This action might not be possible to undo. Are you sure you want to continue?
Ming-Chi Lee Dept. of Computer Science and Information Engineering, National Pingtung Institute of Commerce, Taiwan, R.O.C Abstract This paper investigates how consumers perceive and adopt or resist online shopping from initial intention to continued intention. We proposed a theoretical model which synthesizes the expectation confirmation model (ECM), technology acceptance model (TAM), theory of planned behavior (TPB), and perceived risk to identify the determinants which influence consumers’ online shopping decision processes. The proposed model is then validated empirically using a longitudinal method with a two-stage survey. The results demonstrate that satisfaction has the most significant effect on consumers’ continuance intention, followed by perceived usefulness as a significant but weaker predictor, and also that the initial intention to use online shopping is adversely affected mainly by perceived risk. This study is significant for two reasons. First, it synthesizes the expectation confirmation model, technology acceptance model, theory of planned behavior in a complementary approach and, second, it help bride the existing gap among resistance, initial acceptance and continuance intention to use online shopping. Keywords: Expectation-confirmation model (ECM), Theory of planned behavior (TPB), Technology acceptance model (TAM), Online shopping.
With the rapid growth of the Internet, online shopping has become one of the most profitable e-commerces on the Internet over the last decade (Vijayasarathy 2004). However, online shopping is still confronted with many obstacles. First, considerable Internet users are still reluctant to use online shopping (Forsythe and Shi, 2003; Soopramanien and Robertson, 2007). Second, the acceptance-discontinuance anomaly phenomenon (i.e., users discontinue using online shopping after initially accepting it) frequently occurs (Bhattacherjee 2001). For the former issue, several authors (Forsythe and Shi, 2003; Hsu et al., 2006) have attributed consumers’ reluctance to purchase online to apparent risks (e.g. credit card stolen risk, identity theft risk) but these risks and their potential impact on consumers’ continuance online purchase behavior are unclear. For the latter issue, prior studies have given more attention to examining factors that drive users to “initially adopt online shopping” (i.e., use online shopping for the first time), rather than the factors that influence uses to “continue to use online shopping” after they have adopted it. Although initial acceptance of online shopping is an important first step toward achieving success of online stores, the eventual success of online stores still needs customers to repurchase things online. Therefore, with a better understanding of the consumers’ online purchase decision processes from resistance and initial adoption to continued intention to use online shopping, practitioners should be able to make adjustments in their marketing plans that should help to provide them with a sustainable advantage over their competition. In order to fill this knowledge gap, this study proposed a theoretical model which synthesizes the expectation confirmation model (ECM), technology acceptance model (TAM), theory of planned behavior (TPB), and perceived risk to understand consumers’ online shopping behavior from initial intention to continuance intention. We combine these
theoretical models and perspectives for the following three reasons. First, although previous
research has found ECM to be a robust model for continued IT adoption (Bhattacherjee, 2001), it employs only three variables to explain behavioral intention, namely satisfaction, confirmation, and post-adoption expectations. However, a user’s behavioral intention toward adopting IT will also be affected by other factors, such as the opinions of important individuals (subjective norms) (Fishbein and Ajzen, 1975). Furthermore, even if users have a strong intention to perform a behavior, they may feel that they lack the necessary resources and skills (perceived behavioral control) (Ajzen, 1991), and the use of TPB addresses this gap. Second, while TPB captures the roles of individuals, organizational members, and social influences on behavioral intention, it does not inform us what attitudinal beliefs would affect a user’s attitude toward online shopping (Hsu et al., 2006). According to Taylor’s research (Taylor and Todd, 1995a), TAM provides two attitudinal beliefs, namely perceived ease of use and perceived usefulness as two major antecedents of attitude, which make up precedent factors of attitude for TPB. Moreover, since each theory has distinct roots and is based on a different set of antecedent variables, we contend that they independently provide a partial understanding of users’ cognitive processes related to IT usage. It is therefore possible that, when combined, these theories may collectively provide an improved and more comprehensive understanding of the cognitive processes and behaviors related to IT usage than when each theory considered alone. Third, customers have shown reluctance to complete online purchases, primarily due to risk concerns (Gefen et al., 2003). This is because there is a lot at stake for consumers as they entering into a business relationship with distant, faceless online purchases. While perceived risk has been examined in the online shopping behavior, little has been known about its influence on the consumers’ continuance intention to shop online, which may help understand the acceptance-discontinuance anomaly phenomenon. Therefore, we will also use
perceived risk to examine customer’s initial acceptance and continued usage of online shopping. While prior research has examined ECM, TAM, and TPB independently in explaining information technology (IT) usage, to the best of our knowledge, no study has yet theoretically combined these three models. The primary contributions of this study are its examination of the integration of ECM, TAM, TPB and perceived risk in explaining initial acceptance and continuance intention toward online shopping and an empirical evaluation of which factors are critical to affecting this intention. The findings from this paper may therefore help bridge the existing gap between initial acceptance and continuance streams of online shopping research. The results of this work would be useful for online store practitioners in formulating strategies aimed at increasing online purchases. The rest of the paper is organized as follows. The next section presents our research model and hypotheses, while Section 3 proposes the measurement method and scales. We present the research results in Section 4, followed by discussion in Section 5. Finally, the implications and conclusions of this work are presented in Sections 6 and 7.
2. Theoretical background and literature review
2.1. Expectation-confirmation model (ECM) Expectation-confirmation theory (ECT) is widely used in the consumer behavior literature to study consumer satisfaction, post-purchase behavior, and service marketing in general (Anderson and Sullivan, 1993; Oliver, 1980, 1993). In order to understand users’ continued information technology (IT) usage behavior, Bhattacherjee (2001) adapt the ECT to propose the expectation confirmation model (ECM), which emphasizes the importance of post-usage expectation rather than pre-usage expectation. Therefore, instead of pre-usage expectation, post-usage perceived usefulness, a surrogate for the post-usage expectation is included. Furthermore, because ECM assumes the influence of perceived performance is
already explained by confirmation variable, it does not include the performance variable. The ECM posits that an individual’s intention to continue IT usage is dependent on three variables: the user’s level of satisfaction with the IT; the extent of user’s confirmation of expectations; and post-usage expectations, in the form of perceived usefulness. Fig. 1 presents the ECM.
