An empirical examination of behavioral intention to mobile Internet ...

An empirical examination of behavioral intention to mobileInternet banking: The case of EgyptTarek Taha Ahmed *Mobile Internet banking has changed the business of retail bankssignificantly in terms of cost reduction and increased convenience for thecustomers. However, most of the existing literature focused ondeveloped countries, while the worldwide growth of mobile commercehas shown the need to extend the research to other unstudieddeveloping countries with different cultures and from differentperspectives. Thus, the current paper is one more attempt to fill thesegapes in the current body of literature, specifically in the developingcountries context, by empirically examining and validating factors thataffect customer’s behavioral intention to use mobile Internet banking.This study developed an empirically-based model including factors,which have never been integrated into one framework, to examinationsimultaneously. A richer research methodology is used in our empiricalstudy combining quantitative and qualitative methods to validate theresearch model. Based on our findings the study has made a number ofimportant managerial and academic implicationsField of Research: Marketing1. Introduction:Banking services sector nowadays is increasingly relying on the Internet as anelectronic distribution channel, for delivering services directly to customersthrough bank websites, where customer use web browser to access theiraccounts, which is known as Internet banking, Net banking or web-bankingThese terms are often used interchangeably (e.g. Dimitriadis et al., 2011;Malhotra & Singh, 2010; Oliveira &Hippel, 2011, Shah et al., 2009; Callaway,2011; Lee at al., 2011).The latest generation of Internet banking is the current mobile Internet banking(m-banking), based on the convergence of Internet, wireless technology andmobile devices, in which customers can access bank websites via their mobiledevices and perform a wider range of banking transactions anywhere andanytime, , without visiting banks in person, limited only by coverage provided bythe mobile networks, including, account balance checking, account movementstatement, transferring between accounts, and electronic payment of bills (Lin,2011; Luo et al., 2010; Peevers et al., 2011)._______________* Prof. Dr. Tarek Taha Ahmed, the faculty dean of Pharos University in Alexandria, Egypt, e-mail:Dr.Tarek.Ahmed@pua.edu.eg, ProfDrTarekTaha@gmail.com1

However, the empirical evidence from past research tends to support the viewthat mobile Internet banking reduces service costs for banks and adds value tocustomers by supporting their mobile lifestyle and it is believed to have thepotential to expand significantly in the future despite its relatively slow usage ratecompared with Internet banking users (e.g. Chung & Kwon, 2009; Cruz et al.,2009). Others, as Gu et al. (2009) argued that the trend of mobile bankingindicates a remarkable potential to banking industry and banks have theopportunity to convert cell phone users into banking users.In summary, m-banking has changed the business of retail banks significantly interms of cost reduction and increased convenience for the customers, despite itsinternational adoption rates are still low and many banks try to market it to mobileusers. Therefore, identifying and examining the factors contribute to customer’sintention to use m-banking considered a fundamental requisite for developing thistype of service (e. g. Lin, 2011; Zhou et al., 2010; Luo et al., 2010; Yu & Fang,2009; Petrova & Ya, 2010).2. Research Problem, Objectives and Plan:Despite the progress of research in e-commerce Literature most of the existingprevious works about Internet banking in general and mobile Internet bankingparticularly have focused on developed countries, with a greater predispositiontoward the Internet, and mainly examined customers’ intention from trustperspective, while the worldwide growth of mobile commerce has shown theneed to extend this research to other unstudied developing countries withdifferent cultures and from different perspectives (e.g. Hernandez et al., 2010;Petrova & Yu, 2010; Lee & Chang, 2009; Krauter & Faullant, 2010; Zhou et al.,2010). In addition, our preliminary study revealed that the mobile Internet bankingremains largely unnoticed by a retail banking customers of Egypt.This relatively inadequate coverage poses problems for developing countriesbecause limitations in this area mean difficulties for their banks to properly planand successfully deliver mobile banking service. Thus, the current paper is onemore attempt to fill these gapes in the current body of literature, specifically in thedeveloping countries context, by examining and validating empirically the criticalfactors that affect customer’s behavioral intention to use mobile Internet banking.Specifically, the present investigation adds to literature through achieving thefollowing objectives: (1) identify the potential critical factors that have the mostsignificant influence on customer’s intention to use mobile Internet banking, forenhancing banking B2C mobile commerce applications, (2) determine therelative importance of each of these factors for retaining and attracting mobileusers, (3) develop and validate a mathematical model, that can systematicallypredict the probability of mobile Internet banking usage in the B2C e-commercemarket of Egypt as an example of developing country based on empiricalevidences.2

