Association Between Strategic Values and E-Banking Adoption in Iranian Banks

Association between Strategic values and e-banking adoption in Iranian Banks

Dr. Mohammad Aghdassi,Dr. Lennart Persson,Roja Ghasemi

Abstract

This paper attempts to understand strategic value of e-banking for Iranian banks andexamine the causal effect of perceiving e-banking as a value and its adoption.

We propose an e-banking adoption model that is identifying five factors that have been found to beinfluential in the perception of strategic value of IT: performance support, operational support, managerialproductivity, and strategic decision aids. We also identified eight factors that influenceelectronic banking adoption: organizational readiness, Infrastructural readiness,external dependency, Intangible pressure, persuasive pressure, perceived ease of use,and perceived usefulness.Data are collected via a questionnaire-based survey from Decision maker unit of Iranian Banks.

We can express the result of this study such that bank managers' perception through e-commerce is very positive and effective in their adoption trend. This perception will help them accelerate the adoption process.

Keywords: Perception, Adoption, Strategic value, Banking industry, DMU

Dr. Mohammad Aghdassi: Associate professor of TarbiatModarresUniversity (Iran),

Dr. Lennart Persson: Assistant Professor of Luleå University of technology (Sweden), ,

Roja Ghasemi: Master student of marketing and e-commerce: LuleåUniversity of technology (Sweden) and TarbiatModarresUniversity (Iran),

1. Introduction

“Among the myriad of computer- and telecommunication-based applications inthe modern era, the advent of e-commerce is having the biggest impact onorganizations. E-commerce is changing the way organizations perform their tasks,interact with customers and, in general, do their business.E-commerce enables firms to reduce telecommunication costs, minimize warehousingexpenses, and cut down the distribution chain (QuaddusAchjari 2005)”.“Technological developments particularly in the area of information technology arerevolutionizing banking industries (SohailShanmugham 2003)”.

E-commerce adoption in a bank means using all electronic means of data transmissionand financial transaction bank to bank, bank to customer and customer to customer. Itis not always the matter of monetary transaction sometimes banks grant a credit forsomeone or for some companies that is very valuable in a business network. On theother hand E-banking means to provide facilities for staff to enhance their efficienciesin offering bank services in a branch and also among branches and other banks allover the world. Along with providing hardware and software facilities for customerssuch that they can work with bank and use any services safety, without being there24h a day. By adopting e-commerce, banks can work as facilitators and acceleratorsfor different industries and firms in several scales, small and large, so the businessworld will find a new definition and the global market will become realistic. Bythe adoption of e-commerce their traditional ways of doing business has changed to ahighly dynamic communication that not only is cost-effective but also is revenuegenerating. In this regard many banks and financial institutions are activelydeveloping new way of transaction for themselves and their customers throughout theworld but still it is a very new market to enter and to work in.In spite of the many potential advantages of e-banking, its adoption by Iranian Banksremains limited or none. It is very limited in service channels and very incomplete inoffering e-banking services to customers. Because of this shortcomes, e-banking inIran causes many duplications in work processes and causes customer dissatisfaction.Does this mean that top managers/ owners of Banks do not realize the strategic valueof e-banking to their organization?

This study may identify main factors which facilitate the adoption process bymanagers and experts. This study would be one of the first studies that examine theIranian managers’ attitudes towards e-commerce adoption in Iranian banks.The aim of this study is to examine the determinant factors of strategic value andadoption of electronic commerce as perceived by managers and experts in Iranianbanking industry using the model which is proposed by Grandon and Pearson (2004)for e-commerce adoption. By using this model of e-commerce adoption, we examinedthe causal relationship between determinant factors of strategic value and adoption ofelectronic commerce as perceived by decision maker unit of Iranian Banks. Theobserved research model consists of two main concepts: perception and adoption.Three factors have been found to be influential in the perception of strategic value ofinformation technologies in previous researches: operational support, managerialproductivity, and strategic decision aids, along with five factors that influenceelectronic commerce adoption according to literature: organizational readiness,compatibility, external pressure, perceived ease of use, and perceived usefulness. Wehypothesized a causal link between the perceived strategic value of electroniccommerce and electronic commerce adoption. To validate the research model, wecollected data from decision maker unit of Iranian Banks, both governmental andprivate, that contains Top managers, assistant, managers and co-assistants by using asurvey.

