• III. RESEARCH METHODOLOGY
  • A. Customer Online Banking Studies
  • C. Instrument Development
  • E. Respondents, Primary Data Analysis and Descriptive Statistics
  • IV. EMPIRICAL RESULTS A. Confirmatory Factor Analysis
  • Table 1. Descriptive Statistics (n=199)
  • Measurement Items for Each Construct Item Reliability Factor Loading
  • Trust in Offline Bank (adapted from Doney and Cannon 1997 [items 1, 3]; Plank et al. 1999 [item 2])
  • Flow (adapted from Ghani et al. 1991 [item 1]; Koufaris 2002 [items 2, 3])
  • Structural Assurance (adapted from McKnight et al. 2002)
  • Perceived Website Satisfaction (adapted from McKinney et al. 2002)
  • Perceived Extent of Use (adapted from Devaraj et al. 2002)
  • Table 3. Correlations, Construct Reliability, and Average Variance Extracted
  • H4: Trust in an offline bank positively influences perceived extent of use of an online bank. G. Online to Online Model Relationships




    Download 207.16 Kb.
    bet2/4
    Sana24.03.2017
    Hajmi207.16 Kb.
    1   2   3   4

    H4: Trust in an offline bank positively influences perceived extent of use of an online bank.
    G. Online to Online Model Relationships
    Perceived structural assurance also relates to website satisfaction. Depending upon the extent of customers’ misgivings regarding the purchasing environment, they may be more committed to the website (and thus more satisfied), or less committed to it (and thus less satisfied). It is hard to feel satisfied with a website about which one is concerned in terms of safeguards. Such concerns may increase cognitive overhead. Hence, lack of structural assurance may interfere with the development of website satisfaction. Security signs or logos indicating the existence of website security systems have also been shown to lessen the psychological burden of online customers [34], which should facilitate website satisfaction.
    H5: Online bank structural assurance positively influences perceived online bank website satisfaction.
    Web satisfaction is dependent on both information quality and system quality [69]. Information quality encompasses web design factors and hypertext document quality. System quality embraces user interface, navigation, and information structure. To the extent that a site user experiences flow, the user will feel positive towards the system and information quality of the website, both of which affect web satisfaction. The user experiencing flow will: a) devote attention to the site, b) experience greater coherence, and c) enjoy using the site because of its good content and design. Therefore, flow will be positively related to perceived website satisfaction.
    H6: Online bank system user flow positively influences perceived online bank website satisfaction.
    Enjoyment of the online shopping experience is an important determinant of customer loyalty [28, 49]. Concentration, a measure of flow, has also been found to positively influence the overall experience of computer users and their intention to use a system again [104]. Flow should improve customers’ perceived coherence, which will lead users to want to use the system. As flow becomes greater, customers tend to become more absorbed in the website. Flow is a pleasurable experience, something users desire. Thus, the pleasure behind flow will increase the likelihood that users will utilize the website again.
    H7: Online bank system user flow positively influences perceived extent of use of an online banking system.
    Meanwhile, structural assurance can influence website extent of use. Because online banking activity is especially vulnerable to financial risks [13, 86], trustworthy structural assurance relieves the tension of customers and lowers cognitive overhead, both of which will improve their perceived extent of use. Structural assurance lowers perceived transaction risks [34, 72], which allows customers a more active interaction with online banking websites. Suh and Han [98] found that such structural assurances as nonrepudiation, privacy protection, and data integrity indirectly influenced both intention to use a website and actual use through trust in the website.
    H8: Online bank structural assurance positively influences perceived extent of use of an online banking system.
    Feelings of interactivity (indicating flow [44, 55]) will be influenced by the perceived safety of online banking transactions (indicating structural assurance [34, 72]). One may have trouble reaching a state of flow unless one feels safe about the online vendor’s website. Once the online banking environment is considered safe, one can use the website without concern, making flow more likely.

    H9: Online bank structural assurance positively influences online bank system user flow.

    Many researchers believe that the extent of use varies with the level of satisfaction with the technology [8, 19, 23]. It has been well documented that customer satisfaction affects repeat purchases, product return rates, brand loyalty and the prevalence of word-of-mouth communications [18]. Satisfaction and trust in websites affects purchasing intentions [8, 19, 23, 103]. Greater satisfaction with a website means that customers favor those products, leading to increased intent to use.



