scholarly journals Empirical likelihood inference and goodness-of-fit test for logistic regression model under two-phase case-control sampling

Author(s):  
Zhen Sheng ◽  
Yukun Liu ◽  
Jing Qin
2020 ◽  
Vol 18 (4) ◽  
pp. 25-36
Author(s):  
Oluwayemisi A. Abisuga-Oyekunle ◽  
Mammo Muchie

In South Africa, exploiting economic opportunities in the handicraft sector could create livelihood and employment for ordinary citizens living in rural areas. The potential contribution of handicraft small enterprises to sustainable livelihoods and poverty alleviation is yet to be fully exploited. It is also regarded as a sector with great growth potential, but the degree of support provided to the handicraft sector is low. The study aims to evaluate the socioeconomic factors influencing the viability of handicraft small businesses operating in KwaZulu-Natal. Data collection was drawn from a stratified random sample of 196 handicraft practitioners operating in different areas of KwaZulu-Natal Province with a structured questionnaire. Data analysis was performed with the STATA statistical package. The results obtained from the study have shown that 84 enterprises (42.86%) were not viable, whereas 112 of the 196 handicraft enterprises (57.14%) were viable. The percentage of overall correct classification for this procedure was equal to 77.96%. Percentage sensitivity for the fitted logistic regression model was equal to 60.71%. Percentage specificity for the fitted logistic regression model was equal to 82.14%. The p-value obtained from Hosmer-Lemeshow goodness-of-fit test was equal to 0.0884 > 0.05. This indicates that the fitted logistic regression model is fairly well reliable. The findings from the analysis showed that two factors significantly influenced the viability of handicraft enterprises. These two factors were the belief that handicraft business could sustain the handicraft practitioner, and the level of support for handicraft businesses from non-governmental organizations is decreasing. AcknowledgmentSouth Africa SarChi Chair, Nation Research Fund and Department of Science and Technology, South African, for providing funding for this research.


2004 ◽  
Vol 24 (1) ◽  
pp. 121-130 ◽  
Author(s):  
Nico Nagelkerke ◽  
Jeroen Smits ◽  
Saskia le Cessie ◽  
Hans van Houwelingen

2020 ◽  
Vol 18 (1) ◽  
Author(s):  
Rand Wilcox

For a binary random variable Y, let p(x) = P(Y = 1 | X = x) for some covariate X. The goal of computing a confidence interval for p(x) is considered. In the logistic regression model, even a slight departure difficult to detect via a goodness-of-fit test can yield inaccurate results. The accuracy of a confidence interval can deteriorate as the sample size increases. The goal is to suggest an alternative approach based on a smoother, which provides a more flexible approximation of p(x).


Sign in / Sign up

Export Citation Format

Share Document