94 年 10 月 23 日
報告人姓名
王怡舜 服務機構
及職稱
國立彰化師範大學資訊管理系 副教授
時間 會議
地點
2005 年 7 月 11 日至 2005 年 7 月 13 日
澳洲雪梨
本會核定 補助文號
壹會綜二字第○九三○○三○二 一二號(93.05.25)
NSC 93-2416-H-018-003 會議
名稱
(中文)第四屆行動化企業國際研討會
(英文)The Fourth International Conference on Mobile Business,2005 (ICMB 2005)
發表 論文 題目
(中文)預測台灣消費者行動商務使用意向
(英文)Predicting Consumer Intention to Use Mobile Commerce in Taiwan 報告內容應包括下列各項：
一、參加會議經過
後學於 2005 年 7 月 8 日搭機起程前往澳洲雪梨，會議期間共進行三日之論文研
討，並於7 月 15 日搭機回國。
二、與會心得
此次參與國際學術會議，讓我覺得不應關在象牙塔中進行學術研究，應多與國際學 術接軌與交流，透過和國際學者之討論，可以使自己初步的研究成果得到更多的批評 與指教，使論文的嚴謹度更加提昇，這也有助於日後將研討會論文進一步發表於國際 學術期刊之上。
三、考察參觀活動(無是項活動者省略) 四、建議
非常感謝國科會能夠補助後學出席國際學術會議，確實對於後學有莫大的助益，
期望未來國科會能夠提供更多的經費補助，以鼓勵新進學者參與國際學術會議之論文 發表，除了可以帶回最新的研究資訊與瞭解相關領域目前研究方向與趨勢，更可以接 觸相關領域的學者，拓展國際視野。
五、攜回資料名稱及內容
第四屆行動化企業國際研討會論文集 六、其他
附件為發表之論文
Predicting Consumer Intention to Use Mobile Commerce in Taiwan
Hsin-Hui Lin Yi-Shun Wang
National Taichung Institute of Technology National Changhua University of Education E-mail:[email protected] E-mail:[email protected]
Abstract
Advances in wireless technology have increased the number of people using mobile devices and accelerated the rapid development of mobile commerce conducted with these devices. However, while many companies are vying for market share of the new business opportunities offered by wireless technology, research on m-commerce suggests potential consumers may not adopt these mobile systems, in spite of their availability. Based on an assessment of the existing literature regarding the theory of planned behavior and technology acceptance model, the current research presents an integrated model for predicting consumer intention to use m-commerce systems. It does this by adding (1) a trust-related construct (“perceived credibility”) and (2) resource-related constructs (“perceived self-efficacy”
and “perceived financial resources”) to the TAM, with careful attention to placing these constructs in TAM’s existing nomological structure. Data from 258 users in Taiwan tests against the integrated model using the structural equation modeling approach. The results support the integrated model in predicting consumer intention to use m-commerce.
1. Introduction
Mobile commerce (m-commerce), or electronic commerce (e-commerce) utilizing mobile devices, has become a major topic of interest for the information systems (IS) and marketing research community. As it becomes increasingly evident that PC-based e-commerce has failed to live up to expectations and achieve broad mass adoption, m-commerce has become a key priority for many business organizations [33].
Predictions, based on anecdotal and empirical evidence on the future popularity and volume of m-commerce have been widely presented in academic literature and in the business and technology press. While many authors and research firms believe the demand for
m-commerce services will skyrocket over the next five years, others have been far more conservative in their predictions. One reason is that many empirically optimistic predictions, on the future popularity of m-commerce, rely on indirect units of measurement rather than direct studies based on consumer intention to use m-commerce [7].Anckarand D’Incau also argued that the popularity of m-commerce could not be measured by the popularity of mobile devices [7]. Similarly, the popularity of wired e-commerce cannot be measured by the popularity of computers, as has been proven. While a growing body of literature, matched by limited empirical evidence, has highlighted the valuable elements of m-commerce [16,7], consumers’ primary reasons for adopting and intending to adopt mobile services remain unclear [37,32]. However, building successful strategies for the mobile marketplace unquestionably begins by understanding the factors affecting consumer intention to use m-commerce systems. For several years, businesses have tried to introduce m-commerce systems to improve their operations and reduce costs. Despite all the efforts aimed at developing better and more efficient m-commerce systems, these systems were either ignored by consumers, or seriously under-used, in spite of their availability.