Post-Adoption Expectations (Perceived Usefulness)
Continued IT Usage Intention
Fig. 1 Expectation-Confirmation Model
There are five main hypotheses in the ECM. First, users’ satisfaction with IT has a positive effect on their intention to continue using the IT. Studies in marketing have discovered that the major reason for a consumer’s decision to repurchase products or patronize services is their level of satisfaction (e.g. (Bearden and Teel, 1983; Oliver, 1993; Szymanski and Henard, 2001)). Owing to the similarity between re-purchasing products/services in a consumer context and the continued usage of IT products/services, the ECM posits an equivalent relationship in the latter context. In turn, user’s satisfaction with IT is determined by the user’s confirmation of expectations and their perceived usefulness of IT (which is one type of post-adoption expectation). The confirmation of expectations suggests that users obtained expected benefits through their usage experiences with the IT, and thus leads to a positive effect on users’ satisfaction. On the other hand, based on the expectancy-confirmation
paradigm, users’ post-usage perceived usefulness of IT has a positive effect on their satisfaction with IT by working as a baseline for reference against confirmation judgments. This relationship is supported by the adaptation level theory, which proposes that users perceive stimuli only in relation to an adapted level. 2.2. The technology acceptance model (TAM) Davis (1989) first introduced the TAM as a theoretical extension of the theory of reasoned action (TRA) (Fishbein and Ajzen, 1975) and found that it could better explain user’s acceptance. TAM proposes that two particular beliefs, perceived usefulness and perceived ease of use, are the primary drivers for technology acceptance. Perceived usefulness is defined as ‘‘the degree to which a person believes that using a particular system would enhance his/ her job performance’’, and perceived ease of use is defined as ‘‘the degree to which a person believes that using a particular system would be free of physical and mental effort’’ (Davis, 1989). Further, perceived usefulness and perceived ease of use both affect a person’s attitude toward using the system, and consistent with TRA, these attitudes toward using the system determine behavioral intentions, which in turn lead to actual system use. TAM has been extensively applied to user acceptance research of various types of technologies including e-mail, word processor, world wide web, enterprise resources planning (ERP) systems, and e-commerce (Davis, 1989; Gefen and Straub, 1997; Lu et al., 2009). 2.3. Theory of planned behavior (TPB) Both the theory of planned behavior (TPB) and TAM were developed based on the theory of reasoned action (TRA) (Ajzen and Fishbein, 1980; Fishbein and Ajzen, 1975), which argues that both behavioral attitude and subjective norm affect behavioral intention, which in turn affects the actual behavior. TPB adds to TRA a third factor – perceived behavioral control – that affects behavioral intention and actual behavior (Ajzen, 1991).
Many studies have replicated and investigated these three constructs and found that they are valid in explaining individual intention to use various forms of 2000). 2-4 Perceived risks Since the 1960s, perceived risk theory has been used to explain consumers’ behavior. Considerable research has examined the impact of risk on traditional consumer decision making (Lin, 2008). Peter and Ryan (1976) defined perceived risk as a kind of subjective IT (Liao, 1999; Venkatesh,
expected loss, and Featherman and Savlou (Featherman and Pavlou, 2003) also defined perceived risk as the possible loss when pursuing a desired result. Cunningham(Cunningham, 1967) noted that perceived risk consisted of the size of the potential loss (i.e. that which is at stake) if the results of the act were not favorable and the individual’s subjective feelings of certainty that the results will not be favorable. Most of scholars claimed that consumers’ perceived risk is a kind of a multi-dimensional construct. Six components or types of perceived risk have been identified: financial, performance, social, physical, privacy, and time-loss (Jacoby and Kaplan, 1972; Kaplan et al., 1974; Roselius, 1971) (see Table 1).
3. Research model and hypothesis development
3-1 Research model This study synthesizes the ECM, TAM, TPB and perceived risk to hypothesize a theoretical model to understand consumers’ online shopping behavior from initial intention to continuance intention. Figure 2 illustrates the research model. It asserts that the continued intention to use online shopping is a function of: satisfaction post-usage perceived usefulness and perceived risk. The model further indicates satisfaction mediates the impact of confirmation and post-usage perceived usefulness on the continued intention conduct online
shopping. The proposed constructs and hypotheses are supported by prior studies in information systems literature.
Post-usage Perceived usefulness Pre-usage Perceived usefulness
Attitude Perceived ease of use
Perceived behavior control Fig. 2. Research model and hypotheses.