With these objectives in view, the current paper has been organized as follows:the literature and relevant studies were reviewed and analyzed. Then a researchmodel was proposed and hypotheses were formulated to be tested in the study.This was followed by an explanation of the procedures used to obtain data,measurement, and validation processes, as well as the testing of the hypothesesstated. Finally, based on our findings a series of conclusions with managerialimplications and final thoughts that emphasize the great interest in the topicunder analysis were presented; and then certain limitations and future lines ofresearch with regard to this issue were highlighted.3. Literature Review:Relevant literature and past research were extensively reviewed and integratedsequentially, including a wide range of recently published works, in order todevelop more effectively the study’s hypotheses and the research model. Thisextensive revision, which provided the conceptual foundation for our study,revealed some critical factors that have a significant impact on customers’behavioral Intention to use m-banking. In order to expand the research scope,the current paper has operationalized to major dimensions, to be simultaneouslyexamined.The first dimension included the perceptional factors that relatively under abank’s control and may affect either positively or negatively the customers’intention to use m-banking such as the customer’s perceived trust of the bank asan e-service vendor, perceived usefulness of mobile Internet applications andperceived ease of use, as positive factors, and perceived security risk of m-commerce transactions as a negative factor (see Gu at el., 2009; Chung & Kwon,2009; Yu & Fang, 2009; Shah et al., 2009; Luo et al., 2010). In e-commercecontext, perceived trust is defined as a belief the vendors are willing to behavebased on an individual’s expectation (e.g. Gu at al., 2009).The second dimension included the personal factors that are not under thecontrol of the bank, so they are viewed by many bank practitioners as potentialbarriers to m-banking usage by customers. These factors have been described inprevious studies as the educational level, income; age and mobile Internetexperience (e.g. Proenca & Rodrigues, 2011; Luo et al., 2010; Winley, 2011).In contrast to previous works, the current study extended the research scope bycombining the most critical factors identified in literature and developed anempirically-based model including these factors, which have never beenintegrated into one framework, to examination simultaneously for validation andrelationship.3

4. Developing the research model and Hypotheses:The preceding discussion of literature as well as the feedback obtained from ourpreliminary study formed the theoretical foundation of the research modelillustrated in figure 1, which incorporated many of the relevant features of m-banking identified in the literature and applied these to our local context.Fig.1. The research modelPerceptual DimensionP1 (+)P5 (+)PTR PUS PES PSRP2 (+) P3 (+) P4 (-)Customer’s intention to use mobile Internet Banking (YINT)MIEP6 (+)EDUP7 (+)INCPersonal DimensionP8 (-)AGEWhere:PTR= Perceived trust of bank as an e-services vendorPUS= Perceived usefulness of mobile banking applicationPES= Perceived ease of use of mobile banking applicationPSR= Perceived security risk of m-commerce transactionsMIE= Mobile Internet experienceEDU= Educational levelINC= IncomeAGE= AgeYINT= Intention to use mobile Internet bankingAs seen from the figure, the proposed model, contained 8 structural paths,representing the hypothesized relationships embedded within two dimensionsbetween independent variables as predictors (perceptual and personal factors)and the dependent variable customer’ intention to use m-banking as a criterion.In order to validate this model a number of validity tests have been applied andrelationships among its variables have been empirically examined. Based on thestructural paths of the research model the following hypotheses were formulatedto be simultaneously tested.H 1 : Customers’ intention to use mobile Internet banking is positively influenced by certainperceptual factors such as:H 1a : Perceived trust of bank as an e-services vendor (P1)H 1b Perceived usefulness of mobile banking application (P2)H 1b : Perceived ease of use of mobile banking application (P3)H 2 : Customers’ intention to use mobile Internet banking is negatively influenced by theirperceived security risk of m-commerce transactions (P4)H 3 : Customers’ intention to use mobile Internet banking is positively influenced by certainpersonal factors such as:H 3a : Customer experience with mobile Internet (P5)H 3b : Customer educational level (P6)H 3c : Customer income (P7)H 4 : Customers’ age negatively affects their intention to use mobile Internet banking (P8)Symbolically, the initial multiple regression equation (EQ1) can be presented asfollows to predict the probability of customer’s intention to use m-banking.EQ1:Y INT = a+ b PTR PTR+ b PUS PUS + b PES PES – b PSR PSR + b MIE MIE + b EDU EDU + b INC INC - b AGE AGE4

