Empirical Validation of Online Features in the User Acceptance of Internet Banking Websites


PAULINUS OFEM, PASCHAL OCHANG, LUCKY ONWUZURIKE AND ABSALOM EZUGWU

Department of Computer Science

Faculty of Science, Federal University Lafia P.M.B 146 Lafia

NIGERIA

paulinus.ofem@fulafia.edu.ng

JYRI RAJAMÄKI

Research, Development and Innovations

Laurea University of Applied Sciences

Vanha maantie 9, FI-02650 Espoo

FINLAND

jyri.rajamaki@laurea.fi

http://www.laurea.fi/en/leppavaara


Abstract: - Internet banking was introduced in the developed countries around the 1980s whereas in developing countries like Nigeria, it only started around the year 2000. This study focuses on investigating the online features of Internet banking websites of Nigerian banks and the impact of their relationships with the acceptance behaviors of users from an information systems (IS) view. The empirical results obtained from a sample survey of 55 Internet banking users indicate that technology acceptance model (TAM) is an appropriate model for use in forecasting the acceptance behaviors of Internet banking customers. It also indicates that online features have positive effect on customer’s perceived usefulness, attitude and behavioral intension to adopt Internet banking websites.

Key-Words: - Online features; Internet banking websites; Technology acceptance model; Internet banking, E-banking


1 Introduction

The rapid development and growth of mobile communication networks and the Internet has triggered the adoption of both mobile commerce [1], [2] and electronic commerce in general, giving rise to various means of carrying out commercial transactions through the use of the Internet especially in developing countries. In developing regions like Africa, the amount of mobile phone users outnumber the amount of PC users and bank account holders [3], this implies that the number of mobile banking users will tend to grow [4] thereby enhancing the adoption of e-banking in developing countries. The increasing demand on the use of the Internet in carrying out commercial transactions in Nigeria has encouraged the banking sector to invest heavily in technological infrastructure [5] thereby improving e-banking systems in order to meet up with increasing customer populations and corresponding financial demands. All the existing commercial banks in Nigeria now provide Internet banking websites and other forms of e-banking platforms.

As there are many commercial banks in Nigeria providing e-banking services, we envisage a competition between e-banking providers. Banks in Nigeria are also faced with the challenge of dealing with customer requirements in the case of Internet banking in order to provide optimum services. The acceptance of e-banking in developing regions is still uncertain with [4] pointing out that there are no known actual figures or statistics of mobile banking users in Africa. However, the opportunities of the system growing in under-developed regions are enormous. Several works in the literature of electronic commerce investigated and reported factors that are capable of influencing the satisfaction of users and their usage of Web-based shopping centers. In [6], Web quality factors were generally discussed and also in [7], [8], [9], the success model based on the success factors of Web-based shopping centers was equally presented. We borrow some of the outcomes of these studies into the e-banking domain because it is very similar to the Web-based shopping centers. We specifically base our work on Information Systems (IS) opinions [10], [11], [12], [13], [14] which includes the fact that customer acceptance behaviors can be predicted by exploring the technical specifications of a system via constructs like service quality, system quality and information quality.

To this end, in our study, we first carry out an empirical exploration of the online features of Internet banking websites from the point of view of IS. Secondly, we explore the relationships between the online features and behavioral acceptance of customers via path coefficient analysis using technology acceptance model (TAM) as the theoretical basis for the exploration. We focus on how each online quality factor relates with TAM’s individual constructs.

2 Technology Acceptance Model

In order to understand the user acceptance behaviors of different forms of technologies, several acceptance models have been developed with the technology acceptance model (TAM) [15] being widely and commonly used to investigate user acceptance of various forms of Internet technology driven applications. TAM opines that the use of a system is influenced by user’s behavioral intention to use the system and this behavioral intension is in-turn influenced by user’s attitude to use the system. As we mentioned earlier, Internet banking websites constitute a Web system which enables customers to complete certain financial transactions as they will normally do in a traditional banking setting. Since there is no face-to-face interaction between the bank’s service desk and the customer, the only medium of communication is the user interface provided by the website. It is therefore not out of context to seek to understand customer’s requirements and expectations from their use of the Internet banking platform in order to serve them better. A customer’s positive attitude towards using the Internet banking website significantly and positively influences his behavioral intention to continue using the website.

Previous studies have adopted the TAM in understanding technology acceptance among users especially in the domain of Internet technology-driven applications in recent times [11], [12], [14], [16], [17], [18]. However, this has seen TAM being extended to include other external constructs. Wu and Wang [2] integrated the extended technology acceptance model (TAM2) [19] with the innovation diffusion theory (IDT), risk and cost in order to study the acceptance of mobile commerce. Wu and Chen [20] extended TAM by integrating it with the Theory of Planned Behaviour (TPB) in order to study the acceptance of online tax in Taiwan.