2. Literature

To reach e-commerce objectives, many studies were done to find the problemof e-commerce adoption in different industries. In this regard, numbers of influencingfactors were developed among them for example in the study of (Quaddus & Achjari,2005) Key factors impacting e-commerce are differentiated according to theircontribution to the success of e-commerce and to the locus of impact. Or inthe study of (Hong Zhu 2005), which was for better positioning of firms whenadopting e-commerce for revenue generation, a conceptual model is developed upontechnology diffusion theory, TOE framework, for assessing e-commerce adoption andmigration, incorporating six factors in the technology-context (technology integration,web spending, and web functionalities, electronic data interchange (EDI) use,outsourcing partner usage, and perceived obstacles). It was also attempted to identifythose factors which affect the adoption of e-commerce among SMEs in the study of(Ching Ellis 2004). These factors were described within 3 in dependent variables:

1- Decision maker characteristics (Age, Education, and Cosmopolitanism)

2-Innovation characteristics (Relative advantage, Compatibility, Complexity, Cost effectiveness)

3- Environmental characteristics (Supplier incentives, Customerpressure, and Competitive intensity).

In the study of (Grandon Pearson 2004) aresearch model was developed that suggested three factors that had been found to beinfluential in the perception of strategic value of other information technologies:operational support, managerial productivity, and strategic decision aids. They alsoidentified four factors that influence electronic commerce adoption: organizationalreadiness, external pressure, perceived ease of use, and perceived usefulness. A causallink between the perceived strategic value of electronic commerce and electroniccommerce adoption is hypothesized. The model that they proposed then was completedby adding compatibility to adoption factors as follows in figure 1:

Figure 1: Grandon and Pearson model of adoption

The factors influencing e-commerce adoption in this study have been extracted fromseveral studies in IT adoption. In addition the main structure of (Grandon Pearson)'s model corresponds to the structure of models, theories and frameworkssuch as TAM (technology acceptance model, TPB (theory of planned behavior), Diffusion theory and TOE (Technology-organization-environment) framework. Consequently, it wasinteresting to investigate (Grandon Pearson)'s model as the research model of thisstudy.

Therefore, this study is an investigation of causal effect of perception on e-bankingadoption in banking industry in Iran by using (Grandon et. al)'s adoption model whichhas been tested before in SMEs in US. This study is a combination of two differentstudies that have been represented in an adoption model. The former has been studiedby Subramanian and Nosek (2001) and others (Barua et. Al 1995, Chan 2000) whilethe latter has been investigated by Davis (1989) and others (e.g. Adams et. al 1992,Igbaria et. al 1997, Lederer et. al 2000 and Venkatesh et. al 1996) primarily throughthe technology acceptance model (TAM).

3. Methodology

3.1 Sample

We targeted top managers, assistant, managers, co-assistants and experts ofprivate and Governmental banks in Iran. Co-assistants are those people who have high responsibility after managers in a certain department in a bank. In this study Experts are those people who have more than 10 years work experience and works sometime as managers' arm in implementing their decision but they have lower position than managers and higher than other typical experts. Qualified persons were introduced to us byresearch and development department in each bank. In this study 4 private banks and2 Governmental banks were chosen. Banks were chosen according to theiraccessibility and willingness of corporation.

3.2 Data Collection

We used a self-administered questionnaire method for collecting primary data.More importantly we were replicating a study that had been done in United state with(Grandon et al. 2004)'s questionnaire. Hence in this research we also replicated thesame questionnaire. First, we deduced the questionnaire from the tables that werepresented in (Grandon et al. 2004) paper. The questionnaire then was translated intoFarsi. A covering letter was also provided for the first page. Two hundred questionnaires were distributed to respondents and One hundred and sixtyindividuals completed the survey for a response rate of 80%. To make such highresponse rate we were delivering questionnaires individually by making anappointment.

3.3 Questionnaire design

The questionnaire is designed to poll the opinion of bankers with respect to:

1- Their perception of e-commerce adoption in their banks as a strategic value.

2- Their attitude towards factors affecting e-commerce adoption in their banks.

Respondents were required to complete the survey that had the following majorsections (see Appendix A).

• Eight demographic questions (respondent’s gender, age, education, yearsof work in present position, years of work in present firm, department,position in the department, and department responsibility ).
• Two questions about the technology in the organization (presence ofweb site, and utilization of e-commerce).
• Sixteen questions asking the extent to which e-commerce is perceivedas contributing to strategic value.
• Twenty-nine questions to measure the factors involved in e-commerceadoption.

