Integrative Framework

---

Figure 2 presents an integrative framework for online consumer behavior research. The framework builds upon prior research and integrates research findings across studies to develop a coherent and comprehensive understanding of the online consumer behavior research conducted in the IS field. The framework is also grounded on several theoretic perspectives developed outside of the online consumer behavior research, such as IS success model (DeLone & McLean 1992; Seddon 1997), SERVQUAL (Pitt et al., 1995), and TAM (Davis, 1989; Davis et al., 1989). Table 5 provides a listing of representative studies that support the linkages contained in the integrative framework. Each element of the framework and relationships between them are further.

Table 5: Empirical Literature Supporting Integrative Framework

Relationship Between Variables	Representative Studies
System Quality → Ease of Use	Otto et al., 2000
System Quality → Usefulness	Han and Noh, 2000
System Quality → Trust	Grazioli and Jarvenpaa, 2000
System Quality → Shopping Enjoyment	Koufaris, 2002
Information Quality → Ease of Use	Agarwal and Venkatesh, 2002
Information Quality → Usefulness	Agarwal and Venkatesh, 2002
Vendor and Channel Characteristics → Usefulness	Liao and Cheung, 2001
Vendor and Channel Characteristics → Trust	Jarvenpaa et al., 2000
Ease of Use → Web Use	Chen et al., 2002; Moon and Kim, 2001
Ease of Use → Online Purchase	Chen et al., 2002; Devaraj et al., 2002
Ease of Use → Post Purchase	Devaraj et al., 2002
Usefulness → Web Use	Chen et al., 2002; Moon and Kim, 2001
Usefulness → Online Purchase	Chen et al., 2002; Devaraj et al., 2002
Usefulness → Post Purchase	Devaraj et al., 2002
Trust → Online Purchase	Bhattacherjee, 2002; Grazioli and Jarvenpaa, 2000; Jarvenpaa et al., 2000
Shopping Enjoyment → Online Purchase	Koufaris, 2002; Koufaris et al., 2002
Shopping Enjoyment → Post Purchase	Koufaris, 2002; Koufaris et al., 2002

Figure 2: Integrative Framework

Dependent Variables

Consistent with our review of the online consumer behavior research, the framework groups dependent variables into three categories: web use, online purchase, and post-purchase. Studies have examined these behaviors independently or in combination with each other. However, an interesting aspect that has not been explicitly addressed in literature is the relationship between these behaviors. The framework proposes significant links between web use and online purchase, between online purchase and post-purchase, and between post-purchase and use. First, frequent use of the system is likely to lead to online purchasing. Companies on the Internet try to increase traffic and make their web sites "sticky", so that users can spend more time on the Web. Liang & Lai (2002) report that consumers are more likely to shop at well-designed web sites. As the customer has to interact with the system to execute an online purchase, the use of the web site is a crucial precursor to online purchase. Second, online purchase is likely to become a repeated pattern of behavior if customers are satisfied with their purchase. Online purchasing offers the opportunity to assess the quality of product and vendor service, as well as to experience the convenience of online transactions. Thus, the experience from the purchase becomes a determinant of post-purchase decision variables such as channel preference, switching, attrition, and re-visitation. Finally, post-purchase is likely to influence the level of web use. Customers need to resolve post-purchase issues, receive technical support, and check product updates through the use of the Web. Further, as they are satisfied with the purchase, they will continue using the system to repeat the purchase.

Mediating Perceptual Variables

Extending the prescriptions of the TAM, which theorizes usefulness and ease of use as fundamental mediating perceptions through which external factors influence usage behavior (Davis et al., 1989), the framework conceptualizes usefulness, ease of use, trust, and shopping enjoyment as perceptual variables that mediate the effects of system quality, information quality, service quality, and vendor and channel characteristics. Studies based on Flow Theory have found shopping enjoyment as a mediator between various predictor variables and intention to return (Koufaris 2002; Koufaris et al., 2002). Trust-related literature emphasizes trust as a key mediating variable (Bhattacherjee, 2002; Gefen et al., 2003; Grazioli & Jarvenpaa, 2000; McKnight et al., 2002). Relationships between the mediating variables have also been found. The relationship between usefulness and ease of use is well established (Davis et al, 1989; Venkatesh & Davis, 2000). Gefen et al. (2003) found that ease of use, trust, and usefulness are related. However, the relationship between shopping enjoyment and usefulness, while implicitly referred to, has not been empirically examined.

Predictor Variables

Based on our review of the studies and the theoretical perspectives presented earlier, we propose that many variables used as predictors of online consumer behavior can be classified into system quality, information quality, and service quality. Other factors such as vendor and channel characteristics, consumer demographics and traits, and the social context of the consumer were also addressed in the studies and are included in the integrative framework.

System Quality

System quality captures the user perceptions regarding the effectiveness of system attributes. The infusion of technology in the interaction between the consumer and the vendor increases the importance of the technology-enabled interface with which the consumers have to interact. Navigation, interface layout, download speed, digital seals, and value added mechanisms are some factors that constitute the notion of system quality (Han & Noh, 2000; Liao & Cheung, 2001; Liu & Arnett, 2000; Westland & Au, 1998).