Therefore, the primary objective of this research is to understand the acceptance of m-commerce from consumers’perspectives and to identify the factors that can predict their intention to use business-to-consumer m-commerce systems. While extensive research on the technology acceptance model (TAM) has explained why individuals voluntarily accept or reject information systems, TAM has some limitations. These include its omission of the importance of a trust-based construct in the context of electronic or mobile commerce, and that it assumes a lack of barriers capable of preventing individuals from using an IS if they so choose. Based on the literature concerning theory of planned behavior (TPB) and TAM, this study presents an integrated model for predicting the consumer intention to use m-commerce by adding (1) a trust-related construct
(“perceived credibility”) and (2) resource-related constructs (“perceived self-efficacy”and “perceived financial resources”) to the TAM, with careful attention given to placing these constructs in TAM’s existing nomological structure. An important goal throughout this work is to develop a model capable of providing useful information to m-commerce practitioners, while, at the same time, maintaining TAM’s theoretical and psychometric rigor. By explaining usage intention from consumers’perspectives,thefindingsofthisresearch will not only help m-commerce practitioners to develop better user-accepted m-commerce systems, but also provide insights into how to promote new IT to potential customers.
2. Theoretical foundations
M-commerce acceptance is of particular interest to the current study, and represents a fundamental managerial challenge in m-commerce implementation.
A review of prior studies suggests the theoretical foundations of the hypotheses formulations. Toward this direction, this study examines two prevalent theories (i.e., TAM and TPB) for investigating individual IT or IS acceptance in the m-commerce context.
2.1. Technology acceptance model
Technology acceptance model [18,19] adapted from the theory of reasoned action (TRA) [6,21], posits that user adoption of a new information system is determined by users’intention to usethesystem,which in turn isdetermined by users’beliefs about the system.
TAM further suggests two beliefs—perceived usefulness and perceived ease of use—are instrumental in explaining the variance in users’ intentions. Perceived usefulness is defined as the extent to which a person believes that using a particular system will enhance his or her job performance, while perceived ease of use is defined as the extent to which a person believes that using a particular system will be free of effort. Among the beliefs, perceived ease of use is hypothesized to be a predictor of perceived usefulness.
Information systems researchers have investigated and replicated the TAM, and agreed that it is valid in predicting individual’sacceptanceofvariouscorporate IT [1,14,20,34]. However, the TAM’s fundamental constructs do not fully reflect the specific influences of technological and usage-context factors that may alter user acceptance [31]. Thus, research is required to seek additional factors that better predict the acceptance of m-commerce by individuals. Prior studies have extended TAM with constructs such as perceived playfulness [31], cognitive absorption [2], and product
involvement and perceived enjoyment [27]. Recently, Gefen et al. added “trust”as a construct to the TAM in the context of online shopping [22]. Wang et al. also successfully introduced the trust-related construct of perceived credibility, as a new TAM factor, to reflect user security and privacy concerns in the acceptance of online banking [41]. Perceived credibility is usually impersonal and relies on reputation based information and economic reasoning [8]. Consequently, perceived credibility is used as a TAM construct to reflect the security and privacy concerns in the acceptance of m-commerce.
