Electronic Commerce Research

, Volume 12, Issue 2, pp 151–175 | Cite as

The mediating role of the dimensions of the perceived risk in the effect of customers’ awareness on the adoption of Internet banking in Iran

Article

Abstract

One of the major issues banks are faced with in providing Internet Banking (IB) services is the adoption of these services by the customers. This study seeks answer to the question that whether bank customers’ awareness of the services and advantages of IB is effective in reducing the negative effect of customers’ perceived risk on their intention of IB adoption. To this end, the two constructs of the dimensions of the perceived risk and IB awareness are simultaneously considered. Besides, in the research model, the effect of IB awareness on each dimension of the perceived risk and the effect of these dimensions on intention of IB adoption by the customers are investigated. The results indicate that IB awareness acts as a factor reducing all dimensions of the perceived risk (including time, financial, performance, social, security, and privacy). In addition, it was found out that except for social risk, other dimensions of the perceived risk have significantly negative effect on the intention of IB adoption. Finally, proving the direct and positive effect of IB awareness on adoption intention, it was concluded that the dimensions of customers’ perceived risk plays a mediating role in the positive effect of IB awareness on IB adoption intention. In this respect, management approaches centered on the concept of IB awareness are offered for reducing the dimensions of customers’ perceived risk.

Keywords

Internet banking Adoption Dimensions of the perceived risk Awareness 

1 Introduction

Development of information technology and emergence of various forms of Internet banking in recent years have substantially changed the ways banks communicate with their customers [28, 71].

Internet Banking (IB) is one of the branches of electronic banking in which the customers have the opportunity of using a broad range of services such as cash transfer, paying the bills, getting information on deposits, investment and cheque services through Internet and the website designed by the banks [64].

IB offers many advantages to both banks and their customers. Regarding the advantages of IB for the banks, one can point to release from temporal and spatial limitations, reducing operational and administrative expenses, awareness from the activities of customers, and creating the potential for expanding the range of services [7, 83]. In addition, IB provides the customers with the possibility of more rapidly conducting a broad range of financial transactions electronically, through the bank website, at any time and place and with less handling fees in comparison to other bank methods [7, 48].

However, despite the increasing number of Internet users and all IB advancements, the number of IB users has not yet increased as expected [82].

One of the factors affecting the adoption of online shopping among customers is the concept of the perceived risk [80]. Researchers of customer behavior have often defined perceived risk as the customers’ perception about lack of trust and the potential adverse effects of purchasing a good or service [55]. Many studies have revealed that customers perceive different dimensions of the risk. The predictive value of each dimension in the total risk and its reductive behavior greatly depend upon the class of the good or service [32]. If online shopping is considered as a short-term relationship between the buyer and seller, the dimensions of the perceived risk in IB, which is a long-term relationship, will be observed in a more complicated form requiring more accurate investigation.

When studying the concept of customers’ perceived risk in various fields, it should be noted that different forms of risk may be perceived independently from each other and their impact may be different, because each of them can be caused by different sources and in a different condition [56]. Although the concept of the perceived risk, as a barrier of IB adoption, has been investigated in relatively many studies; most studies, except for a limited number (e.g. [52, 55]), provide two main forms. First, the concept of the perceived risk has been considered as a single construct rather than a collection of dimensions (e.g. [83, 84]). Second, in most of these studies, the concept of the perceived risk has been regarded as equivalent to one or more specific dimensions from among the dimensions of the perceived risk, such as security and privacy [20, 60]. In this respect, the first aim of this study is to investigate the effects of each dimension of the perceived risk on IB adoption, separately.

Littler and Melanthiou [55] argue that research on the concept of the perceived risk is conducted based on the assumption that customers can have rational evaluations from the effects and probabilities of these events. To put it more simply, it is assumed that customers have the necessary awareness of the various angles of a good or service and obtain ensuring information. However, if the perceived risk is viewed as a collection of dimensions rather than a single construct, the question arises that how much does the concept of awareness affect each dimension of the perceived risk in IB adoption? Finding answer to this question is the second goal of this study.

The third and the main objective of the present study is finding the answer to the question that regarding the results obtained from the first and second objectives, is it possible to consider a mediating role for each dimension of the perceived risk in the relationship between IB awareness and IB adoption. The answer to this question basically specifies two very important concepts that can be regarded as the main contribution of the present study. First, considering the concept of the perceived risk as a set of dimensions, it becomes clear that by mediating which dimension of the perceived risk, IB awareness increases IB adoption. Second, it is determined that which dimension of the perceived risk, plays a higher mediating role in the positive effect of IB awareness on IB adoption intention. Accordingly, the answer to this question can greatly help bank managers and planners in developing marketing strategies using the role of awareness and information on IB. In other words, they will find out that they can increase IB adoption by their customers through focusing more on which aspects of the products, facilities and their advantages.

With respect to the knowledge acquired from the literature on IB adoption, it can be claimed that simultaneously investigating the two concepts of the dimensions of the perceived risk and IB awareness and their relationship in this study is unique.

Findings of Gholami et al. [34] indicated that the changes caused by information and communication technology in less-developed countries have stronger impacts upon human development index scores in comparison to more developed countries. Al-Somali et al. [5] argued that despite the conduction of many studies on IB adoption in various countries of world, few studies can be found that investigate this issue in developing countries of the Middle East. The present article investigates the research model in Iran in order to achieve its objectives. Iran is a developing country in the Middle East. The population of this country in 2009 is about 66.5 million people and the number of Internet users is estimated to be 32.2 million people in this country [43]. According to the report of Iran central Bank [19], the number of IB services users at the end of the first quarter of 2009 was around 5.8 million people (i.e. 8.7 % of the population) who receive these services from 11 governmental and 6 private banks. Comparison of this statistics with the statistics in a developed country like England can be interesting. England with a population of 61.5 million people in 2009 had 48.75 million Internet users [42] and according to the report of Association of Payment Clearing [8], the number of IB users in the first half of 2009 was over 22 million people (i.e. 35.8 % of the population). Comparison of these two statistics clearly indicates that Iran is back-warded with respect to the application of IB. The problem of IB adoption by the customers in Iran is the main problem banks face in expanding these services. Therefore, the findings of this study, besides its theoretical contributions, can help managers of Iranian banks to develop their marketing strategies for more rapidly expanding IB adoption by customers.

This paper is organized as following: Sect. 2 introduces the theoretical background of the study including studies on Internet banking, hypotheses, and research model. Section 3 describes research methodology and Sect. 4 presents data analysis and results of hypotheses testing. Section 5 investigates research findings and offers their theoretical and managerial implications. Finally, Sect. 6 presents the conclusion and offers suggestions for further research in future.

2 Theoretical background

The theoretical framework of this paper consists of four parts. The first part reviews the related literature on IB and models employed by the studies on IB adoption. In the second part, the dimensions of the perceived risk in IB adoption is investigated on the basis of the perceived risk theory (PRT) and the related hypotheses are presented. The third part explains the concept of IB awareness and the related hypotheses. Finally, in the fourth part, the research model is proposed on the basis of the hypotheses presented in parts 2 and 3.