Koufaris (2002) has found support that value added search mechanisms play a significant role in shaping consumer's intention to return to the web site and shopping enjoyment partially mediates the effect. Anecdotal evidence suggests that high performing companies are actively pursuing enhancements in web site features and services that facilitate the consumer purchase experience (Zbar, 2000). Schubert & Selz (1997) structure an extensive list of web site features into three phases common to purchase transactions (information phase, agreement phase, and settlement phase). TAM suggests that system features affect use through the perceptions of ease of use and usefulness (Davis et al., 1989). Therefore, the framework proposes that system quality influences online consumer behavior by altering consumer perceptions of ease of use, usefulness, trust, and shopping enjoyment.

Information Quality

Information quality captures the perceptions of the consumer regarding the characteristics of the web site content, such as accuracy, comprehensiveness, reliability, relevance and usefulness. Agarwal & Venkatesh (2002) found that content was equally important across industries (books, airline, car rental, and automotive) and tasks (customer and investor). Although studies suggested that information quality was an important determination of use and user satisfaction, its impact on purchase or post-purchase behavior was found to be rather moderate. For example, Palmer (2002) and McKinney et al. (2002) found that information quality impacted use, while Ranganathan & Ganapathy (2002) concluded that content was the least important discriminator between subjects with low intent and high intent to purchase. A possible reason for such findings could be the underlying task or product. For example, content may be a dominant factor in the context of web sites that provide information-based services (news, search, legal counseling, article delivery, etc.), while its role in purchasing physical products may be moderate. Liu & Arnett (2000) found a high correlation between information quality and learning capability (r = .72, p < .001). Overall, prior research findings suggest that information quality is an important predictor of online consumer behavior, and its effect may be mediated by user perceptions of usefulness and ease of use.

Service Quality

Service quality measures the perceptions of the consumers regarding their service experience. The peculiar nature of the technology in question (the web site) and the context (online consumer behavior) creates complexity in application of service quality in electronic channels. This issue is also prevalent in the context of other information systems, as pointed out by Seddon (1997) that the system and the IS department are two different entities. Thus, a distinction needs to be made regarding who is providing the service. If the service is being provided by the web site, the elements of service quality such as tangibility, reliability and responsiveness will tend to overlap with system quality. However, if the vendor provides the service, service quality should emerge as a distinct factor. This may be one of the reasons for contradictory findings in the studies on dimensions of SERVQUAL. Consequently, we recommend that researchers make a clear distinction regarding the context and apply SERVQUAL with caution. Most prior studies operationalized SERVQUAL as a set of service functions of a web site. In our framework, we conceptualize SERVQUAL as vendor's effectiveness in providing customer service, rather than web site's effectiveness in providing service functions. When a vendor's service quality changes, it is likely to change user perceptions of trust and usefulness, thereby changing the users intentions to buy online. Thus, the framework proposes that the service quality of the vendor influences online consumer behavior through its effects on trust and usefulness perceptions.

Vendor and Channel Characteristics

Vendor characteristics such as vendor competence, size, reputation, and participation costs have shown consistent results across different studies (Chen & Hitt, 2002; Jarvenpaa et al., 2000). Vendor characteristics such as size and reputation enhance consumer perceptions regarding trust or the integrity of the vendor. Thus, brand issues seem to be as prevalent, if not stronger, in an online context as they are in an offline channel. These results raise concerns regarding the assertions that the Internet provides a level playing field for the companies. It is argued that electronic markets may be more efficient than offline markets (Devaraj et al., 2002). The main arguments presented in favor of such an assertion is that the Internet reduces search costs and makes the delivery processes more efficient, thus resulting in low prices for products. Empirical results show that lower prices play an important role in channel choice decisions (Devaraj et al., 2002; Liao & Cheung, 2001; Liang & Huang, 1998). Furthermore, prior research found that price differentials between online and offline channels (Devaraj et al., 2002) and participation costs (Chen & Hitt, 2002) influenced online behavior. The framework proposes that these characteristics of the vendor and channel impact online consumer behavior by enhancing vendor trust and perceived usefulness of the channel.

Consumer Demographics and Personal Traits

Factors that constitute demographics and personal traits have either been modeled as facilitating factors of certain types of perceptions or as factors that moderate the relationships between the independent and dependent variables. Three important findings have emerged concerning demographics. First, women have been found to be more conservative customers with respect to electronic channels (Slyke et al., 2002). Multiple arguments have been presented for these results. For example, women view shopping as a social activity, and show conservatism toward trying a new technology. Second, lifestyle has been suggested as an important variable. Researchers have found that a wired lifestyle (Bellman et al., 1999) and a net-oriented lifestyle (Kim et al., 2000) play a significant role in determining consumer behavior. Finally, studies have concluded that demographics overall do not explain much variance in behavior (Bellman et al., 1999; Chen & Hitt, 2002). Personal traits such as personal innovativeness, web skills, computer self-efficacy, and affinity with a computer have consistently been found to be significant variables across the studies.

Social Context Variables

Social context consists of external influences (mass media, advertising, and marketing related stimuli) and interpersonal influences (word-of-mouth, friends, relatives and other sources involving a consumer's social network) (Agarwal & Venkatesh, 2002; Limayem et al., 2000; Parthasarathy & Bhattacherjee, 1998). Results were, for the most part, supportive of the significant effects of social context variables on the use of the Internet and online shopping intention. Outside of the online consumer behavior context, Venkatesh & Davis (2000) showed that social norm affected intention to use partially via perceived usefulness. Consistently, the framework includes social context variables as determinants of both perceptual variables and consumer behavior.

---