As aforementioned, TAM has been subjected to several modifications that gave birth to different models. Some of these modifications revolve around the inclusion of external constructs or variables that IS researchers argue can further develop TAM and make it more sustainable. For instance we mentioned that online quality features such as information quality, system quality and service quality have been used as external constructs to successfully explore user acceptance behavior via TAM. Therefore, adopting TAM as a model and as a theoretical foundation towards exploring the acceptance of Internet banking websites in Nigeria is a justified approach since TAM allows the introduction of external variables in determining user acceptance of technology.

3 Conceptual model development

Figure 1 presents our extended research model. The model was originally proposed by Ahn et al. [14] to include offline features. Based on our contextual study, we removed the offline features. The research model is a combination of the well-known technology acceptance model (TAM) as established by Davis [15] and online features as developed and validated in [1], [13], [14], [20]. For the purpose of this study, we have relied on the original constructs of TAM such as perceived ease of use (PEOU), perceived usefulness (PU), attitude and behavioral intension to Use (BIU) in investigating the relationships between external variables and customer’s acceptability of Internet banking websites. Unlike in [14], where the TAM model was used to validate the online and offline features of Internet shopping malls, we set out to validate only the online features which are based on information systems research views in the domain of E-banking and in a Nigerian context. The offline features which are based on marketing research views are not applicable in this context and therefore excluded. The basis of the research analysis is the Nigerian user of E-banking and our sample population includes all categories of people who transact with various E-banking website providers.

Fig.1. Proposed conceptual model (Adapted from [14])




















3.1 The Measurement Development

In the collection of data for the constructs of the underlying research model, we used a questionnaire. The measurement items in the questionnaire were based on the seven-point Likert scale. In [1], [13], [21] the measurement items for the Web quality constructs were developed and validated as the key instruments for measuring the online features.

In order not to reinvent the wheel in the assessment of the measurement items, we adopted almost all the sample measurement items presented by Ahn et al. [14] which had already gone through vigorous and thorough evaluation process undertaken by IS professors and passed. We however rephrased the statements in the questionnaire to reflect our domain and also removed few sub-measurement items that had a lower reliability score based on our small sample data. The original TAM constructs and other related measurements were equally adopted from related studies such as [11], [12], [14], [15]. The final questionnaire was made up of 40 measurement items which are measuring the variables under consideration. Table 1 shows the final questionnaire for online quality factors.

3.2 Online Survey

Since we intend to undertake a wider study which would enable us make a more meaningful generalization of the case of Nigeria as we would be targeting Nigerian online users after this current pilot study is concluded, we employed a Web-based survey for this study using Lime Survey tool. The online survey enabled to recruited respondents across the country who have actually engaged in Internet banking. We targeted all commercial banks in Nigeria that provide e-banking services. The questionnaire opens with a set of instructions to guide the respondents. Respondents were specifically asked to respond to all the statements based on their individual e-banking experiences. They were also instructed to respond to the statements using the Likert scale of 1-7; where 1 corresponds to “strongly disagree” and 7 corresponds to “strongly agree”.

A total of fifty five (55) respondents were recruited, resulting in 55 recorded cases and with no missing values. The cases were first harvested from Lime Survey into a CSV file. The CSV file was then imported into SPSS for initial data pre-processing. The resulting data was then saved as an SPSS file for onward importation into LISREL 9.2 for actual data analysis. Due to the small sample involved, analysis with LISREL software was deemed inappropriate [22], [23]. We however relied on SPSS and AMOS for data analysis. About 61 percent of the respondents were male. Majority of the participants were between the ages of 20-29. The data collected is a reflection of the opinions of actual users of e-banking.

We found that majority of the participants are civil servants and students who mainly use the Internet to do e-banking at home and in the office. The difference between participants who do e-banking at home or the office is negligible. Most of the respondents have a considerable degree of Internet experience of over 4 years and the two most preferred e-banking channels of choice are the ATM and Internet banking followed by mobile banking. The most preferred e-banking provider of choice is Guaranty Trust Bank followed by First Bank and Diamond Bank, both of which share equal percentage.

4 Results

The conceptual model was analysed using IBM’s SPSS Statistics 20 and IBM’s SPSS AMOS 23 software. Maximum likelihood approach was used in estimating the research model. We performed data analysis in two phases. In the first phase, the measurement model was subjected to a critical examination and validation. In the second phase, we performed a structural equation model analysis which is supported in SPSS Amos via path coefficient analysis. This is in order to test the relationships between the constructs in the research model.

