Yekpay Blog

Data analysis - Customer Loyalty - Mobile Payment (3)

Data analysis - Customer Loyalty - Mobile Payment (3)

Determinants of Customer Loyalty Using Mobile Payment Services in Iran

Data analysis

To test the proposed hypotheses, data were collected and analyzed using structural equation modeling (SEM) supported by AMOS with maximum likelihood estimation. SEM is a second-generation multivariate technique that combines multiple regressions with confirmatory factor analysis to estimate a series of interrelated dependence relationships simultaneously. SEM is a widespread technique in several fields including marketing, psychology, social sciences and information systems.Hence , the structural equation model (SEM) is applied to validate the relationship among variables in the research model. This study applies Amos to perform data analyses. The measurement model is used to test the validity and reliability of items and constructs in the research model. Results of the measurement model are expressed (Table 1). A confirmatory factor analysis (CFA) using AMOS was conducted to test the measurement model. Seven model fit measures were employed to assess the model’s overall goodness-of-fit: the ratio of X 2 to degrees-offreedom (d.f.), the goodness-of-fit index (GFI), the adjusted goodness-of-fit index (AGFI), the comparative fit index (CFI), the norm fit index (NFI), the root mean square error of approximation (RMSEA) and the root mean square residual (RMR). As shown in Table 1 goodness-of-fit for the model was met of(X 2 /df =1.38, GFI = 0.97, AGFI = 0.89, CFI = 0.98, NFI = 0.92, RMSEA = 0.048 and RMR = 0.037), following the suggested cut-off value. After the overall model was accepted we evaluated the psychometric properties of the measurement model in terms of reliability, convergent validity and discriminate validity.

Table 1. Fit indices for the measurement

Customer Loyalty MP 01

Table 2 summarizes the results of internal reliability and convergent validity for the constructs. Internal consistency reliability was also included to test unidimensionality that was assessed by Cronbach’s alpha. The resulting alpha values ranged from 0.72 to 0.89, which were above the acceptable threshold of 0.70 (Nunnally and Bernstein, 1994). Internal reliability and convergent validity were estimated by the composite reliability and average variance extracted respectively.

Table 2.The results of internal reliability and convergent validity

Customer Loyalty MP 02

All factors in the measurement model revealed adequate reliability and convergent validity. To examine discriminate validity we compared the squared correlations between constructs and variance extracted for a construct (Fornell and Larcker, 1981). The results showed that the square correlations for each construct were less than the average variance extracted (AVE) by the indicators measuring that construct, as shown in Table 3;therefore, it is concluded that the scales have construct reliability and validity.

Table 3. Mean, standard deviation, average variance extracted and correlations of each construct

Customer Loyalty MP 03

Structural Model

The final step in model estimation is to explore the path significance of each causal relationship for hypotheses and examine the variance explained by each path in the model. The structural parameter estimates and the variance explained are presented in chart 2. Eight hypotheses were supported and three were rejected. The results are reported and depicted in Table 5 and in chart 2 . Table 1 showed the structural model’s overall goodness-of-fit. The values overall provided evidence of a good model fit [(X 2 /df = 1.34, GFI = 0.97, AGFI = 0.88,CFI = 0.98, NFI = 0.91, RMSEA = 0.044 and RMR = 0.037)]. Thus we could proceed to examine the path coefficients. Chart 2 presents the results with a non-significant path as a dotted line and the standardized path coefficients between the proposed constructs. Customer satisfaction was significantly influenced by the proposed quality dimensions except usefulness .The path coefficients from the quality factors all together explained approximately 48 per cent of variance in the observed variance in satisfaction.

Thus Hypotheses 1, 3, 6, 7 and 9 were supported. Moreover customer loyalty towards the bank was significantly influenced by satisfaction and some proposed quality factors except interactivity and responsiveness. The path coefficients of satisfaction and other quality factors accounted for approximately 48 percent of the observed variance in customer loyalty towards the banks. Thus Hypotheses 4, 10 and 11 were supported.

Table 4. The direct, indirect and total effects on user loyalty

Customer Loyalty MP 04

The direct and total effect of customer satisfaction on customer trust was 0.32 present . However, the total effect of customization on customer loyalty towards the bank was 0.36 because customization had a stronger direct effect than customer satisfaction had on loyalty. Similarly the total effect of perceived risk on customer loyalty towards the bank was greater than that of satisfaction on loyalty. Despite showing a weaker direct effect than user satisfaction on customer loyalty, perceived risk still exhibited a slightly stronger total effect on customer loyalty than that of customer satisfaction. The direct, indirect and total effect of all latent variables are summarized in Table 4.

Customer Loyalty MP 05

Chart 2.Structural model


Table 5. Summary of Hypothesis Tests

Customer Loyalty MP 06


The purpose of this study is to develop empirically test factors that would influence customer loyalty in m-payment. The proposed model enhanced the understanding of formation of customer loyalty. This research verifies the effects of satisfaction and proposed quality factors on customers’ loyalty in m-payment. In line with past studies, this research found that satisfaction is an important determinant of customer loyalty. In addition, while previous studies indicated that interactivity and responsiveness directly affect customer loyalty, here it is found that these two factors influence customer loyalty indirectly. Moreover, it is revealed that the determinants of satisfaction, pave the way for a detailed exploration of how to improve customer satisfaction. This study is undertaken to address these issues by identifying nine critical factors that customer Perceive as important for m-payment loyalty. It was concluded that Security, perceived risk , usefulness, customization , Responsiveness, customer satisfaction and ease of use were the most important. Hence, this study is discusses to represent a useful contribution, relative to theoretical considerations and empirical analysis of the factors that will be of value to systems designers and policy makers working within online transacting environments. In Iran m-commerce has created to business opportunities for banks in general. As more customers conduct their business activities on the mobile internet, the demand for mobile banking services will continue to grow. How to retain existing customers and attract new ones is the issues. The results of this study suggest that by focusing on and improving the m-quality factors, banking on the mobile internet can provide a more satisfying experience to customers. In turn customer loyalty may be developed to help banks retain their existing customers.


Ref :

Prof. Dr.Ali Sanayei, Prof.Dr Bahram Ranjbarian, Dr.Ali Shaemi, Azarnoosh Ansari: INTERDISCIPLINARY JOURNAL OF CONTEMPORARY RESEARCH IN BUSINESS, VOL 3, NO 6 (2011)