2.1 Studies on Internet banking

Since the mid 1990s, a radical change has been witnessed in the service channels of the banks introduced by new technologies like Internet banking [64]. This fundamental change led to a new literature in research in the field of banking technology. Akinci et al. [2] point out that four interrelated fields can be identified by reviewing this literature:
  • Retail banking services

  • Structure of bank distribution channels

  • Banks, bank managers, and their attitudes, perceptions, orientation and strategies toward new technologies like IB

  • Customer characteristics including demographical characteristics, attitudes, intentions, adoption, and satisfaction

The field considered in this study is part of the fourth filed, i.e., IB adoption by the customers.
Hernandez and Mazzon [38] categorized IB adoption studies into two categories:
  • Descriptive studies: the aim of which is to identify characteristics, attitudes, reactions, adoption barriers, and features making IB adoption seem attractive to the customers (e.g. [2, 33, 47]).

  • Relational studies: whose aim is to determine variables affecting IB adoption using one of the models for the adoption of new technologies or a combination of them.

In this study, it is tried to look into the issue of IB adoption on the basis of the second approach.
The basic theoretical approaches and models used in the relational studies conducted in the field of adoption of new technologies like IB can be summarized as follows:
  1. 1.

    Theory of Reasoned Action (TRA) [30]

     
  2. 2.

    Diffusion of Innovation (DOI) [67]

     
  3. 3.

    Theory of Planned Behavior (TPB) [1]

     
  4. 4.

    Social Cognitive Theory (SCT) [10]

     
  5. 5.

    Technology Acceptance Model (TAM) [25]

     
  6. 6.

    Commitment-Trust Theory (CTT) [58]

     
  7. 7.

    Perceived Risk Theory (PRT) [45, 49, 69]

     

Many researchers have also tried to use, develop, and combine these models and theories in order to investigate the issue of adoption of new technologies like IB (DTPB: Taylor and Todd [77]; TAM2: Venkatesh and Davis [78]; UTAUT: Venkatesh et al. [79]; Pikkarainen et al. [64]; Chan and Lu [20]; Lai et al. [51]; Ndubisi [61]; Yiu et al. [83]; Zolait and Ainin [86]; Al-Somali et al. [5]; Lee [52]; Polasik and Wisniewski [65]).

The theory used as the basic concept of this study is the perceived risk theory (PRT). The main feature of this theory distinguishing it from other theories and models is the fact that, unlike other theories or models which concentrate on the positive factors influencing customers’ adoption of IB, this theory specifically focuses on the negative factors (risks) which prevent customers from IB adoption. In the next section, this theory is introduced in the field of IB.

2.2 The perceived risk and its dimensions

As it was mentioned earlier, the perceived risk can be defined as the customer’s perception of lack of trust and the potential adverse effects of purchasing a good or service [55]. The perceived risk is considered as an important factor which influences customer behavior [63].

Researchers in the field of as perceived risk theory (PRT) [29] identified perceived risk as the combination of several dimensions [45, 49, 69]. These dimensions include performance, financial, social, psychological, security, privacy, and physical risks. These dimensions have been used and even developed by many researchers (for more information refer to Lim [54]).

Gemünden [32] argued that the predictive value of each dimension in the total risk and its reductive behavior greatly depend upon the class of the good or service. Since the first objective of this study is to investigate the effects of the dimensions of the perceived risk on IB adoption, the literature in this area is reviewed below.

Studies which, based on PRT, have investigated the dimensions of the perceived risk specifically in IB adoption are limited to four studies the details of which are presented in Table 1.
Table 1

Studies specifically investigating the dimensions of the perceived risk in IB adoption

Prior studies

Year

2006

2008

2009

2009

Author(s)

Littler and Melanthiou

Zhao et al.

Lee

Aldas-Manzano et al.

Perceived risk dimensions

Financial

S

X

S

 

Performance

S

X

S

S

Social

X

X

X

S

Time

S

X

S

X

Security

S

X

S

S

Privacy

 

X

S

Psychological

X

X

  

Physical

 

X

  

Note 1: X: Dimensions included in studies. S: Dimensions found to be significant in the studies

Note 2: In Zhao et al. [85] in the content of Chinese culture, none of the dimensions were not identified as important on the basis of the classification of the Table and another classification which is combination of the dimensions under study has been obtained

Note 3: In Lee [52] the security and privacy risks have been introduced as one single security risk

It can be inferred from Table 1 that some concepts of these studies overlap each other. According to the results of these studies and by summarizing the concepts of the dimensions of the perceived risk in them, the dimensions of the perceived risk can be categorized into six groups: time, financial, performance, social, security, and privacy. Considering the general definition of each dimension, provided by Featherman and Pavlou [29] and Lee [52], in the following, each of these dimensions is defined in terms of IB.

Time risk

This risk refers to customers’ concerns about following issues: 1. Too much time spent for learning how to use IB; 2. Too much time devoted for solving problems caused by using IB (such as proving transaction errors), and 3. Too much time which must be spent doing and completing transactions in IB.

Financial risk

This risk refers customers’ concerns about the potential of financial loss which is caused by: 1. Mistake in entering the specifications of transaction such as account number or amount of money by the customer in IB, 2. Impossibility of compensation by bank in case of errors in transaction, and 3. Losing control of personal account resulting in financial loss caused by IB.

Performance risk

This risk refers to factors which may influence efficiency of IB in the consumer’s eyes in the following issues: 1. Lack of system’s good performance due to low download speed, server pauses, or website maintenance operations, and 2. Not meeting customers’ expectations from IB on the basis of the advertised preferences after use.

Social risk

This risk refers to customers’ concerns in the following areas: 1. Negative attitude of family, friends, colleagues toward IB, and losing social position among these groups in case of errors or frauds, and 2. Impossibility of direct contact with bank staff and their help in using IB (see also Al-Somali et al. [5]).

Security risk

Concern about the following issues constitute security risk of IB customers: 1. Lack of security of World Wide Web for sending and receiving financial information (Internet security), and 2. Potential loss resulting from fraud and IB hack threatening the security of the customers (IB security) (see also Sathye, [70]; Aladwani, [3]; Suh and Han, [76]; Westland, [81]; Polasik and Wisniewski, [65]).

Privacy risk

This risk refers to the fact that IB users desire to have control over all aspects of collecting their personal data [4]. Now if users’ private data (privacy) is collected and registered without their awareness, this may worry them [18, 26]. This concern can be referred to as privacy risk.

Lee’s [52] reasoning that IB creates no threat to human life, has excluded physical risk from the dimensions of the perceived risk in IB adoption. In this respect and considering the studies of Littler and Melanthiou [55] and Zhao et al. [85] which have not identified psychological risk as one dimension of the risk affecting IB adoption, physical and psychological risk have been excluded from the list of dimensions of the perceived risk in IB adoption.

Beliefs about the results of behavior, including the perceived risk, are among the main components and bases of the attitude toward behavior [46]. Regarding adoption of technologies like IB, Venkatesh et al. [79] argue that since technology adoption is optional, attitude and intention are positively related to each other. Therefore, dimensions of the perceived risk in IB adoption can influence customers’ perception about IB adoption and negatively affect their intention of adopting such technology [83]. Considering these discussions, the following hypotheses can be put forward:
H1:

customers’ perceived risk of IB (1—Time, 2—Financial, 3—Performance, 4—Social, 5—Security, and 6—Privacy) negatively influences their intention of IB adoption.

In the above hypothesis, time risk is considered as H1-1, financial risk as H1-2, performance risk as H1-3, social risk as H1-4, security risk as H1-5, and privacy risk as H1-6.