    H10: Perceived online bank website satisfaction positively influences perceived extent of use of an online banking system.

    A boundary condition should be noted for H1-H4: the effects of offline trust will likely be strongest when consumers have little experience with online banking. After some period of online bank experience, online banking variables, such as website satisfaction, will play a more important role than will offline bank trust. This boundary condition will be tested using two subsets of customers—first-time users and experienced users. Results will be presented after the hypothesis tests.



    III. RESEARCH METHODOLOGY
    To examine trust transfer across offline and online banking channels, this study employed a field study method, using questionnaire techniques to measure each construct in the model. Questionnaires enable researchers to obtain the beliefs of prospective website users, and those beliefs are especially important to trust-related research, in which perceptions are more important than objective measures [46].
    A. Customer Online Banking Studies
    Online banking transactions have emerged as an important topic in domestic and international financial circles [60] because online banking allows financial transactions to be carried out conveniently in virtual space, without regard for time or geographical location [103]. Previous studies of online banking have focused on four kinds of research: (1) general description and success factors [2, 4], (2) analysis of behavioral issues such as customer adoption [76, 92] and customer attitudes toward the usefulness and willingness to use online banking [59, 60], (3) regional characteristics [51, 91] and (4) other issues, such as security [40, 107]. Banks are attracted to online banking because it reduces costs and can provide a competitive edge [47]. These factors have led to growth in automated banking such that more than half of banking transactions now take place via telephone, ATM, or the Internet [47]. Internet banking is expected to grow strongly in the future, such that banks that do not offer Internet services are likely to lose significant portions of their customers [100]. Internet banking is growing fast in such Asian countries as South Korea, Singapore, and Thailand [91].
    B. Research Setting
    This study used South Korean subjects responding to questions about online banking. Banking is an appropriate test domain for a trust model because of the potential financial risk to customers an online bank presents [98] and because of the need for online banks to keep transactions and sensitive personal/financial information secure. Trust only makes a difference in an environment of uncertainty and risk, and it is difficult for a consumer to control or monitor an online vendor in order to eliminate risk [34]. Banking also makes a good test because its online component is relatively new compared to offline banking and because of the natural link between offline and online banking. South Korea makes an interesting place to test the model because of its rapid adoption of Internet. Besides, South Koreans are processing about 3 banking transactions out of 10 through the online banking system, or 30.9 % rate for online banking usage according to the Bank of Korea’s survey data on September 2005. This indicates that South Koreans are aware of online banking because most South Korean offline banks encourage customers to use online banking so that they can focus on more profitable banking services.