3-2 Hypothesis development 3-2-1 Hypotheses about ECM Prior marketing studies have found that the higher the users’ expectations, the higher are their satisfaction (Oliver and DeSarbo, 1988). Moreover, the IT adoption literature has consistently found that post-usage perceived usefulness is the most important determinant of
users’ adoption intentions (e.g., (Davis et al., 1989; Taylor and Todd, 1995b; Venkatesh, 2000)). As a result, the ECM posits users’ post-usage perceived usefulness of IT has a positive effect on their intention to continue IT usage. Lastly, the ECM posits that the users’ confirmation of expectations will have a positive effect on the post-usage perceived usefulness of IT. Perceived usefulness of IT could thus be adjusted by confirmation experience, particularly when the users’ initial perceived usefulness is not concrete due to the uncertainty over what to expect from using the IT (Bhattacherjee, 2001). Because online shopping is a kind of information technology on the Internet, we derived the following hypotheses from the ECM: H1. Users’ satisfaction with online shopping is positively related to their continuance intention to shop online. H2. Users’ confirmation of expectations is positively related to their satisfaction toward online shopping. H3. Users’ post-usage perceived usefulness of online shopping is positively related to their satisfaction toward online shopping. H4. Users’ post-usage perceived usefulness of online shopping is positively related to their continuance intention to shop online. H5. Users’ confirmation of expectations is positively related to their post-usage perceived usefulness of online shopping. 3-2-2 Hypotheses about TAM Consumers need to see online shopping as a useful information technology that can improve their purchase efficiency, providing compelling advantages such as broader product lines, faster transaction, and greater flexibility. Moreover, online shopping users need feel that the online shopping is easy to use. Both pre-usage perceived usefulness and perceived ease of use are beliefs that, according to TRA, will affect a user’s attitude and initial intention. Thus, we have:
H6. Pre-usage perceived usefulness is positively related to behavioral attitude toward online shopping. H7. Pre-usage perceived usefulness is positively related to the initial intention toward online shopping. H8. Perceived ease of use is positively related to behavioral attitude toward online shopping. In addition, TAM states that perceived usefulness will have a direct effect on user’s behavioral intention, and perceived ease of use affects behavioral intention indirectly through pre-usage perceived usefulness (Davis, 1989). That is, pre-usage perceived usefulness mediates the effect of perceived ease of use on behavioral intention. Many empirical studies have supported this argument (Venkatesh and Davis, 2000; Wu and Chen, 2005). Thus, H9: Perceived ease of use is positively related to pre-usage perceived usefulness of online shopping. 3-2-3 Hypotheses about TPB Attitude (A) refers to ‘‘the degree of a person’s favorable or unfavorable evaluation or appraisal of the behavior in question’’ (Fishbein and Ajzen, 1975). According to the TPB, attitude impacts users’ behavioral intention, which in turn influences their actual behavior. When individuals form positive attitude towards online shopping, they will have a stronger intention toward adopting it, and thus they are more likely to use it. H10. Behavioral attitude toward online shopping is positively related to the initial intention to shop online. Subjective norm refers to ‘‘the perceived social pressure to perform or not to perform the behavior’’ (Ajzen, 1991). In other words, subjective norm is related to the normative beliefs about the expectation from other people. Many Internet users choose to shop online because their friends shop online too, and they recommend it to them. Hence, we propose: H11. Subjective norm is positively related to the initial intention toward online shopping.
Perceived behavioral control refers to ‘‘people’s perception of ease or difficulty in performing the behavior of interest’’. It is associated with beliefs about the presence of control factors that may facilitate or hinder the performance of the behavior in question (Ajzen, 1991, 2002; Liao et al., 2006). In this case, although online shopping is a useful tool to improve shopping efficiency, users still need to have the basic Internet skills to use it. Thus, we posit that: H12. Perceived behavioral control is positively related to the initial intention toward online shopping. 3-2-4 Hypotheses about perceived risk The theory of perceived risk has been applied to explain consumer’s behavior in decision making since the 1960s. Mitchell (1999) suggested that perceived risk is powerful at explaining consumers’ behavior because consumers are more often motivated to avoid mistakes than to maximize utility in purchasing. According to Gefen (2003), an increase in the risk perceived by consumers could reduce their intention to buy through a particular website. Similar logic should hold true for perceived risk toward continuance intention. Thus, we propose that: H13. Perceived risk is negatively related to the initial intention to use online shopping. H14. Perceived risk is negatively related to the continuance intention to use online shopping.
4. Research methodology
4.1 Questionnaire development We used a questionnaire survey with two parts to test our theoretical model. The first part had questions measuring the constructs in the research model, while the second part had demographic questions about the participants. Each item corresponding to the constructs was measured using a seven-point Likert scale, with answer choices ranging from “disagree
strongly” (1) to “agree strongly” (7), and most of these items were adapted from the extant literature. The scale items for perceived usefulness and perceived ease of use were adapted from Davis(1989) and Bhattacherjee (2001), while the scale items for subjective norm, perceived behavioral control, and attitude were adapted from the Taylor and Todd (1995a). In addition, the items measuring perceived risk was adapted from Lee (2009), and the continuance intention, satisfaction, confirmation were adapted from Bhattacherjee (2001). After the questionnaire was drafted, it was first sent to three academic experts on online shopping for their review, and then revised according to their comments and suggestions to make the wording of the items more precise. 4-2 Pilot test The questionnaire was pilot-tested by convenient sampling. There were 150 responses, of which 123 were complete, giving a valid response rate of 82%, and the results of the pilot test were evaluated by using Cronbach’s reliability and factor analysis. The reliability coefficient was first calculated for the items of each construct, and the standard lower bound for Cronbach’s alpha set at 0.7 (Hairs et al., 1998), with items that did not significantly contribute to the reliability being eliminated. A factor analysis was then performed to examine whether the items produced the anticipated number of factors and whether the individual items were loaded on their appropriate factors. All items had high loadings on their related factors and low cross-loadings on other factors, showing good convergent and discriminate validities. 4.3 Sample plan and data collection The participants had prior experience of using the Internet but had no experience of online shopping. They aged from 21 to 60, and their educational backgrounds graduated from junior school graduates to Ph.D. holders. Finally, 62% of the respondents were male. Table 1 presents further demographic details on the sample group.
These e-marketing courses last for three months. During this period, all the participants received a two-stage survey. In the first stage survey, they were given a two-hour training session by three research assistants to help understand the purchasing procedure of the online shopping. Then, they were asked to just visit shopping stores but were not allowed to purchase at that time. After this training, we conducted the questionnaire to measure subjects’ perceptions of pre-purchase beliefs for online shopping such as perceived risk, perceived usefulness, attitude, subjective norm, perceived behavioral control, and initial intention. At the end of first survey, 322 questionnaires were collected. In the second stage survey, subjects who had participated in the first-stage survey were encouraged to purchase online. At the end of the course, we conducted questionnaires to measure subjects’ post-purchase beliefs for online shopping, such as feelings of confirmation, satisfaction, and continuance intention. The returned questionnaires with incomplete or invalid answers were eliminated, and a total of 302 valid responses were received. The results showed that all the subjects accomplished the shopping procedure completely, namely browsing, gathering product information, making product and price comparisons, making purchase decision, and completing a transaction by offering payment and address information. In addition, 288 of the participants shopped online more than once in the period studies.