2.3 Awareness from Internet banking

Howcroft et al. [40] concluded in their studies that one of the most important reasons of customers’ reluctance for adopting IB is their unawareness of its services and advantages. Moreover, Sathye [70] notes that low degree of awareness of Internet banking is a critical factor in causing customers not to adopt Internet banking. Azouzi [9] concluded in his study that awareness of its advantages and services has a significant positive effect on adopting and using Internet banking.

This issue has also been confirmed by Gerrard et al. [33]; Al-Somali et al. [5]. Pikkarainen et al. [64] also comment that the volume of information customers receive about IB is recognized as the main influential factor in adopting this service. Information of customers about service, facilities, advantages, and way of using IB, can be regarded as IB awareness. In this regard, the following hypothesis can be presented:
H2:

IB awareness positively influences bank customers’ intention of IB adoption.

Rogers and Shoemaker’s [68] theory states that before customers become ready to adopt a product or service, they pass through the process of knowledge, persuasion, decision, and confirmation. In other words, they argue that acceptance or rejection of an innovation commences when the customers become aware of the product and its advantages and disadvantages. Thus, customer awareness of the product, its facilities, advantages, and disadvantages is among the initial and important stages in determining innovation of the individual. On the other hand, Aldas Manzano and Navarre [4] found out that customer innovation negatively affects IB risk perception. Besides, Cooper [24] considers risk level as an important factor of innovation adoption by the customer. In this respect, it can be concluded that customer awareness of IB, as an important factor of innovation, can exert a negative effect on customers’ risk regarding IB. Littler and Melanthiou [55], confirming this issue, argue that one of the mediating factors which affect risk perception of the customers is insufficient information about the products and its advantages and disadvantages. Lichtenstein and Williamson [53], also, point to the significance of knowledge and support in informing customers for reducing the risk and increasing their willingness for receiving IB services.

With respect to the previous hypotheses in which the negative effect of the dimensions of the perceived risk and the positive effect of IB awareness on IB adoption intention are considered, it seems logical to consider a negative direction for the effect of IB awareness on the dimensions of the perceived risk. Consequently, the following hypotheses are put forward:
H3:

IB awareness negatively influences the dimensions of the perceived risk for IB adoption (including 1—Time, 2—Financial, 3—Performance, 4—Social, 5—Security, 6—Privacy).

In the above hypothesis, the effect of awareness on time risk is considered as H3-1, financial risk as H3-2, performance risk as H3-3, social risk as H3-4, security risk as H3-5, and privacy risk as H3-6.

2.4 Research model

The research model of present study which is designed on the basis of the hypotheses presented in the previous section is depicted in Fig. 1.
Fig. 1

The proposed research model

3 Research methodology

3.1 Survey administration

The survey method was used for collecting data in order to test the hypotheses. Considering the objectives mentioned in previous sections, the statistical population of the study is composed of Iranian bank customers who do not actively use IB. Three points must be mentioned about the selection of statistical population. First, considering the model and research objectives, the aim is to investigate the dimensions of the perceived risk in IB adoption intention and the effect of IB awareness on the dimensions of the perceived risk before IB adoption. Second, those who actively use IB cannot have a true conception of their perceptions before IB adoption. The reason is the changes of the content of the dimensions of the perceived risk and degree of IB awareness after adoption. Third, as it was mentioned, the statistical population consists of the customers who do not actively use IB. Hence, those who have IB account but do not use it due to various reasons part of which being the dimensions of the perceived risk, are considered in the population. Regarding the defined statistical population, the statistical sample was also selected from among Iranian bank customers who do not actively use IB.

Two methods of self-administered survey and Internet survey were utilized for data collection. This was done in order to reduce the amount of possible bias. In this method, first, five regions in the northern, southern, central, eastern, and western parts of Iran were selected and random sampling was conducted in them. This sampling was done in places like trains (with the permission of Iranian railway organization), commercial centers, governmental organizations, private companies, industrial factories, and universities, and lasted for one month in December 2009. In sum, 462 questionnaires were collected and after removing uncompleted questionnaires 414 completed questionnaires were obtained. In the second method, Internet survey, an online questionnaire was developed, first. In the next stage, the address link of this questionnaire was placed on the home page of two well-known websites about management and banking (www.betsa.ir, www.banki.ir). Also, in order to increase the reply rate, the questionnaire was advertised in three forums. Finally, using this method and during one month, 140 replies were obtained. Thus, a sample of 554 was achieved.

Table 2 presents the demographical features of the respondents. 54 % of the respondents were male and the highest ratio (41.5 %) with respect to age belonged to (26–35) group. Most of the respondents, that is, 36.6 % had bachelor’s degree. 57.8 % of the respondents regularly used Internet and the highest ratio (33.4 %) regarding income belonged to the income group of 250000 to 500000 Tomans. In addition, Table 3 presents the amount of respondents’ use of other forms of banking services in which the average to high usage of bank branches and ATMs, and not using Telephone bank, as well as SMS banking is observable.
Table 2

Demographical features of the respondents

Table 3

Respondents’ use of other banking forms

Usage

Bank branch

Telephone bank

SMS bank

ATM

Frequency

%

Frequency

%

Frequency

%

Frequency

%

No use

33

6.0

331

59.7

435

78.5

22

4.0

Very low

60

10.8

62

11.2

39

7.0

10

1.8

Low

122

22.0

71

12.8

22

4.0

31

5.6

Average

192

34.7

52

9.4

38

6.9

94

17.0

High

93

16.8

27

4.9

12

2.2

167

30.1

Very high

54

9.7

11

2.0

8

1.4

230

41.5

3.2 Measurement development

Measurement items used in this study were extracted from previously validated measures based on the literature review. In order to evaluate the opinions of respondents, a 5-point likert scale ranging from “completely disagree” to “completely agree” was used. For ensuring the reliability of questionnaire content, a pre-test was also administered in a random sample of 10 individuals and based on its results awa4 item which was the only item with negative meaning in AWA construct, was removed from the questionnaire. This was also confirmed by Schriesheim and Hill [72] and Jackson et al. [44]. They argue that measures with adverse scores should not be used in the questionnaire for they reduce the validity of questionnaire items and increase the possibility of systematic error. The items of final questionnaire used for measuring the construct are shown in Table 4.
Table 4

Items used in the questionnaire

Construct

Items

References

Time Risk (TIM)

tim1

In my opinion, learning how to use Internet banking services takes a lot of time

[4, 52, 55, 85]

tim2

A lot of time must be spent for transactions in IB services for banking transactions

tim3

I think I should spend a lot of time solving the problems arising from IB systems (such as proving the payment errors)

Financial Risk (FIN)

fin1

I am afraid of losing my money when transferring money on the Internet due to carelessness or mistakes like erroneous entry of account number or the amount of money

[52, 85]

fin2

I am afraid that I will not be able to get compensation from bank in case of errors

fin3

I am afraid of losing control of my account by using IB systems

Performance Risk (PER)

per1

In my opinion IB systems may not work properly due to low download speed or maintenance operations or they may face server pauses

[4, 52, 55]

per2

IB servers may not work properly and the process of payment may be wrong

per3

I am afraid that IB might not provide the advantages advertised

Social Risk (SOC)

soc1

I am sure that if I decide to use IB and mistake or fraud happen in my Internet banking transactions, I will lose my good position among my friends, family, and colleagues

