Statistical model for estimating the probability of using electronic cards : a statistical analysis of SCF data

The use of electronic banking products are increasingly important aspects of American life. Relatively little attention has been paid to the demand side of the financial service market, most importantly, consumers. To understand significant predictors to influence on the use of electronic payment instruments, we have applied stepwise logistic regression, probit regression and random effect probit model to a set of reported SCF data in U.S. and compared the results. The causal relationship between predictors and the adoption of electronic cards is investigated; a discussion of findings is given.