes through a comprehensive simulation study. Furthermore, the research checks how the performance of the suggested method fluctuates when varying parameter values and time series lengths. The findings of the numerical study demonstrated that the methodology was effective in accurately forecasting future observations for Vector ARMA processes across various values of the parameter and different time series lengths
A comprehensive study of the extended JCA distribution with properties and applications
No full textoai:ojs2.journals.scopua.com:article/6The JCA distribution, which was first proposed by Jamal et al. (2021), was expanded in this paper. The new model is called the extended JCA distribution, or EJCA. For the researchers, this will be more adaptable and appealing. Explicit expressions for the moment generating function, the analytical form of the density and hazard, order statistics, and mode are among the mathematical properties that are derived. The maximum likelihood approach is used to estimate the model parameters. To evaluate the effectiveness of the estimation, a simulation study with a range of sample sizes is conducted. Two applications to the real-world data set demonstrate the proposed family\u27s adaptability. We combined the classical exponential distribution with the JCA distribution. The new model is extremely flexible and combines the features of both models, including exponential with some probabilities and JCA. For the new distribution, we derive several mathematical properties and applications to actual data sets
Impact of Hypertension and Other Comorbidities on Stroke Subtypes among Patients in Anambra State Teaching Hospital Awka, Nigeria
Get PDFStroke continues to be a major cause of illness and death. Its occurrence is greatly affected by modifiable risk factors. This study aimed to describe the demographic and clinical characteristics of stroke patients and to explore the relationship between hypertension and stroke subtype. A retrospective cross-sectional analysis of 264 stroke patients was conducted. Continuous variables using means, medians, and interquartile ranges were summarized. Categorical variables were described with frequencies and percentages. The prevalence estimates with 95% confidence intervals (CIs) were calculated. Fisher’s Exact Test assessed the simple association between hypertension and stroke subtype (ischemic, haemorrhagic, transient ischemic attack [TIA]). A multinomial logistic regression model, adjusted for age and gender were applied, to estimate adjusted odds ratios (aORs) with 95% CIs, using ischemic stroke as the reference category. The average age was 59.1 years (SD = 16.3), and mean BMI was 29.9 (SD = 6.36), indicating a predominance of overweight and obesity (BMI ≥ 25). Hypertension was the most common comorbidity (91.3%, 95% CI: 87.2–94.4), followed by hyperlipidaemia (97%) and high cholesterol (35%). Type 2 diabetes (11%) and heart disease (26.9%) were less common. Fisher’s Exact Test showed no statistically significant association between hypertension status and stroke subtype (p = 0.362). Stroke subtype distribution was as follows: ischemic stroke (58.7%), haemorrhagic stroke (39.0%), and transient ischemic attack (TIA) (2.3%). In logistic regression analysis, hypertension was not significantly related to haemorrhagic stroke (aOR = 1.47, 95% CI: 0.58–3.76) or TIA (aOR = 0.55, 95% CI: 0.06–5.22) compared to ischemic stroke. Stroke patients in this group were mostly older, obese, and had high blood pressure. Although hypertension was quite common, it did not significantly vary by stroke subtype, highlighting its role as a universal risk factor across different types of strokes
An Enhanced Burr Type III Distribution: Simulation Studies and Practical Applications
Get PDFIn this paper, we introduce a new three-parameter distribution, the New Exponentiated Burr Type III distribution (NEBIII), which is a member of the Generalized (G)-family of continuous distributions. The mathematical features that are derived include the mode, actuarial measures, order statistics, the analytical forms of the density and hazard functions, and explicit formulations for the moment generating function (MGF). The Maximum Likelihood Estimation (MLE) approach is used to estimate the model parameters. A simulated research with different sample sizes is used to evaluate the estimation\u27s efficacy. The versatility and adaptability of the proposed distribution family are demonstrated on four real-world data sets. Additionally, we examine a Mixture of the Exponential and Exponentiated Burr Type III Distributions, identifying various mathematical characteristics and demonstrating their application to actual data
A New Generalization of Exponentiated Exponential Distribution using Quantile Functions
Get PDFGeneralising existing probability distributions increases their appeal to researchers and expands their applicability to real-life situations by adding flexibility to the existing models. In this study, a new generalised class of the exponentiated exponential (EE) distribution is introduced, referred to as the T-Exponential Exponential{Y} or T-EE{Y} class of distributions. By utilising the quantile functions of several well-known continuous distributions in the T-EE{Y} framework, six distinct subclasses have been developed. Various statistical properties such as quantiles, mode, incomplete moments, entropy, and mean deviation have been derived for these subclasses. Additionally, four specific member models within the proposed class have been explored. The study demonstrates that the members from the T-EE{Y} class exhibit flexible shapes, including unimodal, bimodal, symmetrical, and skewed (both right and left) forms. Parameter estimation is performed using the maximum likelihood estimation method, and the effectiveness of the estimators is evaluated through a Monte Carlo simulation study. To evaluate the practical applicability of the proposed class of distributions, three real-world datasets are analysed. The members of the proposed class consistently outperform several existing distributions in modelling lifetime data, showcasing its significance and versatility
Statistical analysis to assess the satisfaction level of foreign students in China concerning accommodation provided by the university
Get PDFThe focus of this study was the assessment of the satisfaction level regarding accommodation provided by the University to International students studying in China. Southwest University of Political Science & Law, China was taken as a case study to complete this study. A questionnaire was designed and 152 international students studying at the Yubei campus were approached personally. The students filled out the questionnaires based on the main five variables i.e. overview, bedrooms, dormitory services, study environment, and cultural exchange. SPSS and SmartPLS were used for data analysis. The factor loadings for all items exceed the threshold of 0.6, indicating strong associations between the items and their corresponding constructs. The Variance Inflation Factor (VIF) values, which measure multicollinearity, are all below 3, confirming that there are no significant redundancy issues among the items. Cronbach’s alpha values for all constructs are above 0.7, demonstrating good internal consistency reliability. The Average Variance Extracted (AVE) values, which assess convergent validity, are all above 0.5, indicating that each construct explains more than half of the variance in its items. About a majority of SWUPL international students show considerably high satisfaction with university housing facilities, along with the student assistance structure and out-of-classroom possibilities. The positive experiences reported by students directly related to the strong cooperation of SWUPL’s International Student Office and the university’s commitment to academic and non-academic activities