[55, 85]

soc2

I think if I use IB systems, people will not admire me for using it

soc3

Using IB systems, I will not be able to have direct relations with bank staff and use their helps and this gives me an unpleasant feeling

Security Risk (SEC)

sec1

I feel unsecured about sending and receiving my financial information on IB systems

[4, 52, 55, 85]

sec2

I think IB systems can easily be accessed by unauthorized people like hackers

sec3

In my opinion, World Wide Web is not a safe and appropriate place for financial transactions

Privacy Risk (PRI)

pri1

Use of Internet for financial transactions increases the possibility of unwanted emails

[4, 85]

pri2

I think if I use IB systems, it might be possible for bank to make my personal information accessible for other organizations or companies without my consent

pri3

I think if I use IB systems, my privacy is threatened due to illegal use of my personal information

Awareness (AWA)

awa1

I think I get enough information about the services of IB systems

[5, 64]

awa2

I think I get enough information about the advantages of IB systems

awa3

I think I get enough information about the ways of opening account and using IB systems

awa5

In general, I have enough information about IB

Intention to use (INT)

int1

I intend to use IB regularly in future

[5, 52, 66, 83]

int2

I intend to use IB for quick and easy access to my bank information in future

int3

I am going to use IB for my bank transactions in future

int4

I think that I will use IB more than bank branches in future

Note: Item awa4 has been removed from the questionnaire

Cook et al. [23] commented that the least number of items required for achieving internal reliability is three items on each construct. This limitation has been considered in this questionnaire.

4 Analysis

The two-stage process suggested by Anderson and Gerbing [6] was utilized for data analysis. In the first stage, the measurement model is analyzed and in the second, the structural relations among the hidden constructs are investigated. In the first stage, it is considered that the collection of indicators of each construct must uniquely measure the construct related to them. That is to say, in this stage it is evaluated whether indicators provided for presenting each construct really indicate them or not. The precision with which indicators indicate their constructs is also reported. The second stage of the model is investigated when the set of constructs is employed for measuring the multi-dimensional construct. In this stage, it is assessed whether the constructs are good measures for the multi-dimensional construct or not and a report of the precision of the construct in introducing multi-dimensional construct is presented. Besides, based on the results obtained from this stage, the research hypotheses are tested.

4.1 Analysis of measurement model

In order to test the measurement model and establish discriminant and convergent validity of the constructs, Confirmatory Factor Analysis (CFA) was conducted using LISREL 8.54. The following indices were considered for evaluating the measurement model.
  • All factor loadings must be meaningful and greater than 0.4 [39].

  • All reliability measures must be greater than 0.7 [35, 62].

  • Convergent Validity: constructs must have at least a 0.5 the Average Variance Extracted (AVE) [31].

  • Discriminant Validity: The AVE of the construct must be greater than the variance shared between a particular construct and other construct in the model [52].

It must be mentioned that the convergent and discriminant validity are different forms of construct validity. Convergent validity shows high correlation of indicators of a construct in comparison to correlation of the indicators of another construct [74]. Discriminant validity indicates minor correlation among the indicators of a construct and those of another construct [37].

While construct validity refers to the measurement among constructs, reliability is the outcome of measurement within a construct. That is to say, the indicators selected for a given construct are considered together to identify the erroneous indicators of the construct. In fact, reliability is used to investigate the coordination of indicators with the constructs with which they must be measured [75].

In the results obtained which are presented in Tables 5 and 6, all these recommended levels have been observed.
Table 5

Validation of the final measurement model, reliability and convergent validity

Construct

Item

Factor Loading

T-Value

CA

CR

AVE

Time Risk (TIM)

tim1

0.64

14.06

0.74

0.75

0.51

tim2

0.84

17.69

tim3

0.64

14.08

Financial Risk (FIN)

fin1

0.85

24.24

0.91

0.91

0.78

fin2

0.90

26.27

fin3

0.90

26.64

Performance Risk (PER)

per1

0.81

21.45

0.86

0.86

0.67

per2

0.89

24.24

per3

0.74

19.17

Social Risk (SOC)

soc1

0.72

15.98

0.75

0.76

0.52

soc2

0.82

17.85

soc3

0.61

13.70

Security Risk (SEC)

sec1

0.78

21.14

0.89

0.90

0.74

sec2

0.86

24.34

sec3

0.94

27.52

Privacy Risk (PRI)

pri1

0.91

26.77

0.92

0.92

0.79

pri2

0.90

26.37

pri3

0.85

24.41

Awareness (AWA)

awa1

0.92

27.31

0.90

0.90

0.70

awa2

0.88

25.38

awa3

0.82

23.09

awa5

0.71

18.80

Intention to use (INT)

int1

0.87

12.57

0.93

0.93

0.77

int2

0.86

12.95

int3

0.89

11.71

int4

0.89

11.87

Recommendation

0.70

0.70

0.50

Notes: CA=Cronbach’s Alpha; CR=Composite Reliability; AVE=Average Variance Extracted; p<0.001

Table 6

Discriminant validity of constructs

Construct

INT

TIM

FIN

PER

SOC

SEC

PRI

AWA

INT

0.86

       

TIM

−0.43

0.71

      

FIN

−0.28

0.17

0.88

     

PER

−0.35

0.21

0.29

0.82

    

SOC

−0.20

0.12

0.17

0.21

0.72

   

SEC

−0.26

0.17

0.23

0.29

0.17

0.86

  

PRI

−0.19

0.12

0.16

0.20

0.12

0.16

0.89

 

AWA

0.38

−0.35

−0.48

−0.60

−0.35

−0.48

−0.34

0.82

Note: Diagonal elements (in bold) are the square root of average variance extracted (AVE). Off-diagonal elements are the correlations among constructs. For discriminant validity, diagonal elements should be larger than the off-diagonal elements

4.2 Analysis of the structural model

In order to analyze the structural model proposed, structural equation modeling (SEM) was utilized and its fitness was evaluated through Chi-Square test. When the sample size is between 75 and 200, the Chi-Square value of a rational index shows the fitness. However, for models with larger n, the value of Chi-Square is always statistically significant. (Of course this is not totally agreed upon by the researchers and its rejection or acceptance is debated. For more information please refer to Barrett [12]. In addition, Chi-Square is influenced by the amount of correlations in the model. The higher these correlations, the weaker the fitness will be [16, 50]. In this regard, other fitness indices are considered in this study, while being based on Chi-Square, the effect of sample size has been modified in them. In general, it must be stated that these fitness indices are used for measuring the fitness if data with the given structural model. The values obtained for these indices and their recommended levels are indicated in Table 7. These indices generally show that the model has acceptable fitness.
Table 7

Goodness of fit indices for structural model

Fit indices

Structural model

Recommended value

References

NFI

0.96

>0.9

[14, 21]

NNFI

0.97

>0.9

[21]

CFI

0.98

>0.9

[13, 36]

GFI

0.94

>0.9

[21, 73]

AGFI

0.92

>0.8

[21, 73]

RMSEA

0.05

<0.06

[41]

χ2/df

2.50

<5

[15, 57]

<3

[21, 73]

4.3 Hypothesis testing

The results of SEM used in data analysis are shown in Table 8. Also, the results of structural model can be observed in Fig. 2. In general, the structural model can be evaluated by two indices. The first one are the path coefficients (β) which show the strength of relations between independent and dependent variables, and the second are the values of R2 which show the values of variances explained by independent variables and reflect the predictive power of the model. As it can be seen from the Table, intention of using IB is significantly influenced by IB awareness and time, performance, security, finical, as well as privacy risks; they, jointly, explain 76.3 % of the total variance in intention. On the one hand, social risk does not significantly influence intention of using IB. Thus, all H1 and H2 hypotheses expect for H1-4 are supported. On the other hand, IB awareness has significant negative influence on all dimensions of the perceived risk. The values of variance percentage explained by awareness in all these aspects are high. Consequently, all hypotheses from H3-1 to H3-6 are supported.
Fig. 2

Results of structural modeling analysis

Table 8

Assessment of the structural model

No.