5. Research Methodology:A richer research methodology is used in our empirical study combiningquantitative and qualitative methods to validate the research model. Thus, theresearch process involved multi-stage procedures as follows:5.1 Preliminary qualitative study: In this stage, a series of in-depth interviewswere carried out to enhance our understanding of the nature and essence of thephenomenon under investigation, the issues arising from this stage, were usedas a basis for the next quantitative study.5.2. Quantitative study and sampling: The quantitative stage in the form ofquestionnaire survey was conducted to collect empirical data from retail bankingcustomers of banks operating in Egypt. The list of banks registered with theCentral Bank of Egypt at 30 August 2010 (http://www.cbe.org.eg/links.htm)served as the sampling frame for this study. Simple random sampling wascarried out in order to gain as many representative samples as possible.5.3. Instrument development and Validity: To develop our instrument a number ofprior relevant studies and corresponding scales were reviewed to ensure that acomprehensive list of measures were included. Multi-items measures weregenerated for each construct and assessed for the reliability and content validity.The questionnaire consisted of two parts including a portion for respondent’spersonal data and another dealt with the perceptual variables. Eachquestionnaire items of perceptual constructs were measured on 7-point Likertscales, in which 1 indicates “completely disagree” and 7 “completely agree”. Prior tothe conduct of a formal survey, a pre-test was carried out to validate the initialversion of the survey questionnaire. The results from the pre-test study led to thefinal version of the questionnaire, and the instrument has confirmed contentvalidity.5.4. Research design and Reliability: The research design for this study involveda cross-sectional survey methodology, which was conducted betweenNovember, 2010 and Jan, 2011. The questionnaire was originally developed inEnglish, and subsequently translated into Arabic language. Among a total of 700questionnaires that were personally distributed, 328 valid responses werereceived and used in data analysis, after discarding some erroneous responses,achieving a 46 percent usable response rate for the overall survey.Despite the relatively low response rate, which thought to be expected for thelocal context, the fact that the respondents were as representative of thepopulation as possible, led to their contribution being regarded as providinginformation applicable to the larger population. The reliability and internalconsistency of instruments was assessed using Cronbach's alpha coefficient. Allalpha values exhibited an acceptable degree of reliability (alpha > 0.7).5

6. Data Analysis and model testing:The empirical data obtained were processed using statistical software packages(SPSS). Multiple regression analysis with its associated statistical inference tests(F test and t-test on b) was performed, to investigate the strength and direction ofthe hypothesized relationships among the model’s constructs. Eight independentvariables were used as regressors, while the dependent variable was served asregress.6.1. Multicollinariy: To determine whether any multicollinearity effects existed,total correlation matrix of the research model was reviewed in-depth, and theresults showed that there was no evidence detected of severe multicollinearityproblem among regressors. The results of testing each of the four hypothesesare given below.6.2. The results of hypotheses testing: As shown in Table 1, the results ofmultiple regression analysis provide strong support for the hypotheses; also thesignificant testing findings presented in table 2 provided further evidence for thisacceptance.Table 1: Summary output of the multiple regression analysisCoefficients a Symbols ValuesRegression StatisticsMultiple correlation coefficient Multiple R 0.94723367Coefficient of multiple determination R 2 0.89725163Adjusted R Square Adjusted R 2 0.89467487Standard Error SEE 0.69580488Observations N 328ANOVA bRegression SS reg 1348.667684Residual SS res 154.442072Total SS total 1503.109756F-test overall model F 348.209028766214*Degrees of freedom df 1, df 2 8, 319a Criterion variable: Y EPIb. Predictors: (constant), PTR, PUS, PES, PSR, MIE, EDU, INC, AGE*Significant at (p < 0.0000 level)Table 2. Variables included in the research model equationRegressionBetat-testFactors CoefficientsCoefficientsSymbol Value Symbol Value Value Sig.PTR B PTR 0.53877344 Beta PTR 0.521 10.029357 0.000000000PUS B PUS 0.34908702 Beta PUS 0.192 4.5785714 0.000000000PES B PES 0.19532828 Beta PES 0.149 4.6270737 0.000000000PSR B PSR -0.09029877 Beta PSR -0.087 -3.849057 0.000120679MIE B MIE 0.1627428 Beta MIE 0.180 3.8876813 0.000116659EDU B EDU 0.03003920 Beta EDU -0.025 1.0052776 0.31552511INC b INC 0.03373515 Beta INC 0.021 0.6049265 0.54565783AGE b AGE -0.04707059 Beta EDU -0.029 -1.356486 0.17590354Intercept a 0.66889313df n-k-1 319Notes: n=sample size, k=No. of independent v6