A seven-point Likert scale (from strongly disagree to strongly agree) was utilized tomeasure the questions about perceived strategic value and adoption of e-commerce.

3.4 Pilot testing

As this questionnaire was verified and used before, a pilot was not required.

However as it was translated into Farsi, we sent it out to 25 people, all expertsand specialized in IT, EC and Banking to make certain that no problem was raisedduring the translation process. Having the pilot testing completed and feed backreceived, some changes were made through the questionnaire. We also added sevenquestions according to the responders’ feedback:

1. E-banking should support linkage with other network suppliers
2. Havingskillful human resource is an important factor in our decisionto adopt e-banking
3. Having require under structure is an important factor in our decisionto adopt e-banking
4. Adopting e-banking is depends on infrastructure of organizationalprocess
5. Having legal infrastructure readiness is an important factor in ourdecision to adopt e-banking
6. Having telecommunication infrastructure readiness is an importantfactor in our decision to adopt e-banking
7. Having technical infrastructure readiness is an important factor in ourdecision to adapt e-banking

3.5 Reliability and validity

As we dispensed the questionnaires directly or indirectly to the qualified persons andhad a chance to be with them while they were filling out the questionnaires or a contactnumber to call them after distribution we really did not face the subject error. For reducing thesubject bias we tried to make sure that their answers were consideredconfidential. Since the questionnaire was designed in a survey format we did not face withobserver error or the observer bias.Table 1 shows that alpha values range from 0.73 to 0.75 for the perceived strategic value and0.7 to 0.90 for the adoption of e-banking factors. “The scale reliabilities are unusually good compared to the acceptable 0.7 level for field research(Nunnally, 1978)”.

Table 1 reliability analysis, alpha test

Variables / Reliability
Organizational Support (OS) / 0.75
Managerial Productivity (MP) / 0.73
Decision Aid (DA) / 0.74
Organizational readiness (OR) / 0.74
Compatibility (CC) / 0.81
External Pressure (EP) / 0.70
Ease of Use (EU) / 0.84
Perceived Usefulness (PU) / 0.90

4. Results

4.1. Demographics and descriptive statistics

The 160 surveys were returned over an 8-week period. Results indicated that theDMU's of banks were well educated, with over 50% holding a bachelor degree or a master'sdegree. The majority were maleand between 20 and 40 years of age.Table 2 shows other demographics.

Table 2: demographics graphics of responders

4.2. Statistical analysis

The instrument used in this study was adopted from Grandon et al's (2004) study. In order to test the model, a statistical analysis was conducted in two stages. The firststep employed factor analysis to measure whether the number of factors and loadings of itemsinvolved in the two main constructs (perceived strategic value and adoption) conform to theproposed model. Since we were also interested in exploring how the perceptions of strategicvalue influence the decision to adopt e-commerce, canonical analysis was utilized in thesecond step. This technique involves developing a linear combination of independent variables (strategic value variables) and dependent variables (adoption variables) to maximizethe correlation between the two sets (J.F. Hair, et al. 1998). This method was alsoconducted in the (Grandon et al.)'s study to investigate the causal relationship betweenvariables.

4.3. Factor analysis

4.3.1. Perceived strategic value construct

A factor analysis was run using SPSS 13. The factor analysis used principalcomponents in order to extract the maximum variance from the items. To minimize thenumber of items that have high loading on any given factor, a varimax rotation wasutilized. All items measuring the perception of strategic value of e-commerce wereconsidered during the first run and by using the Kaiser Eigenvalues criterion over 1 (see table 3); weextracted three factors that collectively explained 66.13% of the variance in all itemsin one factor, organizational support. Hence, Organizational Support was broke upinto three main factors, Performance support, Operational Support and RelationshipSupport respectively, that each of them can be considered to be a value of e-banking.Table 4 shows rotated component matrix of these three factors:

Table 3: KMO and Bartlett's test

Table 4: components' loadings

(Organizational Support)

component
Factor 1 / Factor 2 / Factor 3
OS2 / 0.861 / .114 / .081
OS3 / 0.823 / .151 / .104
OS1 / 0.554 / .128 / .154
OS8 / 0.209 / 0.805 / -.038
OS4 / 0.049 / 0.771 / .146
OS5 / 0.295 / 0.600 / .427
OS6 / 0.128 / .228 / 0.849
OS7 / 0.145 / -.002 / 0.849