From

To

β

R2

t-value

p-value

Supported?

H1-1

TIM

INT

−0.54

0.76

−13.54

0.000**

Yes

H1-2

FIN

−0.21

−4.19

0.000**

Yes

H1-3

PER

−0.31

−6.39

0.000**

Yes

H1-4

SOC

−0.05

−1.59

0.11

No

H1-5

SEC

−0.29

−6.14

0.000**

Yes

H1-6

PRI

−0.13

−2.16

0.03*

Yes

H2

AWA

0.30

4.43

0.000**

Yes

H3-1

AWA

TIM

−0.33

0.51

−6.19

0.000**

Yes

H3-2

FIN

−0.47

0.62

−10.41

0.000**

Yes

H3-3

PER

−0.59

0.75

−12.66

0.000**

Yes

H3-4

SOC

−0.34

0.52

−6.52

0.000**

Yes

H3-5

SEC

−0.48

0.63

−10.28

0.000**

Yes

H3-6

PRI

−0.33

0.61

−7.35

0.000**

Yes

Note: **Significance at p<0.001, *Significance at p<0.05

Baron and Kenny [11] argue that it is necessary for both direct and indirect effect to be considered simultaneously for completely testing mediation. To this aim, three concept of total, direct, and indirect effects were taken into account in research model. Except for AWA, all constructs of the research model only have direct effect on INT. The only construct having both direct and indirect effect is AWA. The direct, indirect and total effects of awareness on intention to use of IB were (0.302, ρ<0.01), (0.655, ρ<0.01) and (0.957, ρ<0.01). The indirect effect of AWA on INT shows a stronger effect than the direct effect, exhibiting that dimensions of the perceived risk were also the key mediators to influence INT. This indirect effect of AWA on INT has respectively been through PER (0.182, ρ<0.01), TIM (0.176, ρ<0.01), SEC (0.139, ρ<0.01), FIN (0.101, ρ<0.01), PRI (0.042, ρ<0.05) and SOC (0.015, ρ=0.09). In this regard, performance, time, security, financial, and privacy risk act respectively as mediators in the positive effect of IB awareness on intention to use IB.

5 Discussion

The model investigated in this study examines the effect of IB awareness on each dimension of the perceived risk in IB adoption as well as the effect of each of these dimensions on customers’ intention. It should be remembered that considering the values of R2, the proposed model has a high explanatory power. It was found out in previous section that performance, time, security, financial, and privacy risk act respectively as mediators in the positive effect of IB awareness on intention to use IB. In this section, the results are investigated based on each dimension of the perceived risk.

It was found out that performance risk plays the highest mediating role. This risk is the second risk in terms of importance in intention of use. The concern of customers in this dimension of risk is that IB server does not work as expected and due to server pauses or maintenance and troubleshooting activities, it usually faces problems and they always witness error message when using the server. Besides, they worry about the low download speed of IB website. Also, the customers are afraid that it is possible for the bank not to provide the services and advantages it advertises. Furthermore, it has been proven that IB awareness plays an important and significant role in reducing the performance risk. Regarding the above mentioned concerns, banks must remember to offer and expand IB services on the basis of their timed strategic and operational programs and put emphasis on their advertising process at this point. This emphasis ensures the customer that all technical and non-technical aspects will be observed in offering these services and the possibility of problems like server pauses will be low when they use such services. Administering traffic management and server supporting systems in IB systems and appropriate advertisement can also reduce this risk among customers and increase their willingness to adopt IB.

The second risk plays the most mediating role is time risk. It is the most important risk which influences intention of using IB. Therefore, the most important concern of bank customers is the great amount of time spent for transactions or solving possible problems arising from them. Besides, customers worry that learning how to use IB is difficult and time-consuming. On the other hand, it was proved that awareness from the advantages, services, and ways of using IB plays a significant role in reducing time risk. Therefore, banks must take technical actions to reduce the possibility of delay in payments and transaction times and inform the customers of these activities. Also, necessary guides and instructions about the possible problems arising at the time of use must be prepared and provided to the customers to ensure them that they will not spend much time solving these problems. Establishing and expanding telephone guide centers can also be useful in this regard. As Yiu et al. [83] argue, designing demonstrations and advertisements for their use among bank customers can ensure them that they can easily learn how to use IB.

The security risk is the third risk playing a mediating role. It is the third risk bank customers are faced with in adopting IB. This risk is the customers’ concerns regarding Internet security and security of IB website. Also, it was proved that awareness has a significant role in reducing this risk. Banks are now using various solutions such as firewalls, filtering routers, callback modems, encryption biometrics, smart cards, digital certificates [59], and two-factor authentication systems [65] for creating security in IB systems. But these concepts are not understandable for many customers. Giving information about these issues both in technical and non-technical terms ensures customers that bank is trying to provide the security of IB system in the best way possible. Another strategy regarding awareness that, through creating sense of security, can encourage customers for adopting IB is giving information about various frauds in the area of IB such as phishing [17] and providing the customers with guidelines for protection against these frauds. These cases and third-party trust certification bodies [4, 84] can increase customers’ awareness of security issues and, by reducing this risk in their minds, increase their willingness to adopt IB.

The fourth risk, among the dimensions of the perceived risk, in which IB awareness plays a part, is financial risk. This is the fourth risk negatively affecting intention to use IB. Customers’ concern about impossibility of getting compensation from bank in the case of mistake, possibility of erroneous transactional data entry, and fear of losing control over personal account using IB are among the main concerns of customers in this risk. Also, it was proved that awareness plays a significant part in reducing financial risk. Solutions proposed that can reduce this risk and encourage customers for adopting IB include: 1. Developing and offering guidelines and instructions which explain customer rights and banks responsibilities in the field of IB; 2. Giving information about the articles of electronic commerce law that refer to the validity of electronic documents resulting from IB transactions; 3. Planning, administering, and advertising about consumer reassurance programs such as after sale redress policy; 4. Giving information about the features of possibility of confirming transaction at the same time it is going on which minimizes error possibility; and 5. Expanding IB services which increase user’s authority in managing personal account (such as continuous payment planning and giving appropriate information about it).