    C. Instrument Development
    A questionnaire was designed based on the research model (Figure 3). Respondents answered questions based on a five-point Likert-type scale ranging from “strongly disagree” (=1) to “strongly agree” (=5). This study’s trust measure combines trust in the bank itself, the bank’s services, and the bank’s tellers together so that the concept of offline trust is not limited to a single object. (See Table 2 for items.) The first item reflects the honesty/integrity of the offline bank and the second item reflects the bank’s benevolence, in terms of how the bank’s services meet the consumer’s needs [25]. The third item reflects general trustworthiness [84]. Thus, our measures cover two of the three most commonly researched aspects of trust, benevolence and honesty/integrity, while not covering competence, the third aspect [67, 71].
    Flow is measured by the most applicable online aspects of the flow concept: consumer enjoyment, concentration/attention, and perceived control [37, 55]. These are key aspects of flow and follow the theory base on which the hypotheses were developed. They form a unity in terms of being closely related to the psychological aspects of flow.
    Customers who use the Internet for financial transactions have been known to feel threatened by potential online risks [13, 86], such as possible loss of personal property or identity. In light of the general information available about security for Internet-based financial transactions, we adapted three structural assurance items from [71] to measure perceived safety in the vendor’s bank in the midst of perceived online banking risks.
    Several factors determining website satisfaction have been identified, including web design [62], content [85], user interface [94], navigation and information structure [69]. This study adopts three measures representing online banking website satisfaction: information quality, system quality, and overall satisfaction with a website (as noted in [69]). This study measures perceived extent of use based on the items in [23]. Among them, two seem most relevant to perceived extent of use of online banking – intention to reuse, and frequency of use – in order to apply to an online banking case (see Table 2).
    The instruments were translated from English into Korean by bi-lingual (Korean and English) authors. To assure the translation was correct, the authors had 10 MIS doctoral students review the translated instruments and suggest adjustments for subtle nuances in Korean expression. Then, to obtain further translation equivalence, the authors compared the Korean-translated instruments with the same instruments published in Korean in a top-tier South Korean journal, to assure they would convey the original English meaning of the instruments precisely. Translation equivalence from English to Korean language was in this way accomplished in line with other cross-cultural research [77].
    D. Pretest
    To test the psychometric properties of the constructs [97], a questionnaire pretest was given to 68 graduate students enrolled in classes administered by the study authors. The respondents were volunteers and were not told the study’s objective. Respondents were asked to visit their bank’s website to gauge the convenience and services it offered. The convergent validity and unidimensionality of each construct was verified using a principal component factor analysis for factors with eigenvalues above 1, using a varimax rotation. Each item loaded on the intended construct and each Cronbach’s alpha exceeded 0.80. The highest cross-loadings were in the 0.4 range. Since the exploratory factor analysis results appear acceptable for a pretest, we decided to use these measurements (Table 2).
    E. Respondents, Primary Data Analysis and Descriptive Statistics
    Respondents included 325 company workers, housewives and college students. 300 were selected from the Seoul metropolitan area in South Korea, using a stratified quota sampling method. Seoul was first divided into two parts-Southern area and Northern area in order to consider economic status. Then five districts were selected from each group on a random basis, totaling 10 sampling districts. 30 customers were chosen randomly from each district, totaling 300 respondents. Another 25 respondents were MBA students enrolled in the first author’s university. In order to assure the sample was representative, we tried to obtain 165 males and 160 females. We also targeted 125 in the 20s age group, 110 in the 30s age group, and 90 in the 40s age group in order to cover those most interested in online banking. The questionnaire was divided into two parts. The first asks questions about offline banking experience, and the second about online banking experience. Before respondents answered the questions, they were asked to select one bank with which they did offline (but not online) business, and to visit its website to check out its basic banking functions and services. Respondents then answered the questionnaire. The entire exercise took approximately 20 minutes to complete. Among the 325 distributed questionnaires, 85 were not completed validly, and 41 were not returned, resulting in 199 valid responses for a 61.2 % valid return rate. The demographic characteristics of the 199 respondents are shown in Table 1.
    IV. EMPIRICAL RESULTS
    A. Confirmatory Factor Analysis
    The research model has five constructs with interrelated dependence relationships or causal paths, requiring a structural equation model (SEM) analysis [9, 41]. SEM analysis requires constructs to be assessed rigorously by confirmatory factor analysis (CFA) [1, 29, 41, 93], to examine convergent and discriminant validity. CFA results were obtained using software package LISREL 8.30.
    (1) Sample Size:
    Experts consider the minimum sample size for reliable SEM analysis to be from 100 [9], to 150 [3] to over 200 [10]. Because our research model is relatively simple, with only five constructs, the sample size of 199 is considered adequate.
    (2) Item Reliability:
    Item reliability denotes the amount of variance in an item due to its underlying construct. Table 2 shows that t-values for all of the standardized factor loadings of items are significant at p<.01, indicating item reliability. Additional evidence was found using average SMC (squared multiple correlation), which denotes the explanatory power of items related to latent variables. SMC figures in Table 2 lie between .73 and .84, indicating item reliability.
    Table 1. Descriptive Statistics (n=199)

    Construct

    Mean

    Standard Deviation

    Trust in Offline Bank

    3.13

    .66

    Flow

    2.77

    .95

    Structural Assurance

    3.05

    1.02

    Perceived Website Satisfaction

    2.83

    1.02

    Perceived Extent of Use

    3.03

    .83

    Demographic

    Characteristics




    Sex

    Male

    Female

    107 (54%)

    92 (46%)

    Job

    Company Workers

    Housewives

    College Students

    109 (55%)

    54 (27%)

    36 (18%)

    Age

    20-29

    30-39

    Over 40

    46 (23%)

    85 (43%)

    68 (34%)

    Online Banking Experience

    Below 1 month including first-time user

    1-12 months

    Over 1 year

    113 (57%)

    22 (11%)

    64 (32%)


    Table 2. CFA Results

    Measurement Items for Each Construct

    Item Reliability

    Factor Loading

    Std. Errors

    Std.