A strong significant meaningful correlation is found between customers’ intentionto use m-banking and the above mentioned independent variables (Multiplecorrelation coefficient: Multiple R= 0.94723367). The F statistic value(F=348.209028766214 at p < 0.0000 level) is statistically significant indicating thatthe results of the model could hardly have occurred by chance. The coefficient ofdetermination, multiple R-square showed that these predictor factors explainedthe major proportion (89.72 %) of the variability observed (R 2 =0.901606115),which reinforce our confidence in the hypotheses testing results and providessupport for the above mentioned association.Furthermore, the adjusted R 2 of the model, which is a more conservativeestimate of variance by considering error variance, is 0.89467487. Thus, theoverall explanatory power of the research model is considered high and quitecapable of explaining the variance of customers' intention. Using the values ofthe regression coefficients, shown in table 2, the future intention of using m-banking can be predicted, in this study, by the following equations (EQ2):EQ2:Y INT = 0.67 + 0.54 PTR + 0.35 PUS + 0.19 PES - 0.09 PSR + 0.16 MIE + 0.03 EDU+ 0.03 INC - 0.05 AGEBased on the standardized Beta coefficients of each predictor variables and t-tests in Table 2 it can be stated that that within 8 independent variables, includedin EQ2, only 4 variables considered critical highly significant predictors positivelyaffecting the criterion variable Y INT . Those variables are perceived trust (Beta PTR =0.521, p < 0.0000), perceived usefulness (Beta PUS= 0.192 p < 0.0000), mobile Internetexperience (Beta MIE = 0.180, p < 0.0000) and perceived ease of use (Beta PES = 0.149, p< 0.0000), While perceived security has a significant negative impact (Beta PSR =0.087, p< 0.0000). More specifically perceived trust was found to be the mostimportant determinant of customers’ intention to use m-banking (highest betavalue and t-value). This implies that if customers trust the bank as an e-serviceprovider, they become more willing to use m-banking offered by this bank.7. Conclusion and Implications:As stated previously, the main objective of this paper was to contribute to boththeory and practice of m-banking applications and to help address some gaps inthe current body of literature, through expanding the research in this area bydeveloping a comprehensive empirical model that can predict and indentify thecritical factors that have the most influence on the customers’ behavioralintention to use mobile Internet banking, which have never been integratedbefore into one framework, to examination simultaneously for validation andrelationship. More specifically, this study has made a number of importantmanagerial and academic implications7

In term of managerial implications, our empirical results suggested that in orderto obtain higher m-banking usage rate, specifically in developing m-commercemarket, practitioners must continuously work to improve the perceived trust levelof their toward the bank as an e-service provider and special attention should bepaid to secure mobile banking transactions.From an academic and research standpoint, this study provides empiricalevidences and validation for the existing specialized literature concerning m-commerce. Our findings supported the research model that has a high overallexplanatory power, and reinforced its robustness in predicting customers’behavioral intention. Our research attempted to integrate and encompass themost frequently cited factors in the literature, and applied them in the localcontext in order to best examine the phenomenon. Therefore, the proposedmodel contained variables with different dimensions that have not been testedsimultaneously in previous works.8. Limitation and Future research:Although this paper is differentiated from other previous work and expanded theresearch scope, as in any study, there are a few limitations that should beconsidered when interpreting the results and implications. First, the researchmodel was validated using sample data gathered from Egypt and therefore thefindings may be specific to the culture in this developing country. Since the studyis cross-sectional in design, a further examination of our argument using alongitudinal study is recommended in the future to investigate our model indifferent time periods. Finally, we must point out that although the majority of thehypothesized relationships were validated, and significant, and the proposedmodel yielded a relatively high level of coefficient of multiple determination; theobtained value of R 2 indicated that there is still need to find additional variables,to improve our model’s ability to predict the future customers’ behavioral intentionto use m-banking. Nevertheless, further researches could examine the currentproposed model in other countries with different cultures, and makecomparisons, to see whether it can be applied, also future studies can usedifferent methodologies and samples to test whether it makes sense.9. References:Callaway, S., (2011), Internet banking and performance: The relationship of website traffic rank and bank performance, American Journal of Business, Vol.26(1), pp:12-25Chung, N. and Kwon, S., (2009), The effects customers’ mobile experience andtechnical support on the intention to use mobile banking, Cyber Psychology& Behavior, Vol. 12 (5), pp: 539-543Cruz, P., Laukkanen, T. and Munoz, P., (2009), Exploring the factors behind theresistance to mobile banking in Portugal, International Journal of E-servicesand Mobile Applications, Vol. 1(4), pp: 5332-53528