The fifth risk which plays mediating role in the relationship between IB awareness and intention of IB adoption is privacy risk. It is the fifth and the last risk influencing intention of IB adoption. Banks’ illegal use of customers’ private information and its results such as receiving unwanted emails or endangering privacy constitute the main concerns of customers in this risk. The role of awareness in reducing this risk was also proved. Bestavros [15] offered recommendations for reducing this risk which were also emphasized by Aldas-Manzano et al. [4]. Considering these recommendations and emphasizing the role of awareness in reducing this risk, the following solutions are offered: 1. Giving information about articles of electronic commerce law that places responsibility of disclosing IB customers’ private information upon bank; 2. Placing the issue of protecting customers’, particularly IB customers, privacy in the quality policy of banks and giving information about it; 3. Giving information and creating trust regarding protection of customers’ privacy at the time of opening account; and 4. Designing and launching IB system in a way which does not require sending email to customers.

On the basis of the results obtained, it became clear that social risk has minor effect on customers’ intention of using IB. This means that customers are not afraid of the negative attitude of their family, friends, or colleagues, as well as loosing physical contact with bank staff. This finding is in line with the findings of Lee [52] regarding this risk. On the other hand, the role of IB awareness in reducing social risk is proved. These results can be justified in two ways. The first is that sufficient IB awareness of respondents’ relatives caused them to have positive view toward IB which made them regard this risk as minor in their responses. Another justification is that according to the findings of Venkatesh and Davis [78] and Lee [52], social norms, in spite of significant effect on services having obligatory application, have less influence upon the intention of using services having optional usage like IB.

In most solutions suggested in this section for reducing the dimensions of the perceived risk, especial attention has been paid to the concept of advertising. Pikkarainen et al. [64] argued that:

“Banks should now concentrate in their advertising more on informative issues rather than on building only brands with less informative advertisements”.

As it was mentioned before, the concept of IB awareness was emphasized in the findings and focusing on customers’ awareness from each dimension of risk, the effect of awareness on intention of IB adoption was explained. This awareness which can be explained as giving information rather than merely advertising can be created through various channels like TV, radio, magazines, brochures, shows in ATMs, bank website, weblogs, E-mails, short messages, or public training courses.

6 Conclusion and suggestions

This study followed three aims. The first aim was to investigate the effect of each dimension of the perceived risk on bank customers’ intention of using IB based on the perceived risk theory. The results showed that except for social risk, other dimensions of risk including time, performance, security, financial, and privacy has significant negative effects on IB use.

The second aim was to examine the effect of the concept of IB awareness on each dimension of the perceived risk in IB adoption. The findings indicated that IB awareness reduces all aspects of the perceived risk (respectively, performance, security, financial, social, privacy, and time).

The third aim of this study was to determine if it was possible on the basis of the results to consider a mediating role for each dimension of the perceived risk in the relationship between the concept of awareness and IB adoption. According to the results, it was found out that IB awareness has a significant role in increasing the intention of using IB. Considering this finding and the results of the first and second aims, it can be concluded that IB awareness, regarding its direct positive effect on the intention, by reducing the dimensions of the perceived risk, indirectly reduces their negative effect on the intention and thus results in the increase in customers’ willingness to adopt IB. According to these results, it was found out that Customers’ IB awareness affects intention to use IB through the mediation of performance, time, security, financial, and privacy risks, respectively. (Of course, it must be mentioned that considering the results of the first aim, this effect is not true in the case of social risk.) Figure 3 clearly shows this mediating effect.
Fig. 3

The mediating role of dimensions of the perceived risk in the relationship between concept of awareness and IB adoption

Finally, regarding these results, solutions based on the concept of IB awareness were offered applying which within the framework of marketing strategies like pull and push [27] and customer targeting [22] banks can reduce the perceived risk of their customers for adopting IB.

With respect to the limitations of this study, suggestions can be proposed for developing the model under investigation both from a theoretical and survey point of view.

As it was mentioned in Sect. 4.3, 23.7 % of the variance of intention of research was not explained by the dimensions of the perceived risk and IB awareness. This can be attributed to not entering other factors influencing the intention to use IB in research model. In this respect and in order to reduce this value, one suggestion for future research can be made as simultaneous investigation of the dimensions of the perceived risk and other variables affecting intention to use IB according to models and theories proposed in the area of IB adoption.

In this study, the research model is investigated cross-sectionally. This means that the research model is investigated according to views expressed by the respondents at one point of time. This approach, as one of the common approaches, was selected due to theoretical and survey limitations. Another approach to be utilized is longitudinal approach. The method used by this approach is to collect the views of respondents through a survey. Then, by dividing the respondents into two test and control groups during a time period, comprehensive information on IB is provided to test groups and at the end of this period, the views of respondents are measured again. Through this method and by comparing the results obtained from two groups, the effect of IB awareness on the dimensions of the perceived risk and intention to use IB can be investigated.

As it was mentioned in Sect. 3.1, the statistical population of the research is composed of bank customers who do not actively use IB. The sampling was also conducted from this population. According to the reasons pointed out in this section, it is argued that excluding customer actively using IB does not cause bias in the results obtained; rather, it leads to better explanation of the results. However, this issue can be considered as one limitation of the present study. For an in depth analysis, It is suggested that future studies embark on comparing the results obtained from these two statistical population.

This study was conducted in Iran with its unique geographical, cultural, and economic features. Considering previous studies on similar models and theories, it seems that changes in the mentioned features would result in change in the type and power of the relations within the model. Thus, another suggestion for future research is to investigate cultural, national, geographical, and economic restrictions of population regarding the concepts of awareness, dimensions of the perceived risk, and the relationship among them. In other words, interpretation of the results of research model by focusing upon cultural, national, and economic features of the countries can pave the way for expanding an area of IB adoption studies, called comparative studies on IB adoption based on target community.

