Binary Logistic Model to Identify the Factors Associated with Households with Bank Accounts in Nepal

Authors

  • Santosh Kumar Shah Lecturer, Central Department of Statistics, Tribhuvan University, Kathmandu, Nepal

DOI:

https://doi.org/10.3126/qjmss.v2i2.33304

Keywords:

Multiple Logistic Regression, Goodness of Fit, Osius and Rojek Test, Stukel’s score test, Receiving Operating Curve

Abstract

Introduction: Banks play an important role in ensuringthe economicand social stability, and the sustainablegrowth of the economy. The savings and other accounts in financial institutions, including banks, finances, microfinances and cooperatives, enable people to execute important financial functions. Thus, households that have accounts in any of financial institutions can have access to various banking services.

Objective: The objective of the study is to identify the factors associated with households having bank accounts in Nepal.

Methods: The analysis is based on household data extracted from the dataset of Nepal Demographic and Health Survey, 2016. The dependent variable is dichotomous, as the households with bank accounts and without bank accounts in any formal financial channels. In order to identify the factors associated with households receiving financial services in Nepal, multiple logistic regression models were developed by examining the model adequacy test.

Results: The study finds that a total of 66.9% of the households had bank accounts. Several variables were found to be 1% of significance level. The predictive power of the model is found to be 31.2% and multicollinearity among the independent variables was absent. The Hosmer-Lemoshow goodness of fit test revealed that the data were poorly (p-value=0.056) fitted by the model. However, Osius-Rojek goodness of fit test (z=0.11; p-value=0.911), Stukel test (Z=0.683, p-value=0.494), likelihood ratio test (χ2=2770; p-value<0.0001) and area under receiver operating curve (79.8%) revealed that fitted model was good.

Conclusion: Multiple logistic regression model revealed that in mountainous and hilly regions, women-headed households have less chances of not having bank accounts compared to the Terai region and men-headed households. The chances of having a bank account in province-2 is even worse than in Karnali and other provinces. The odds of not having bank accounts gradually decreased with the increase in size of agricultural land, wealth index, increase in family size and the number of family members who have completed secondary education.

Downloads

Download data is not yet available.
Abstract
309
pdf
554

Downloads

Published

2020-12-15

How to Cite

Shah, S. K. (2020). Binary Logistic Model to Identify the Factors Associated with Households with Bank Accounts in Nepal. Quest Journal of Management and Social Sciences, 2(2), 323–336. https://doi.org/10.3126/qjmss.v2i2.33304

Issue

Section

Research Papers