In analyzing the collected data, we followed the two-step procedure suggested by Anderson and Gerbing (1988). First, we examined the measurement model to measure convergent and discriminant validity. We then examined the structural model to investigate the strength and direction of the relationships among the theoretical constructs. 5.1 Analysis of the measurement model Convergent validity was assessed based on the criteria that the indicator’s estimated
coefficient was significant on its posited underlying construct factor. We evaluated the measurement scales using the three criteria suggested by Fornell and Lacker (1981), namely: (1) (2) (3) All indicator factor loadings ( ) should be significant and exceed 0.5 Construct reliabilities should exceed 0.8 Average variance extracted (AVE) by each construct should exceed the variance due to
measurement error for the construct (e.g., AVE should exceed 0.5) The Cronbach’s alpha scores, shown in Table 2, indicated that each construct exhibited strong internal reliability, which all the standard factor loading ( ) values in confirmatory factor analysis of the measurement model exceeded 0.5 and were significant at p=0.001. In addition, the composite reliabilities of constructs ranged from 0.81 to 0.93, and the AVE, ranging from 0.57 to 0.83, was greater than the variance due to measurement error. Therefore, all three conditions for convergent validity were met. ---Insert Table 2 about here--Discriminant validity assesses the extent to which a concept and its indicators differs from another concept and its indicators (Bagozzi et al., 1991). According to Fornell and Larcker (1981), the correlations between items in any two constructs should be lower than the square root of the average variance shared by items within a construct. As shown in Table 3, the square root of the variance shared between the construct and its items was greater than the correlations between the construct and any other construct in the model, satisfying Fornell and Larckers’ (1981) criteria for discriminant validity. All diagonal values exceeded the inter-construct correlations, and thus the results confirmed that our instrument had satisfactory construct validity.
Post-usage Perceived usefulness Pre-usage Perceived usefulness
Attitude Perceived ease of use
Perceived behavior control Fig. 3. Results of structural modeling analysis.
5.2 Analysis of the structural model We assessed the overall goodness of fit using the chi-square test. The chi-square test assesses the adequacy of a hypothesized model in terms of its ability to reflect the variance and covariance of the data. Due to its tendency to be sensitive to sample size, other fit indices (namely GFI, AGFI, CFI, NFI, and RFI) were considered in conjunction with the chi-square. For the statistical significance of parameter estimates, t values were used. The results of
structural equation modeling obtained for the proposed conceptual model revealed a ratio of chi-square to the degree of freedom ( 2 / df ) of 2.04 ( p < 0.001), goodness-of-fit index (GFI)
of 0.91, adjusted goodness-of-fit index (AGFI) of 0.85, comparative fit index (CFI) of 0.95, normed fit index (NFI) of 0.95, relative fit index (RFI) of 0.94, and root mean square error of approximation (RMSEA) of 0.05. Generally, fit statistics greater than or equal to 0.9 for GFI, NFI, RFI, and CFI indicate a good model fit (Hairs et al., 1998). Furthermore, RMSEA values ranging from 0.05 to 0.08 are acceptable (Hairs et al., 1998), indicating that our model fit was acceptable. The other fit indices, except AGFI, indicated that our proposed model obtained an adequate model fit. The low AGFI values may have been due to the small sample size used. 5.3 Hypotheses testing The fourteen hypotheses presented above were tested collectively using the structural equation modeling (SEM) approach (Bagozzi et al., 1991), also performed using AMOS 6. The path significance of each hypothesized association in the research model and variance explained (R2 value) by each path were examined, and Fig. 3 and Table 4 show the standardized path coefficients and path significances. All fourteen hypothesized associations were strongly significant at p<0.05, except for one link between perceived risk and continuance intention. The continuance intention to use online shopping in this study was jointly predicted by satisfaction ( 0.53, p 0.01 ), perceived usefulness
( 0.37 , p 0.001 ) and perceived rik ( 0.12 , p 0.12 ) and these variables together explained 80% of the variance of intention to use ( R 2 0.80, coefficient of determination). In addition to its direct effects, post-usage perceived usefulness also had a significant indirect effect on continuance intention ( 0.12 , p 0.05 ) via the satisfaction, explaining 18% of the variance in the dependent variables. Satisfaction, in turn, was predicted by confirmation ( 0.31 , p 0.01 ) and post-usage perceived usefulness ( 0.22 , p 0.001 ). Confirmation also had a small indirect effect ( 0.06 , p 0.05 ) on satisfaction, via the
post-usage perceived usefulness construct (see Table 4). Initial intention was predicted by perceived risk ( 0.41 , p 0.01 ), pre-usage
perceived usefulness ( 0.21 , p 0.05 ), subjective norm ( 0.22 , p 0.05 ) and perceived behavioral control ( 0.25 , p 0.01 ), and together these variables explained 67% of the total variance. To further assess the significance of indirect effects of predictor variables on intentions to use online shopping, a decomposition of the effects analysis was conducted (see Table 5). A discussion of these effects for online shopping continuance is presented in next Section.