References

  1. 1.
    Ajzen, I. (1985). From intentions to actions: a theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action control: from cognition to behavior (pp. 11–39). Heidelberg: Springer. Google Scholar
  2. 2.
    Akinci, S., Aksoy, S., & Atilgan, E. (2004). Adoption of Internet banking among sophisticated consumer segments in an advanced developing country. International Journal of Bank Marketing, 22(3), 212–232. CrossRefGoogle Scholar
  3. 3.
    Aladwani, A. M. (2001). Online banking: a field study of drivers, development challenges and expectations. International Journal of Information Management, 21(3), 213–225. CrossRefGoogle Scholar
  4. 4.
    Aldas-Manzano, J., Lassala-Navarre, C., Ruiz-Mafe, C., & Sanz-Blas, S. (2009). The role of consumer innovativeness and perceived risk in online banking usage. International Journal of Bank Marketing, 27(1), 53–75. CrossRefGoogle Scholar
  5. 5.
    Al-Somali, S. A., Gholami, R., & Clegg, B. (2009). An investigation into the acceptance of online banking in Saudi Arabia. Technovation, 29(2), 130–141. CrossRefGoogle Scholar
  6. 6.
    Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: a review and recommended two-step approach. Psychological Bulletin, 103(3), 411–423. CrossRefGoogle Scholar
  7. 7.
    Angelakopoulos, G., & Mihiotis, A. (2011). E-banking: challenges and opportunities in the Greek banking sector. Electronic Commerce Research, 11(3), 297–319. CrossRefGoogle Scholar
  8. 8.
    Association of Payment Clearing Services (Apacs). http://www.ukpayments.org.uk/media_centre/press_releases/-/page/871/. Last accessed on 2010-10-24.
  9. 9.
    Azouzi, D. (2009). The adoption of electronic banking in Tunisia: An exploratory study. Journal of Internet Banking and Commerce, 14(3), 1–11. Google Scholar
  10. 10.
    Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs: Prentice-Hall. Google Scholar
  11. 11.
    Baron, R., & Kenny, D. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology. Monograph Supplement, 51(6), 1173–1182. CrossRefGoogle Scholar
  12. 12.
    Barrett, P. (2007). Structural equation modeling: Adjudging model fit. Personality and Individual Differences, 42(5), 815–824. CrossRefGoogle Scholar
  13. 13.
    Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. CrossRefGoogle Scholar
  14. 14.
    Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88(3), 588–606. CrossRefGoogle Scholar
  15. 15.
    Bestavros, A. (2000). Banking industry walks ‘tightrope’ in personalization of web services. Bank Systems + Technology, 37(1), 54–56. Google Scholar
  16. 16.
    Bollen, B., & Long, B. C. (1993). Testing structural equation models. Newbury Park: Sage. Google Scholar
  17. 17.
    Butler, R. A. (2007). Framework of anti-phishing measures aimed at protecting the online consumer’s identity. Electronic Library, 25(5), 517–533. CrossRefGoogle Scholar
  18. 18.
    Castañeda, J. A., & Montoro, F. J. (2007). The effect of Internet general privacy concern on customer behavior. Electronic Commerce Research, 7(2), 117–141. CrossRefGoogle Scholar
  19. 19.
    Central Bank of I.R. Iran. http://www.cbi.ir/simplelist/2546.aspx. Last accessed on 2010-10-24.
  20. 20.
    Chan, S., & Lu, M. (2004). Understanding Internet banking adoption and use behavior: a Hong Kong perspective. Journal of Global Information Management, 12(3), 21–43. CrossRefGoogle Scholar
  21. 21.
    Chau, P. Y. K. (1997). Re examining a model for evaluating information center success using a structural equation modeling approach. Decision Sciences, 28(2), 309–334. CrossRefGoogle Scholar
  22. 22.
    Chiang, M. H., & Teng, C. I. (2003). Installed base collapsing strategies with network externalities. Journal of Management, 20(5), 829–857. Google Scholar
  23. 23.
    Cook, J. D., Hepworth, S. J., Wall, T. D., & Warr, P. B. (1981). The experience of work. San Diego: Academic Press. Google Scholar
  24. 24.
    Cooper, R. G. (1997). Examining some myths about new product winners. In R. Katz (Ed.), The human side of managing technological innovation (pp. 550–560). London: Oxford University Press. Google Scholar
  25. 25.
    Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of computer technology. Management Information Systems Quarterly, 13(3), 319–340. CrossRefGoogle Scholar
  26. 26.
    DePallo, M. (2000). National survey on consumer preparedness and e-commerce: A survey of computer users age 45 and older. Washington: AARP. Google Scholar
  27. 27.
    Dowling, G. R. (2004). The art and science of marketing. Oxford: Oxford University Press. CrossRefGoogle Scholar
  28. 28.
    Eriksson, K., Kerem, K., & Nilsson, D. (2008). The adoption of commercial innovations in the former Central and Eastern European markets: The case of Internet banking in Estonia. International Journal of Bank Marketing, 26(3), 154–169. CrossRefGoogle Scholar
  29. 29.
    Featherman, M. S., & Pavlou, P. A. (2003). Predicting e-services adoption: a perceived risk facets perspective. International Journal of Human-Computer Studies, 59(4), 451–474. CrossRefGoogle Scholar
  30. 30.
    Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: an introduction to theory and research. Reading: Addison Wesley. Google Scholar
  31. 31.
    Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. CrossRefGoogle Scholar
  32. 32.
    Gemünden, H. G. (1985). Perceived risk and information search: a systematic meta-analysis of the empirical evidence. International Journal of Research in Marketing, 2(2), 79–100. CrossRefGoogle Scholar
  33. 33.
    Gerrard, P., Cunningham, J. B., & Devlin, J. F. (2006). Why consumers are not using Internet banking: a qualitative study. The Journal of Services Marketing, 20(3), 160–168. CrossRefGoogle Scholar
  34. 34.
    Gholami, R., Dolores Anon Higon, M., Hanafizadeh, P., & Emrouznejad, A. (2010). Is ICT the key to development? Journal of Global Information Management, 18(1), 66–83. CrossRefGoogle Scholar
  35. 35.
    Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multi-variate data analysis with readings. Englewood Cliffs: Prentice-Hall International. Google Scholar
  36. 36.
    Hatcher, L. (1994). A step-by-step approach to using the SAS(R) system for factor analysis and structural equation modeling. Cary: SAS Institute. Google Scholar
  37. 37.
    Heeler, R. M., & Ray, M. L. (1972). Measure validation in marketing. Journal of Marketing Research, 9(4), 361–370. CrossRefGoogle Scholar
  38. 38.
    Hernandez, J. M. C., & Mazzon, J. A. (2007). Adoption of Internet banking: proposition and implementation of an integrated methodology approach. International Journal of Bank Marketing, 25(2), 72–88. CrossRefGoogle Scholar
  39. 39.
    Hinkin, T. R. (1995). A review of scale development practices in the study of organizations. Journal of Management, 21(5), 967–988. CrossRefGoogle Scholar
  40. 40.
    Howcroft, B., Hamilton, R., & Hewer, P. (2002). Consumer attitude and the usage and adoption of home-based banking in the United Kingdom. International Journal of Bank Marketing, 20(2), 111–121. CrossRefGoogle Scholar
  41. 41.
    Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to under parameterized model misspecification. Psychological Methods, 3(4), 424–453. CrossRefGoogle Scholar
  42. 42.
    Internet Word State (UK). http://www.Internetworldstats.com/eu/uk.htm. Last accessed on 2010-10-24.
  43. 43.
    Internet Word State (IRI). http://www.Internetworldstats.com/me/ir.htm. Last accessed on 2010-10-24.
  44. 44.
    Jackson, P. R., Wall, T. D., Martin, R., & Davids, K. (1993). New measures of job control, cognitive demand, and production responsibility. Journal of Applied Psychology, 78(5), 753–762. CrossRefGoogle Scholar
  45. 45.
    Jacoby, J., & Kaplan, L. B. (1972). The components of perceived risk. Paper presented at the 3rd Annual Convention of the Association for Consumer Research, Chicago, Illinois, 2–6 November. Google Scholar
  46. 46.