Dimitriadis, S, Kouremenos, A. and Kyrezis, N. (2011), Trust-basedsegmentation: Preliminary evidence from technology-enabled bankchannels, International Journal of Bank Marketing, Vol. 29, pp: 5-31GU, J., Lee, S. and Suh, Y., (2009), Determinants of behavioral intention tomobile banking, Expert Systems with Applications, Vol. 36, pp: 11605-11616Hernandez, B., Jimenez, J. and Martin, M., (2010), Customer behavior inelectronic commerce: The moderating effect of e-purchasing experience,Journal of Business Research, Vol. 63, pp 964-971Krauter, S. and Faullant, R., (2010), Internet trust as a specific form oftechnology trust and its influence on online banking adoption, InternationalJournal of Dependable and Trustworthy Information Systems, Vol. 1(4), pp:43-60Lee, K. and Chung, N., (2009), Understanding factors affecting trust in andsatisfaction with mobile banking in Korea: A modified Delone and McLean’smodel perspective, Interacting with Computers, Vol. 21, pp: 385-392Lee, K., Tsai, M. and Lanting, M., (2011); From marketplace to marketspace:Investigating the customer switch to online banking, Electronic CommerceResearch and Applications, Vol.10, pp:115-125Lin, H., (2011), An empirical investigation of mobile banking adoption: The effectof innovation attributes and knowledge-based trust, International Journal ofInformation Management, Vol. 31, pp: 252-260Luo, X., Li, H., Zhang, J. and Shim, J., (2010), Examining multi-dimensional trustand multi-faced risk in initial acceptance of emerging technologies: Anempirical study of mobile banking services, Decision Support Systems, Vol.49, pp: 222-234Malhotra, P. and Singh, B. (2010), An analysis of Internet banking offerings andits determinants in India, Internet Research, Vol. 20, pp: 87 – 106Oliveira, P. and Hippel, E., (2011), Users as service innovators: The case ofbanking services, Research Policy, Vol. 40(6), pp: 806-818Peevers, G., Douglas, G., Marshall, D. and Jack, M., (2011), On the role of SMSfor transaction confirmation with IVR telephone banking, InternationalJournal of Bank Marketing, Vol.29(3) pp:206 – 223Petrova, K. and Yu, S., (2010), SMS Banking: An exploratory investigation of thefactors influencing future use, International Journal of E-services and MobileApplication, Vol. 2 (3), pp: 19-43.Proença, J. and Rodrigues, M., (2011), A comparison of users and non-users ofbanking self-service technology in Portugal, Managing Service Quality, Vol.21(2), pp:192-210Shah, M., Mohsin, M., Mahmood, Z. and Aziz, R., (2009), Organizational Barriersin offering E-banking, Journal of Electronic Commerce, Vol. 7 (2), pp: 67-82Winley, G., (2011), Factors associated with the use of personal Internet bankingin Thailand, Journal of Electronic Commerce in Organizations, Vol. 9(2),pp:15-409

Yu, T. and Fang, K., (2009), Measuring the post-adoption customer perception ofmobile banking services, Cyber Psychology & Behavior, Vol. 12 (1), pp: 33-35Zhou, T., Lu, Y. and Wang, B., (2010), Integrating TTF and UTAUT to explainmobile banking users adoption, Computer in Human Behavior, Vol. 26, pp:760-76710