---Insert Table 4 about here-----Insert Table 5 about here---
The results of this study provide support for the research model and for the hypotheses regarding the directional linkage among the model’s variables. The overall explanatory power of our research model had an R-square of 75% for initial intention to use online shopping and an R-square of 80% for continued intention, suggesting that the extended ECM model is capable of explaining a relatively high proportion of variation of intention to use online shopping. Several significant results could be derived from our research framework, and these are presented below. We examined the effects of satisfaction, perceived usefulness, and perceived risk on the continuance intention with regard to online shopping. The results show that satisfaction is the strongest predictor of users’ continuance intention ( 0.53 , p<0.01), followed by post-usage perceived usefulness ( 0.37 , p<0.05) as a significant but weaker predictor, and perceived risk ( 0.11 , p=0.14) as an insignificant predictor. Satisfaction explains 43% of
the variance of continuance intention relative to the other predictors, which jointly explained 33% of the variance. Users who are dissatisfied with online shopping may stop using it, despite having positive perceptions with regard to other elements of the experience. In other words, dissatisfaction is the necessary condition for online shopping discontinuance. Therefore, satisfaction may be the key to explaining the online shopping
acceptance-discontinuance anomaly, a little-understood phenomenon in the literature. Although both the post-usage and pre-usage perceived usefulness were identified in this study as important determinants of initial intention and continuance intention toward online shopping, the post-usage perceived usefulness has a stronger effect than pre-usage perceived usefulness. This is consistent with Bhattacherjee (2001), which found there are often huge changes in consumers’ expectations following their consumption experience, and these changes then have an impact on subsequent cognitive processes. The pre-usage perceived usefulness is a cognitive belief, formed potentially via second-hand information from referent others, the media, advertising or other sources. These influence sources may be biased, and hence user attitude potentially may be inaccurate, unrealistic, and uncertain. In contrast, post-usage perceived usefulness is grounded in users’ first-hand experience, and is therefore more realistic, unbiased, and less susceptible to change. In view of this, online store practitioners should adopt a two-fold strategy in order to increase the initial adoption and continued usage of online shopping: specifically, they should inform new (potential) users of the potential benefits of online shopping use, and educate old (continued) users on how to use online shopping more effectively so as to maximize their perceived usefulness with such methods. It is worth noting that the perceived risk has not significant influence on the continuance intention. One possible explanation for this is that once users decide to continue to repurchase online, they are satisfied with and /or perceive the usefulness of such purchases, and so are
willing to accept the related risks. In contrast, perceived risk has a significantly negative impact on the initial intention to use online shopping. This implies that before adopting online shopping, concerns about perceived risk are foremost in the minds of online consumers, and thus may be the prime reason for not undertaking such purchases. Consequently, practitioners need to develop for risk-reducing strategies that might assist in inspiring higher confidence in potential customers. The results of this study also support the idea that the user’s confirmation and post-usage perceived usefulness are key determinants of satisfaction. In addition, we find that confirmation has the primary effect on satisfaction. Combined with the earlier finding that satisfaction is the primary antecedent of continuance intention, these two results confirm the saliency of the expectation-confirmation paradigm in understanding users’ continuance intention toward online shopping. While all the TAM and ECM variables are significant in this study, they implicitly assume that behavior is volitional. However, online shopping users face several novel constraints, such as the impersonal nature of the online environment, certain necessary resources and skills (perceived behavioral control), and the influence of the opinions of other important persons (subjective norms). These issues call for the inclusion of TPB in the online shopping adoption model, and both subjective norms and perceived behavioral control are verified as having a significant influence on continuance intention. This finding implies that when users find that people around them have adopted online shopping services, they will be more willing to use them. Online shopping providers may thus use a positive word-of-mouth strategy to enhance the awareness of their services and promote their benefits. They may also need to consider how to bring more positive experiences to their existing customers to retain their acceptance, rather than relying on the mass media to deliver reinforcing message (Bhattacherjee, 2000).
7-1 Implications for Academics In terms of theory building, this study attempts to develop a new theory by grounding a new variable in an integration of two schools of the nomological structure model of theory of reasoned action (TRA) as well as a derivation of expectation disconfirmation theory (EDT) and applying them into a new context. It is important to note that the new variable, perceived risk, is compatible with the TAM, TPB and ECM variables. This approach is likely to ensure a stable theory development. Hence, the proposed model makes an important contribution to the emerging literature on online shopping. The present study has two implications for future online shopping research. First, the empirical results show that the unified model supports all the hypotheses and has good explanatory power, implying that the integration of ECM, TAM, and TPB provides a model with a theoretical basis to explain online shopping. This approach may provide an initial blueprint for the further integration of other theoretical acceptance models. For example, information technology (IT) research (Venkatesh et al., 2003) has already yielded many competing models such as innovation diffusion theory (IDT), social cognitive theory (SCT), expectation disconfirmation theory (EDT) and theory of reasoned action (TRA), each with different sets of acceptance determinants. It is anticipated that this study may encourage other research that integrates these competing models into unified ones. Second, while the results show that the factors in TAM, TPB and ECM all have significantly direct or indirect effects on continuance intention to use online shopping, satisfaction has the strongest effect. Since confirmation and expectations are critical antecedents to satisfaction, future research may explore what factors influence these variables and how they can be manipulated in order to improve eventual user experience with online
shopping, and hence its subsequent continuance.
7-2 Implications for practitioners The results of this study shed light on some important issues related to consumers’ initial adoption and continuance intention toward online shopping. First, while the intention to use online shopping is significantly influenced by many factors, it is adversely affected mainly by perceived risk. This implies that controlling the risk of online shopping is particularly important for managers as they decide how to allocate resources to retain and expand their current customer base. However, building a risk-free online shopping transaction environment is a difficult thing. Therefore, online stores need to search for risk-reducing strategies that might assist in inspiring high confidence in potential customers. This study suggests that they should consider focusing on the prevention of fraud, identity theft, and financial loss. For example, building secure firewalls to avoid intrusion, developing methods for strengthening encryption, and authenticating websites are all measures that should be undertaken. In addition, this study suggests that online shopping companies could develop trust-building mechanisms to attract customers, such as statements of guarantee, increased familiarity through advertising, and long-term customer service. It is worth noting that because online shopping is a less verifiable and controllable environment, online shopping customers usually have difficulty in asking for compensation when dissatisfaction for products/services occurred. Thus, we suggest that online shopping companies should provide customers with legal digital receipts or a guarantee for every transaction in order to increase confidence in such services.