4.1. Measurement Model

Our sample data was tested for suitability for factor analysis and the result showed a KMO and Bartlett’s test values which were below the recommended values. This again we attributed to the small sample size involved. Exploratory factor analysis (EFA) and LISREL analysis was therefore discontinued. We hope to employ LISREL analysis in our future wider study that will accommodate a larger sample size.

A reliability test is usually given as a measure of the degree of consistency between measurement items corresponding to a scale [24]. We used Cronbach’s alpha coefficient to measure the internal consistency of the scale. According to Nunnaly et al. [25], a reliability score of 0.70 can be accepted as a minimum though Lance et al. [26] argued that this acceptable minimum can also be misleading. A review of past literature shows 0.70 is the most common acceptable minimum score and we have adopted it as a standard in this study. Due to the small sample size and also low item-total results, we dropped the following measurement items: SYSQ8 (multimedia) for system quality, PEOU5 (mental effort) for ease of use construct, and lastly PU4 (saves money) for perceived usefulness. All constructs indicated a reliability of between 0.80 and 0.90 except service quality which has 0.75. The reliability of the scale was therefore found to be sufficient.

4.2 Structural Model

Given the success of the reliability test, we proceeded with path coefficient analysis to test the general model fit. Table 2 presents a summary of the overall model fit indices for the proposed research model. Again, due to the small sample size, we did not expect our Chi-square value to be high. A χ2 value of 19.329, with 7 degrees-of-freedom and at a significance level of 0.007 was obtained. This is considered significant. The values for GFI, CFI and NFI were above the recommended value of 0.90. The RMSEA, AGFI and TLI values were at their marginal acceptable levels. For a small size of ours, we believe they should be acceptable. It is worthy of mentioning that SPSS AMOS has the capability of suggesting significant relationships between constructs in order to produce a perfect model based on the sample data. This is usually given in the modification indices. A significant relationship was found between system quality and attitude. We effected this suggestion and the outcome is what is summarized and presented in Table 3. A χ2 value of 7.761, with 6 degrees-of-freedom and at a significance level of 0.256 was obtained. A p-value of <0.05 is usually recommended. The GFI, NFI, TLI and CFI values were all above the recommended value. We note that the RMSEA for this computer suggested perfect model is 0.074. According to the indices obtained, there is however, no significant difference between our model and the AMOS’s suggested model.

4 Discussion

First, there is quite no significant difference between our model and AMOS’s suggested model fit. Following from the structural model, we deduce the following; system quality indicates a strong relationship with perceived usefulness (β=0.27, p<0.01) and perceived ease of use (β=0.19, p<0.01), information quality has strong positive influence on perceived usefulness ((β=0.23, p<0.01) and a rather negative influence on ease of use (we expected the reverse). Service quality has significant influence on perceived usefulness (β=0.18, p<0.01) and a positive relationship with ease of use (β=0.14, p<0.01). System quality shows a slightly higher impact on usefulness and behavioral intention to use internet banking out of the entire external variables considered. This impact rate is quite insignificant when compared with information quality. The results obtained shows that system quality, information quality and service quality mediated by ease of use indicates a strong influence on usefulness, which in turn shows a strong influence on attitude and behavioral intention to use e-banking websites.

The results so far obtained indicate that Interne banking users have identified with online features of Internet banking websites in the e-banking domain, and this is consistent with similar research reported in [14]. In the e-banking domain, the bank’s client acts as a system user. The online features of any e-banking website play a very significant role in determining the behavioral intention of customer’s continuous use of the Web platform for financial transactions.

Based on our small sample data and the results obtained, TAM produced outcomes which are quite consistent with testing the behavioral intention of bank customers to use e-banking websites. The attitude towards using Internet banking for example, has a great impact on user’s behavior to use e-banking websites. User’s attitude is mediated by his or her perceived ease of use and perceived usefulness. We obtained path coefficients of 0.46, 0.68 and 0.27 between ease of use and usefulness, usefulness and attitude, and between usefulness and behavioral intention to use. These results are consistent with the submissions of Lin et al. [12] and Ahn et al. [14], thereby confirming that perceived usefulness has a greater and significant relationship with perceived ease of use, and perceived ease of use has a greater and significant impact on perceived usefulness. By implication, perceived usefulness will continue to remain a major determinant of e-banking website usage while perceived ease of use will indirectly affect user’s intention to use e-banking websites. We therefore deduce that perceived usefulness is influenced by perceived ease of use. In the same vein, users who find it easy to use e-banking websites will equally find e-banking websites very useful.

References:

[1] Barnes, S., Vidgen, R., 2001. An evaluation of cyber-bookshops: the WebQual method. Int. J. Electron. Commer. 6, 11–30. doi:10.1080/10864415.2001.11044225.