    Loading

    t-value

    Average SMCa

    Cronbach’s alpha

    Trust in Offline Bank (adapted from Doney and Cannon 1997 [items 1, 3]; Plank et al. 1999 [item 2])

    .79

    .91

    This bank keeps the promises it makes to me.

    1.00

    -

    .88

    -

    This bank’s services meet my needs.

    1.06

    .05

    .94

    19.62**

    This bank’s teller is trustworthy.

    .98

    .06

    .86

    16.76**

    Flow (adapted from Ghani et al. 1991 [item 1]; Koufaris 2002 [items 2, 3])

    .84

    .94

    During my visit to this online banking website, I found a lot of interesting content.

    1.00

    -

    .91

    -

    During my visit to this online banking website, my attention was focused on online banking activity.

    1.02

    .05

    .92

    21.06**

    During my visit to this online banking website, I felt in control.

    1.02

    .05

    .93

    21.29**

    Structural Assurance (adapted from McKnight et al. 2002)

    .73

    .88

    This online banking website has enough safeguards to make me feel comfortable using it for my personal business.

    1.00

    -

    .88

    -

    I feel assured that the legal and technological structures of this online banking website adequately protect me from Internet problems.

    .97

    .07

    .85

    15.59**

    In general, this online banking website is a robust and safe environment in which to transact business.

    .95

    .06

    .84

    15.21**

    Perceived Website Satisfaction (adapted from McKinney et al. 2002)

    .81

    .92

    I feel satisfied with the information quality offered by this online banking website.

    1.00

    -

    .91

    -

    I feel satisfied with the system quality by this online banking website.

    .98

    .05

    .89

    19.02**

    After using this online banking website, I feel very satisfied.

    .99

    .05

    .90

    19.78**

    Perceived Extent of Use (adapted from Devaraj et al. 2002)

    .76

    .85

    I intend to use the services offered by this online banking site again.

    1.00

    -

    .90

    -

    I intend to visit this online banking website as often as possible.

    .93

    .08

    .84

    12.13**

    ** p <0.01, a SMC = Squared Multiple Correlation

    (3) Construct Reliability, Convergent Validity, and Average Variance Extracted (AVE):


    Table 3 shows construct reliability and AVE figures. Reliability is a necessary condition for evaluating convergent validity. Construct reliability estimates range from .86 to .94, and all are greater than .70. The AVEs, which should meet a .50 standard, fall between .75 and .84, indicating convergent validity.
    Table 3. Correlations, Construct Reliability, and Average Variance Extracted


    Intercorrelations between Constructs

    Flow

    Structural Assurance

    Perceived Website Satisfaction

    Perceived Extent of Use

    Trust in Offline Bank

    Flow

    1.00













    Structural Assurance

    .72

    1.00










    Perceived Website Satisfaction

    .73

    .79

    1.00







    Perceived Extent of Use

    .58

    .56

    .52

    1.00




    Trust in Offline Bank

    .69

    .79

    .76

    .69

    1.00

    Construct Reliability (>.70)

    .94

    .89

    .93

    .86

    .92

    Average Variance Extracted (>.50)

    .84

    .75

    .81

    .76

    .80

    (4) Discriminant Validity:


    Since intercorrelations between constructs are relatively high (refer to Table 3), common method bias may exist. In order to detect this, a discriminant validity test was performed in accordance with [29], one of the more statistically rigorous methods of doing so. In this test, the squared correlations between two constructs must be lower than the corresponding AVE. Table 3 shows that the AVE figures, ranging from .75 to .84, all exceed the squared correlations between the five constructs, the highest of which is .63, confirming discriminant validity of the proposed constructs. Thus, the five constructs possess adequate convergent and discriminant validity for further SEM analysis.


    Download 207.16 Kb.
    1   2   3   4




    Download 207.16 Kb.

    Bosh sahifa
    Aloqalar

        Bosh sahifa



    H4: Trust in an offline bank positively influences perceived extent of use of an online bank. G. Online to Online Model Relationships

    Download 207.16 Kb.