8. Limitations and suggestions future research
Several limitations and suggestions for future research are worth noting. First, some threats to internal validity may need to be taken into account with a longitudinal study such as
this one, including history, maturation, testing, mortality and regression (Bhattacherjee and Premkumar, 2004). To minimize the impact of jeopardizing factors, we adopt several proactive steps, as suggested by Bhattacherjee and Premkumar (2004). The time period between pre-usage and usage stage is three months, and this period is short enough to avoid the effects of history and maturation, but long enough to get rid of effect of testing. In addition, our high response rates between both stages demonstrate a low mortality effect. Second, because we collected the data for the independent and dependent variables from the same respondents, concerns about common method bias could arise (Woszczynski and Whitman, 2004), and we thus conducted Harmon’s one-factor test (Podsakoff et al., 2003) to assess the risk of this. We entered all the variables into a factor analysis. These factors emerged with the first factor, accounting for 23.4% of the variance in the variables. Because more than a single factor emerged from the factor analysis and no general factor accounted for the majority of the variance in those variables, we saw no evidence to suggest the presence of common method variance bias. Third, about 62% of the respondents were male in this empirical study, and thus the gender distribution was not symmetrical. Therefore, the results of the current empirical study might tend to model the specific behavior of men, rather than general behavior of all users. Much evidence has shown that gender differences can cause discrepancies in the effects of satisfaction, perceived behavioral control, and subjective norms on a user’s behavioral intention (Armitage et al., 2002; Liao et al., 2006). Other references also show that the effects of ease of use and perceived usefulness on behavioral intention can be moderated by gender difference (Gefen and Straub, 1997). Accordingly, further research may be needed to examine the moderating effect of gender difference on the behavioral performance of online customers.
Ajzen, I. (1991), "The Theory of Planned Behavior", Organizational Behavior and Human Decision Processes, Vol. 50 No. 2, pp. 179-211. Ajzen, I. (2002), "Perceived behavior control, self-efficacy, locus of control, and the theory of planned behavior", Journal of Applied Social Psychology, Vol. 32 No. 1, pp. 1-20. Ajzen, I. and Fishbein, M. (1980), Understanding attitudes and predicting social behavior., Englewood Cliffs, NJ: Prentice-Hall. Anderson, E.W. and Sullivan, M.W. (1993), "The antecedents and consequences of customer satisfaction for firms," Marketing Science, Vol. 12 No. 2, pp. 125-143. Anderson, J.C. and Gerbing, D.W. (1988), "Structural equation modeling in practice: A review and recommended two step approach", Psychological Bulletin, Vol. 103 No. 3, pp. 411-423. Armitage, C.J., Norman, P. and Conner, M. (2002), "Can the theory of planned behavior mediate the effects of age, gender and multidimensional health locus of control?", British Journal of Health Psychology, Vol. 7 No. 2, pp. 299-316. Bagozzi, R.P., Yi, Y. and Phillips, L.W. (1991), "Assessing Construct Validity in Organizational Research", Administrative Science Quarterly, Vol. 36 No. 3, pp. 421-430. Bearden, W.O. and Teel, J.E. (1983), "Selected determinants of consumer satisfaction and complaint reports", Journal of Marketing Research, Vol. 20 No. 1, pp. 21-28. Bhattacherjee, A. (2000), "Acceptance of Internet applications services: the case of electronic brokerages.", IEEE Transactions on Systems, Man, and Cybernetics --- Part A: Systems and Humans, Vol. 30 No. 4, pp. 411-420. Bhattacherjee, A. (2001), "Understanding information systems continuance: An expectation confirmation model", MIS Quarterly, Vol. 25 No. 3, pp. 351-370. Bhattacherjee, A. and Premkumar, G. (2004), "Understanding changes in belief and attitude toward information technology usage: a theoretical model and longitudinal test", MIS Quaterly, Vol. 28 No. 2, pp. 229-254. Cunningham, S.M. (Ed.). 1967. The major dimensions of perceived risk. In: Cox, D.F. (Ed.), Risk Taking and Information Handling in Consumer Behaviour.: Harvard University Press, Boston, MA. Davis, F.D. (1989), "Perceived usefulness, perceived ease of use, and user acceptance of information technology.", MIS Quarterly, Vol. 13 No. 3, pp. 319-340. Davis, F.D., Bagozzi, R.P. and Warshaw, P.R. (1989), "User acceptance of computer technology: a comparison of two theoretical models", Management Science, Vol. 35 No. 8, pp. 982-1003. Entrust (2008), "Pshing attack", http://www.entrust.com/phishing/index.htm, Vol. No., pp. Featherman, M.S. and Pavlou, P.A. (2003), "Predicting e-services adoption: a perceived risk facets perspective", International Journal of Human-Computer Studies, Vol. 59 No. 4, pp. 451-474. Fishbein, M. and Ajzen, I. (1975), Belief, attitude, intention, and behavior: An introduction to theory and research., MA: Addison-Wesley. Fornell, C. and Larcker, D.F. (1981), "Evaluating Structural Equation Models with Unobservable Variables and Measurement Error", JMR, Journal of Marketing Research, Vol. 18 No. 1, pp. 39-47. Forsythe, S.M. and Shi, B. (2003), "Consumer patronage and risk perceptions in Internet shopping", Journal of Business Research, Vol. 56 No. 2, pp. 867-875. Gefen, D., Karahanna, E. and Straub, D.W. (2003), "Trust and TAM in online shopping: An integrated model", MIS Quaterly, Vol. 27 No. 1, pp. 51-90.