    Jarvenpaa, S. L., & Todd, P. A. (1997). Consumer reactions to electronic shopping on the World Wide Web. Journal of Electronic Commerce Research, 1(2), 59–88. Google Scholar
  47. 47.
    Jun, M., & Cai, S. (2001). The key determinants of Internet banking service quality: A content analysis. International Journal of Bank Marketing, 19(7), 276–291. CrossRefGoogle Scholar
  48. 48.
    Kalakota, R., & Whinston, A. (1997). Electronic commerce: a manager’s guide. Reading: Addison Wesley. Google Scholar
  49. 49.
    Kaplan, L. B., Szybillo, G. J., & Jacoby, J. (1974). Components of perceived risk in product purchase: a cross validation. Journal of Applied Psychology, 59(3), 278–291. CrossRefGoogle Scholar
  50. 50.
    Kenny, D. A. (2001). Measuring model fit. New York: Wiley. Google Scholar
  51. 51.
    Lai, V. S., & Li, H. (2005). Technology acceptance model for Internet banking: an invariance analysis. Information & Management, 42(2), 373–386. CrossRefGoogle Scholar
  52. 52.
    Lee, M. C. (2009). Factors influencing the adoption of Internet banking: An integration of TAM and TPB with perceived risk and perceived benefit. Electronic Commerce Research and Applications, 8(3), 130–141. CrossRefGoogle Scholar
  53. 53.
    Lichtenstein, S., & Williamson, K. (2006). Understanding consumer adoption of Internet banking: an interpretive study in the Australian context. Journal of Electronic Commerce Research, 7(2), 50–66. Google Scholar
  54. 54.
    Lim, N. (2003). Consumers’ perceived risk: sources versus consequences. Electronic Commerce Research and Applications, 2(3), 216–228. CrossRefGoogle Scholar
  55. 55.
    Littler, D., & Melanthiou, D. (2003). Consumer perceptions of risk and uncertainty and the implications for behavior towards innovative retail services: the case of Internet banking. Journal of Retailing and Consumer Services, 13(6), 431–443. CrossRefGoogle Scholar
  56. 56.
    Mandrik, C. A., & Bao, Y. (2005). Exploring the concept and measurement of general risk aversion. Advances in Consumer Research, 32, 531–539. Google Scholar
  57. 57.
    Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis to the study of self-concept: First- and higher order factor models and their invariance across groups. Psychological Bulletin, 97(3), 562–582. CrossRefGoogle Scholar
  58. 58.
    Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(1), 20–38. CrossRefGoogle Scholar
  59. 59.
    Mukherjee, A., & Nath, P. (2003). A model of trust in online relationship banking. International Journal of Bank Marketing, 21(1), 5–15. CrossRefGoogle Scholar
  60. 60.
    Narayanasamy, K., Rasiah, D., & Tan, T. M. (2011). The adoption and concerns of e-finance in Malaysia. Electronic Commerce Research, 11(4), 383–400. CrossRefGoogle Scholar
  61. 61.
    Ndubisi, N. O. (2007). Customers’ perceptions and intention to adopt Internet banking: the moderation effect of computer self-efficacy. Journal of Knowledge, Culture and Communication (AI & SOCIETY), 21(3), 315–327. Google Scholar
  62. 62.
    Nunnally, J. C. (1978). Psychometric theory (Vol. 2). New York: McGraw-Hill. Google Scholar
  63. 63.
    Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: integrating trust and risk with the technology acceptance model. International Journal of Electronic Commerce, 7(3), 101–134. Google Scholar
  64. 64.
    Pikkarainen, T., Pikkarainen, K., Karjaluoto, H., & Pahnila, S. (2004). Consumer acceptance of online banking: an extension of the technology acceptance model. Internet Research, 14(3), 224–235. CrossRefGoogle Scholar
  65. 65.
    Polasik, M., & Wisniewski, T. P. (2009). Empirical analysis of Internet banking adoption in Poland. International Journal of Bank Marketing, 27(1), 32–52. CrossRefGoogle Scholar
  66. 66.
    Prompattanapakdee, S. (2009). The adoption and use of personal Internet banking services in Thailand. The Electronic Journal on Information Systems in Developing Countries, 37(6), 1–31. Google Scholar
  67. 67.
    Rogers, E. M. (1983). Diffusion of innovations. New York: Free Press. Google Scholar
  68. 68.
    Rogers, E. M., & Shoemaker, F. (1971). Communications in innovation. New York: Free Press. Google Scholar
  69. 69.
    Roselius, T. (1971). Consumer rankings of risk reduction methods. Journal of Marketing, 35(1), 56–61. CrossRefGoogle Scholar
  70. 70.
    Sathye, M. (1999). Adoption of Internet banking by Australian consumers: an empirical investigation. International Journal of Bank Marketing, 17(7), 324–334. CrossRefGoogle Scholar
  71. 71.
    Sayar, C., & Wolfe, S. (2007). Internet banking market performance: Turkey versus the UK. International Journal of Bank Marketing, 25(3), 122–141. CrossRefGoogle Scholar
  72. 72.
    Schriesheim, C. A., & Hill, K. (1981). Controlling acquiescence response bias by item reversal: The effect on questionnaire validity. Educational and Psychological Measurement, 41(4), 1101–1114. Google Scholar
  73. 73.
    Segars, A. H., & Grover, V. (1993). Re-examining perceived ease of use and usefulness: a confirmatory factor analysis. Management Information Systems Quarterly, 17(4), 517–525. CrossRefGoogle Scholar
  74. 74.
    Spector, P. E. (1992). Summated rating scale construction: An introduction. Sage University paper series on quantitative applications in the social sciences. Newbury Park: Sage. Google Scholar
  75. 75.
    Straub, D., Boudreau, M. C., & Gefen, D. (2004). Validation guidelines for IS positivist research. Communications of the Association for Information Systems, 13, 380–427. Google Scholar
  76. 76.
    Suh, B., & Han, I. (2002). Effect of trust on customer acceptance of Internet banking. Electronic Commerce Research and Applications, 1(3), 247–263. CrossRefGoogle Scholar
  77. 77.
    Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: a test of competing models. Information Systems Research, 6(2), 144–176. CrossRefGoogle Scholar
  78. 78.
    Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: four longitudinal field studies. Management Science, 45(2), 186–204. CrossRefGoogle Scholar
  79. 79.
    Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: toward a unified view. Management Information Systems Quarterly, 27(2), 425–478. Google Scholar
  80. 80.
    Vrechopoulos, A. P., Siomkos, G. J., & Doukidis, G. I. (2001). Internet shopping adoption by Greek consumers. European Journal of Innovation Management, 4(3), 142–153. CrossRefGoogle Scholar
  81. 81.
    Westland, J. C. (2002). Transaction risk in electronic commerce. Decision Support Systems, 33(1), 87–103. CrossRefGoogle Scholar
  82. 82.
    White, H., & Nteli, F. (2004). Internet banking in the UK: why are there not more customers? Journal of Financial Services Marketing, 9(1), 49–56. CrossRefGoogle Scholar
  83. 83.
    Yiu, C. S., Grant, Y. K., & Edgar, D. (2007). Factors affecting the adoption of Internet banking in Hong Kong: implications for the banking sector. International Journal of Information Management, 27(5), 336–351. CrossRefGoogle Scholar
  84. 84.
    Yousafzai, S. Y., Pallister, J. G., & Foxall, G. R. (2003). A proposed model of e-trust for electronic banking. Technovation, 23(11), 847–860. CrossRefGoogle Scholar
  85. 85.
    Zhao, A. L., Hanmer-Lloyd, S., Ward, P., & Goode, M. M. H. (2008). Perceived risk and Chinese consumers’ Internet banking services adoption. International Journal of Bank Marketing, 26(7), 505–525. CrossRefGoogle Scholar
  86. 86.
    Zolait, A. H. S., & Ainin, S. (2008). Incorporating the innovation attributes introduced by Rogers’ theory into theory of reasoned action: An examination of Internet banking adoption in Yemen. Computer and Information Science, 1(1), 36–51. Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2012

Authors and Affiliations

  1. 1.School of Management and AccountingAllameh Tabataba’i UniversityTehranIran
  2. 2.Department of Financial EngineeringUniversity of Science and CultureTehranIran

Personalised recommendations