[2] Wu, J., Wang, S., 2005. What drives mobile commerce? Inf. Manag. 42, 719–729. doi:10.1016/j.im.2004.07.001

[3] Porteous, D., 2006. The Enabling Environment for Mobile Banking in Africa, Bankable Frontiers Associates Boston USA. London.

[4] Baptista, G., Oliveira, T., 2015. Understanding mobile banking: The unified theory of acceptance and use of technology combined with cultural moderators. Comput. Human Behav. 50, 418–430. doi:10.1016/j.chb.2015.04.024

[5] Ayo, C., Adewoye, J., Oni, A., 2010. The state of e-banking implementation in Nigeria: A post-consolidation review. J. Emerg. trends … 1(1), 37–45.

[6] Liu, C., Arnett, K. P., 2000. Exploring the factors associated with Web site success in the context of electronic commerce, Information and Management 38(23–33).

[7] Torkzadeh G., Dhillon G., 2002. Measuring factors that influence the success of internet commerce, Information Systems Research 13 (2) 187–204

[8] Koufaris, M., Kambil, A., LaBarbera, P. A., 2002. Consumer behavior in Web-based commerce: an empirical study, International Journal of Electronic Commerce 6 (2)15–138.

[9] Aladwani, A. M., Palvia, P. C., 2002. Developing and validating an instrument for measuring user-perceived Web quality, Information and Management 38 (467–476).

[10] DeLone, W.H., McLean, E., 2003. The DeLone and McLean model of information systems success: A ten-year update. J. Manag. Inf. Syst. 19, 9–30. doi:10.1073/pnas.0914199107

[11] Lederer, A., Maupin, D., Sena, M., Zhuang, Y., 2000. The Technology Acceptance Model and the World Wide Web. Decis. Support Syst. 29, 269–282. doi:10.1016/S0167-9236(00)00076-2

[12] Lin, J.C., Lu, H., 2000. Towards an understanding of the behavioural intention to use a web site. Int. J. Inf. Manage. 20, 197–208.

[13] Ranganathan, C., Ganapathy, S., 2002. Key dimensions of business-to-consumer web sites. Inf. Manag. 39, 457–465. doi:10.1016/S0378-7206(01)00112-4

[14] Ahn, T., Ryu, S., Han, I., 2004. The impact of the online and offline features on the user acceptance of Internet shopping malls, in: Electronic Commerce Research and Applications. pp. 405–420. doi:10.1016/j.elerap.2004.05.001

[15] Davis, F.D., 1989. Perceived Ease of Use, and User Acceptance of Information Technology. MIS Q. 13, 319–340. doi:10.2307/249008

[16] Bhattacherjee, A., 2001. An empirical analysis of the antecedents of electronic commerce service continuance, Decision Support Systems. 32, 201–214.

[17] Kim, D.J., Ferrin, D.L., Rao, H.R., 2008. A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and their antecedents. Decis. Support Syst. 44, 544–564. doi:10.1016/j.dss.2007.07.001

[18] Raida, R.E., Néji, B., 2013. The Adoption of the E-Banking: Validation of the Technology Acceptance Model. Technol. Invest. 04, 197–203. doi:10.4236/ti.2013.43023

[19] Venkatesh, V., Davis, F., 2000. Theoretical extension of the Technology Acceptance Model: Four longitudinal field studies. Manage. Sci. 46, 186–204. doi:10.1287/mnsc.46.2.186.11926

[20] Wu, I.L., Chen, J.L., 2005. An extension of Trust and TAM model with TPB in the initial adoption of on-line tax: An empirical study. Int. J. Hum. Comput. Stud. 62, 784–808. doi:10.1016/j.ijhcs.2005.03.003

[21] Palmer, J.W., 2002. Web site usability, design, and performance metrics. Inf. Syst. Res. 13, 151–167.

[22] Chin, W.W., 1998. Issues and Opinion on Structural Equation Modeling. MIS Q. 22, 1.

[23] Gefen, D., 2000. E-commerce: the role of familiarity and trust. Omega Int. J. Manag. Sci. 28, 725–737. doi:10.1016/S0305-0483(00)00021-9

[24] Pallant, J., 2004. SPSS survival manual. Open University Press, Mcgraw-Hill, London.

[25] Nunnally, J.C., Bernstein, I., 1994. Psychometric Theory, Rdsepiucsforg. McGraw-Hill, Sydney.

[26] Lance, C.E., Butts, M. M., Lawrence, C.M., 2006. The Sources of Four Commonly Reported Cutoff Criteria: What Did They Really Say? Organ. Res. Methods 9, 202–220. doi:10.1177/1094428105284919

[27] Ives, B., Olson, M.H., Baroudi, J.J., 1983. The measurement of user information satisfaction. Commun. ACM 26, 785–793. doi:10.1145/358413.358430