Gefen, D. and Straub, D.W. (1997), "Gender differences in the perception and use of email: An extension to the technology acceptance model", MIS Quarterly, Vol. 21 No. 4, pp. 389-400. Hairs, J.F., Anderson, R.E. and Tatham, R.L. (1998), Multivariate Data Analysis, 5th ed. New York:Macmillian. Hsu, M.H., Yen, C.H., Chiu, C.M. and Chang, C.M. (2006), "A longitudinal investigation of continued online shopping behavior: An extension of the theory of planned behavior. ", International Journal Human-Computer Studies, Vol. 64 No. 6, pp. 889-904. Jacobs, P. (1997), "Privacy: what you need to know", InfoWorld, Vol. 19 No. 44, pp. 111-112. Jacoby, J. and Kaplan, L., B (1972), the components of perceived risk, Association for Consumer Research, Chicago. Kaplan, L.B., Szybille, G.J. and et al. (1974), "Components of perceived risk in product purchase: a cross validation", Journal of Applied Psychology, Vol. 59 No. 3, pp. 278-291. Kuisma, T., Laukkanen, T. and Hiltunen, M. (2007), "Mapping the reasons for resistance to Internet banking: A means-end approach", International Journal of Information Management, Vol. 27 No. 2, pp. 75-85. 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, Vol. 8 No. 3, pp. 130-141. Liao, C., Chen, J.-L. and Yen, D.C. (2006), "Theory of planning behavior (TPB) and customer satisfaction in the continued use of e-service: An integrated model", Computers in Human Behavior, Vol. 23 No. 6, pp. 2804-2822. Liao, S. (1999), "The adoption of virtual banking: an empirical study", International Journal of Information Management, Vol. 19 No. 1, pp. 63-74. Lin, W.-B. (2008), "Investigation on the model of consumers' perceived risk --- integrated viewpoint", Expert Systems with Applications, Vol. 34 No. 1, pp. 977-988. Littler, D. and Melanthiou, D. (2006), "Consumer perceptions of risk and uncertainty and the implications for behaviour towards innovative retail services: The case of Internet Banking", Journal of Retailing and Consumer Services, Vol. 13 No. 6, pp. 431-443. Lu, Y., Zhou, T. and Wang, B. (2009), "Exploring Chinese users' acceptance of instant messaging using the theory of planned behavior, the technology acceptance model, and the flow theory", Computers in Human Behavior, Vol. 25 No. 1, pp. 29-39. Mitchell, V.W. (1999), "Consumer perceived risk:conceptualizations and models", European Journal of Marketing, Vol. 33 No. 2, pp. 163-195. Oliver, R.L. (1980), "A cognitive model for the antecedents and consequences of satisfaction," Journal of Marketing Research, Vol. 17 No., pp. 460-469. Oliver, R.L. (1993), "Cognitive, Affective, and Attribute Bases of the Satisfaction Response", The Journal of Consumer Research, Vol. 20 No. 3, pp. 418-430. Oliver, R.L. and DeSarbo, W.S. (1988), "Response Determinants in Satisfaction Judgments", The Journal of Consumer Research,, Vol. 16 No. 4, pp. 495-507. Peter, J.P. and Ryan, M.J. (1976), "An investigation of perceived risk at the brand level", Journal of Marketing Research, Vol. 13 No., pp. 184-188. Podsakoff, P.M., MacKenzie, S.B. and Lee, J.Y. (2003), "Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies", Journal of Applied Psychology, Vol. 88 No. 5, pp. 879-903. Reavley, N. (2005), "Securing online banking", Card Technology Today, Vol. 17 No. 10, pp. 12-13.
Roselius, T. (1971), "Consumer Rankings of Risk Reduction Methods", Journal of Marketing Vol. 35 No. 1, pp. 56-61. Soopramanien, D.G.R. and Robertson, A. (2007), "Adoption and usage of online shopping: An empirical analysis of the characteristics of ''buyers'' ''browsers" adn "non-internet shoppers"", Journal of Retailing and Consumer Services, Vol. 14 No. 1, pp. 73-82. Szymanski, D.M. and Henard, D.H. (2001), "Customer satisfaction: a meta-analysis of the empirical evidence", Journal of the Academy of Marketing Science, Vol. 29 No. 1, pp. 16-35. Taylor, S. and Todd, P.A. (1995a), "Assessing IT usage: the role of prior experience", MIS Quarterly, Vol. 19 No. 4, pp. 561-570. Taylor, S. and Todd, P.A. (1995b), "Understanding information technology usage: a test of competing models", Information Systems Research, Vol. 6 No. 2, pp. 144-176. Venkatesh, V. (2000), "Determinants of perceived ease of use: integrating control, intrinsic motivation, and emotion into the technology acceptance model.", Information Systems Research, Vol. 11 No. 4, pp. 342-365. Venkatesh, V. and Davis, F.D. (2000), "A theoretical extension of the technology acceptance model: four longitudinal field studies", Management Science, Vol. 46 No. 2, pp. 186-204. Venkatesh, V., Morris, M.G., Davis, G.B. and Davis, M.D. (2003), "User acceptance of information technology: Toward a unified view", MIS Quarterly, Vol. 27 No. 3, pp. 425-478. Woszczynski, A.B. and Whitman, M.E. (Eds.). 2004. The problem of common method variance in IS research: The Handbook of Information Systems Research. Idea Group Inc., Hershey, PA, pp.66-77. Wu, I.L. and Chen, J.L. (2005), "An extension of Trust and TAM model with TPB in the initial adoption of online tax: An empirical study", International Journal of Human-Computer Studies, Vol. 62 No. 6, pp. 784-808.