Managerial productivity and Strategic Decision aids' item were both coveringmarginally 60% of the cumulative variance of all items and remained the same asbefore.Therefore the model of the Perceived Strategic Value factors of e-banking would beas shown in figure 2, in this figure also each factors loading has been shown:

Figure 2: PSV revised model after factor analysis

4.3.2. Adoption construct

As well as Perception variables, Adoption variables were also analyzed usingprincipal component factor analysis. By using the Kaiser Eigenvalues criterion, weextracted two factors that collectively explained 63.73% of the variance in all items inOrganizational Readiness factor. Hence, Organizational Readiness was broke up intotwo main factors, Organizational readiness and Infrastructural Readinessrespectively, that each of them can be considered to be influencing in e-bankingadoption (Table 5).

Table 5: components' loadings

(Organizational Readiness)

component
Factor 1 / Factor 2
OR4 / .884 / -.040
OR2 / .844 / .069
OR3 / .744 / .140
OR1 / .620 / .202
OR6 / .112 / .871
OR5 / .145 / .870
OR7 / .043 / .644

Using the Kaiser Eigenvalues criterion External Pressure was also broke up into threemain factors, Intangible Pressure, External Dependency and Persuasive Pressure thatcorrespondingly explained 71% of the cumulative variance of all items (Table 6).

Table 6: components' loadings

(External Pressure)

component
Factor 1 / Factor 2 / Factor 3
EP5 / 0.837 / 0.221 / -0.053
EP6 / 0.834 / -0.056 / 0.102
EP1 / 0.001 / 0.893 / -0.055
EP4 / 0.170 / 0.633 / 0.424
EP2 / -0.035 / 0.088 / 0.846
EP3 / 0.565 / 0.008 / 0.568

Other factors in adoption construct, Compatibility, Perceived Ease of Use andPerceived Usefulness, were marginally covering 60%, 62 and 68% of the cumulativevariance of all items in each factor respectively.Therefore the model of factors affecting e-banking in Iran with related loading wouldbe as shown in figure 3:

Figure 3: Adoption revised model after factor analysis

4.4. Canonical analysis

Considering the fact that perception was independent variable in the Grandonand Pearson's model and adoption was the dependent one, we had to consider theimpact of perception factors on adoption factors and among several statistical toolsand techniques the canonical analysis was the most suitable ones.The “Canonical analysis is a multivariate statistical model that studies theinterrelationships among sets of multiple dependent variables and multipleindependent variables. By simultaneously considering both, it is possible to controlfor moderator or suppressor effects that may exists among various dependentvariables (Mahmood et al. 1993)”. In canonical analysis there are criterion variables(dependent variables) and predictor variables (independent variables). “The maximumnumber of canonical correlations (functions) between these two sets of variables is thenumber of variables in the smaller set (Green et al. 1966)”.In our case, the number of variables for the perception of strategic value construct isfive while the number of variables in the adoption construct is eight (figure 4). Thus,the number of canonical functions extracted from the analysis is five; i.e., the smallestset. Canonical correlation is a measure to interpret the canonical functions. Canonicalcorrelation size also gives the measures of overall model fit given.

Figure 4: the revised model after factor analysis

In order to answer research question (Is there any association between factors of perceived strategic value and adoption?) and to investigate the causal relationship between two set of variables and factors in the model, we uses canonical correlation running STATISTICA 5.5A.

In order to test the significance of the canonical functions we followed the same path in Grandon and Pearson's canonical analysis section that was also followed the guidelines given by Hair et al. They suggest three different measures to interpret the canonical functions:

(a)The significance of the F-value given by Roy’s gcr (see table 7)

(b)The measures of overall model fit given by the size of the canonical correlations (see table 8); and

(c)The redundancy measure of shared variance (see table 9).

Table 7: multivariate test of significance

Note that the strength of the relationship between the canonical covariates is given bythe canonical correlation (Grandon et al. 2004).Even though the multivariate test of significance shows that the canonical functions,taken collectively, are statistically significant at the 0.01 level, from the overall modelfit (Table 8) it can be concluded that only the first canonical function is significant (P< 0.01). This conclusion is consistent with the canonical R² values showed in Table 8.

Table 8: Measures of overall model fit