2.2. Theory of Planned Behavior
Theory of planned behavior extends from TRA by incorporating an additional construct, namely, perceived behavior control, to account for situations in which an individual lacks substantial control over the targeted behavior [5]. According to TPB, an individual’s behavior can be explained by his or her behavioral intention, which is jointly influenced by attitude, subjective norms, and perceived behavioral control. Ajzen [5] indicated that perceived behavioral control is most compatible with Bandura’s [9,10]
concept of perceived self-efficacy which “is concerned with judgments of how well one can execute courses of action required to deal with prospective situations”
[10]. Prior research has confirmed the critical role that computer self-efficacy plays in understanding individual acceptance of IT [4,26,24,12]. Thus, perceived self-efficacy of m-commerce will be an important knowledge resource for consumers in adopting m-commerce. On the other hand, several consumers confirmed, during our qualitative interviews with them, that financial considerations, including handset, subscription, service, and communication fees, might influence their behavioral intentions to use m-commerce. Mathieson et al. also found that hardware/software and financial resources are important for users in adopting an information system [30]. In our opinion, the practicality of this study would have been diminished had we omitted the influence of financial considerations. Consequently, considering the model parsimony and the resources required to use m-commerce, the current study extends TAM by adding one internal behavioral control factor “perceive dself-efficacy”and one external behavioral control factor
“perceived financial resources” to reflect people’s concerns about their knowledge and financial resources needed to use m-commerce. By focusing on resources specific to m-commerce usage, researchers can better delineate factors that managers may have some degree of control over. Measuring perceptions of the
availability of specific resources would help m-commerce practitioners identify opportunities for interventions that could increase system use.
3. Research model and hypotheses
The research model tested in this study is shown in Figure 1. In the extended model, like many other studies of TAM [28,11], the “attitudes” construct is removed in an attempt to simplify the model. The proposed constructs and hypotheses are based on prior studies in the information systems literature [3,4,13,18,19,24,25,26,29,38,39,40,41].
H1: Increases in perceived usefulness increases behavioral intention to use m-commerce.
H2: Increases in perceived ease of use increases perceived usefulness of m-commerce.
H3: Increases in perceived ease of use increases perceived credibility of m-commerce.
H4: Increases in perceived ease of use increases behavioral intention to use m-commerce.
H5: Increases in perceived credibility increases behavioral intention to use m-commerce.
H6: Increases in perceived self-efficacy increases perceived ease of use of m-commerce.
H7: Increases in perceived self-efficacy increases behavioral intention to use m-commerce.
H8: Increases in perceived financial resources increases behavioral intention to use m-commerce.
4. Research design and method 4.1. Measures of the constructs
Items selected for the constructs were mainly adapted from prior studies to ensure content validity.
TAM’s instruments are psychometrically sound. The perceived ease of use and perceived usefulness
instruments show good convergent and discriminant properties [1,15,18], are internally reliable [18,19,29]
and demonstrate predictive validity [35]. Items for the perceived ease of use and perceived usefulness were taken from the previous validated inventory and modified to fit the specific technology studied. The items to measure behavioral intention were taken from previous applications of TAM [3,39]. The items for the perceived self-efficacy construct were adapted from the original instrument of computer self-efficacy developed by Compeau and Higgins [17]. Perceived credibility was measured by two items adapted from Wang et al.
[41] to reflect specific user beliefs concerning the security and privacy protection of m-commerce. Finally, perceived financial resources were measured by two statements specifically developed for this study. Likert scales (1~7), with anchors ranging from “strongly disagree” to “strongly agree” were used for all questions with the exception of those items for measuring perceived self-efficacy whose anchors ranged from “not at all confident” to “totally confident.”2
4.2. Data collection procedure
Data used to test the research model was gathered from a sample of respondents attending an e-commerce exposition and symposium held in Taiwan. The questionnaire was self-administered by the respondents.
The questionnaire consists of the measures and a request for demographic information. Respondents were asked to circle the response which best described their level of agreement with the statements. A total of 573 approaches were made to obtain 258 completed surveys. Reason for nonparticipation was mainly due to a lack of time to complete the survey. A total of 63 percent of the completed surveys were from male respondents. Respondents ranged from 18 to 45 years of age (mean = 32 years); 35 percent had completed one college or university degree; and a further three percent had completed post-graduate degrees.
Table 1. Fit indices
2The measures used in this study are available from the authors upon request.
5. Data analysis and results 5.1. Measurement model
A confirmatory factor analysis using LISREL 8.3 was conducted to test the measurement model. As shown in Table 1, all the model-fit indices exceeded their respective common acceptance levels, suggested by previous research, and thus demonstrating that the measurement model exhibited a fairly good fit with the data collected. Therefore, we could proceed to evaluate the psychometric properties of the measurement model in terms of reliability, convergent validity, and discriminant validity.