Appendix A. The questionnaire
Constructs Questionnaire items Pre-usage perceived usefulness (adapted from Davis (1989)) If I were to adopt online shopping, it would enhance my efficiency. If I were to adopt online shopping, the effect of purchasing goods or services would improve. If I were to adopt online shopping, it would enable me to improve the ability of a commercial transaction. Post-usage perceived usefulness (adapted from Bhattacherjee (2001)) Using online shopping help me purchase products/services efficiently. I find online shopping useful in my life. In my opinion, using online shopping increase my effectiveness in buying products/services. Perceived ease of use (adapted from Davis (1989)) I’ll find using online shopping easy for me. Learning how to use online shopping will be easy for me. It’ll be easy to find products that I want to buy using online shopping. Attitude (adapted from Davis (1989)) Using online shopping is a good idea. I feel good about using online shopping. I like using online shopping. Intention (adapted from Davis (1989)) I intend to purchase online. It will be worth to shop online. It will be desirable to use online shopping. Perceived risk (adapted from Lee (2009)) I think that online shopping is risky because it may lead to financial loss for me. I think that online shopping is risky because it may incur fraud or the hacker invades. I think that online shopping is risky because it may cause identity theft. I think that online shopping is risky because the products/services delivered may fail to meet to my expectations. I think that online shopping is risky because it may lead to a time loss for me. Subjective norm (adapted from Taylor & Todd (1995b)) People important to me support my use of online shopping. People who influence me think that I should use online shopping. People whose opinions I value prefer that I should use online shopping Perceived behavioral control (adapted from Taylor & Todd (1995b)) Using online shopping was entirely within my control. I had the resources, knowledge, and ability to use online shopping. I would be able to use online shopping well. Confirmation (adapted from Bhattacherjee (Bhattacherjee, 2001)) My experience with using online shopping was better than I expected. The service level provided by online shopping was better than I expected. The online shopping can meet demands in excess of what I required for the products/services. Satisfaction (adapted from Bhattacherjee (2001)) I am satisfied with the performance of online shopping. I am pleased with the experience of using online shopping. My decision to use online shopping was a wise one. Continuance Intention (adapted from Bhattacherjee (2001)) I intend to continue using online shopping on a regular basis in the future. I will frequently use online shopping in the future. If I could, I would like to discontinue my use of online shopping (reverse coded).
Table 2: Construct reliability and convergent validity
Construct/ Indicators Pre-usage perceived usefulness Intention Item PU1 PU2 PU3 INT1 INT2 INT3 PEOU1 PEOU2 PEOU3 AT1 AT2 AT3 CI1 CI2 CI3 SN1 SN2 SN3 PBC1 PBC2 PBC3 IT1 IT2 IT3 CF1 CF2 CF3 FR1 FR2 FR3 PR1 PR2 PR3 PR4 Factor loading 0.855 0.826 0.827 0.844 0.833 0.832 0.846 0.843 0.865 0.848 0.851 0.869 0.857 0.845 0.874 0.817 0.862 0.860 0.937 0.861 0.870 0.927 0.907 0.902 0.769 0.815 0.773 0.876 0.846 0.844 0.873 0.839 0.833 0.817 t-value 19.110 18.339 19.516 13.433 15.232 15.223 19.518 18.978 19.267 19.168 18.838 18.764 12.531 11.312 9.334 18.537 18.941 18.857 19.801 19.139 19.419 20.503 21.100 20.906 9.863 10.583 11.676 17.789 18.179 17.654 16.705 16.916 13.443 12.876 Composite reliability (CR) 0.91 Average variance extracted 0.7335 Cronbach’s alpha
Perceived ease of use
Perceived behavioral control
Post-usage perceived usefulness
Table 3: Correlation Matrices and Discriminant Validity
Construct Pre-usage perceived usefulness 0.833 0.33 0.53 0.21 0.24 0.26 0.31 0.27 0.28 0.43 -0.35 Post-usage perceived usefulness Perceived ease of use Subjective Norm Perceived Behavior control Attitude Initial intention Continued Intention Confirm -ation Satisfaction Perceived risk
Pre-usage Perceived usefulness Post-usage perceived usefulness Perceived ease of use Subjective norm Perceived behavioral control Attitude Initial intention Continuance intention Confirmation Satisfaction Perceived risk
0.855 0.43 0.41 0.34 0.63 0.22 0.69 0.52 0.61 -0.23 0.867 0.32 0.34 0.51 0.42 0.55 0.43 0.23 -0.24 0.851 0.35 0.54 0.51 0.67 0.12 0.20 -0.23 0.856 0.39 0.31 0.51 0.22 0.20 -0.26 0.85 0.42 0.62 0.54 0.58 -0.62 0.84 0.53 0.23 0.33 -0.31 0.90 0.51 0.24 -0.24 0.88 0.23 -0.21 0.77 -0.02 0.85
Note: All correlations significant at p< 0.05 except where noted. Diagonal elements are square roots of average variance extracted
Table 4: Summary of hypotheses tests
Hypotheses H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 satisfaction → continuance intention confirmation → satisfaction perceived usefulness → satisfaction post-usage perceived usefulness →continuance intention confirmation → post-perceived usefulness pre-usage perceived usefulness → attitude pre-usage perceived usefulness → initial intention perceived ease of use → attitude perceived ease of use → pre-usage perceived usefulness attitude → initial intention subjective norm → initial intention perceived behavior control → initial intention perceived risk → initial intention perceived risk → continuance intention
0.53 0.31 0.22 0.37 0.33 0.11 0.21 0.31 0.41 0.19 0.22 0.25 -0.41 -0.11
p-value 0.008** 0.007** *** *** 0.04* 0.02* 0.04* 0.03* 0.04* 0.02* 0.04* 0.002** 0.002** 0.14
Results Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No
a. Standardized estimates are shown. b. * p 0 .05 , * * p 0 .01, * * * p 0 .001
Table 5. Direct, indirect and total effects --- Estimates
Intention C r i t Direct effects 0.183*** Indirect effects Total effects Direct effects
Satisfaction Indirect effects Total effects 0.22***
Continuance Intention Direct effects 0.37*** Indirect effects 0.12* Total effects 0.49*
Post-usage perceived usefulness Confirmation Satisfaction Perceived risk Perceived ease of use Pre-usage perceived usefulness Attitude Subjective norm Perceived behavioral control *
0.31** -0.41** 0.15* 0.21* 0.02* 0.15* 0.23*
0.19* 0.22* 0.25*
0.19* 0.22* 0.25* 0.11* 0.103** 0.11* 0.103**
significant at p
< 0.05. < 0.01.
** significant at p
*** significant at p < 0.001
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.