Examining Determinants of Customer Bank Selection Using Logistic Regression
Mohamed Fofana
School of Science, Zhejiang University of Science and Technology, Hangzhou, Zhejiang, 310023, PR China.
Shaowei Sun *
School of Science, Zhejiang University of Science and Technology, Hangzhou, Zhejiang, 310023, PR China.
*Author to whom correspondence should be addressed.
Abstract
In increasingly competitive banking markets, understanding the determinants of customer choice between public and private institutions is essential for strategic positioning and market competitiveness. Despite extensive research on consumer banking behavior, limited empirical work has systematically examined how service quality, risk perception, reputation, and cost considerations jointly influence bank type selection using rigorous statistical methods. This study examines the key factors influencing individuals' choice between public and private banks, specifically analyzing four independent variables: value-added services, perceived risk, institutional reputation, and perceived costs. A binary logistic regression approach is applied to data collected from 341 respondents to assess the individual and combined effects of these factors on banking preferences. The findings demonstrate that value-added services (OR=1.374, p=.005) and institutional reputation (OR=1.248, p=.037) significantly increase the likelihood of selecting private banks, while perceived risk (OR=0.600, p<.001) and perceived costs (OR=0.782, p=.008) significantly reduce this likelihood. The model achieved 76% classification accuracy and explained 16.5% of variance in bank choice (Nagelkerke R²=.165). These results confirm that consumers place strong emphasis on service enhancements and brand credibility while remaining highly sensitive to risk exposure and cost considerations. The study advances empirical understanding of consumer behavior in the banking sector by quantifying the relative importance of these factors and offers insights relevant to banking strategy, competition, and customer-oriented policy design.
Keywords: Bank choice, binary logistic regression, value added services, perceived risk, perceived costs