Reliability and convergent validity of the factors were estimated by composite reliability and average variance extracted. (see Table 2). Composite reliability for all the factors, in our measurement model, was above 0.70. The average extracted variances were all above the recommended 0.50 level [23], signifying that more than one-half of the variances observed in the items were accounted for by their hypothesized factors.
Convergent validity can also be evaluated by examining the factor loadings and squared multiple correlations from the confirmatory factor analysis.
Following Hair et al.’s recommendation[23], factor loadings greater than 0.50 are considered very significant. All of the factor loadings of the items in the research model were greater than 0.70. Also, squared multiple correlations between the individual items and their a priori factors were high (above 0.50 in all cases). Accordingly, all factors in the measurement model had adequate reliability and convergent validity.
Table 2. Reliability, average variance extracted, and discriminant validity
Factor CR PU PEU PC PSE PFR BI PU 0.89 0.72
PEU 0.81 0.13 0.59 PC 0.73 0.14 0.08 0.57 PSE 0.83 0.12 0.21 0.02 0.63 PFR 0.75 0.11 0.02 0.18 0.00 0.60 BI 0.86 0.49 0.28 0.32 0.20 0.25 0.76 CR: Composite reliability; PU: Perceived usefulness; PEU:
Perceived ease of use; PC: Perceived credibility; PSE:
Perceived self-efficacy; PFR: Perceived financial resources;
BI: Behavioral Intention
Diagonal elements are the average variance extracted. Off-diagonal elements are the shared variance.
To examine discriminant validity, we compared the shared variances between factors with the average variance extracted of the individual factors. This analysis shows that the shared variance between factors
were lower than the average variance extracted of the individual factors, which confirms discriminant validity (see Table 2). In summary, the measurement model demonstrated adequate reliability, convergent validity, and discriminant validity.
Note: t-values for standardized path coefficients are described in parentheses.
Figure 2. Hypotheses testing results 5.2. Structural model
A similar set of fit indices was used to examine the structural model (see Table 1). Comparison of all fit indices, with their corresponding recommended values, provided evidence of a good model fit. Based on such confirmation, we could proceed to examine the path coefficients of the structural model.
Properties of the causal paths, including standardized path coefficients, t-values, and variance explained for each equation in the hypothesized model are presented in Figure 2. As expected, the findings support hypotheses H1, H4, H5, H7, and H8 in that perceived usefulness, perceived ease of use, perceived credibility, perceived self-efficacy, and perceived financial resources all had significant effects on behavioral intention. Altogether, they account for 68 percent of the variance in behavioral intention with perceived usefulness (β=0.46) contributing more to intention than perceived ease of use ( β =0.21), perceived credibility (β=0.28), perceived self-efficacy (β=0.18), and perceived financial resources (β=0.28).
In addition, hypotheses H2, H3, and H6 were also supported. Perceived self-efficacy was found to have a significant influence on perceived ease of use ( β
=0.48), which, in turn, had a positive effect on both perceived usefulness ( β =0.39) and perceived credibility (β=0.31).
6. Implications for Practice
This study proposed an integrated model to explain and predict consumer intention to use m-commerce systems based on the TAM and TPB. The findings of this study strongly support the appropriateness of using the integrated model to understand the acceptance of m-commerce by individuals. Perceived usefulness, ease of use, credibility, self-efficacy, and financial resources were observed to have positive influences on behavioral intention. We also found both perceived credibility and perceived financial resources to have stronger effects on behavioral intention than the traditional TAM variable, perceived ease of use. That is, cost, security and privacy issues are important concerns for consumers in using m-commerce. Given that the usage of m-commerce is completely voluntary, and lacks of organizational resource support and that the target user group consists of a large number of people with diversified backgrounds, the findings of this study suggest that making the system easy to interact with is an insufficient means of attracting more users to use m-commerce. Developing m-commerce systems with valuable functions and trustworthy security and privacy protection is very important to users. In addition, the m-commerce authorities need to decrease user perception of the financial costs associated with using m-commerce through promotion and pricing strategies.
On the other hand, perceived self-efficacy had a significant effect on perceived ease of use, which, in turn, had positive influences on perceived usefulness,
On the other hand, perceived self-efficacy had a significant effect on perceived ease of use, which, in turn, had positive influences on perceived usefulness,