رضایت کاربر از سیستم های پشتیبانی تصمیم گیری مبتنی بر وب : نقش سوابق شناختی

Websites play a critical role in attracting customers and providing information to assist Web customers in decision making. Despite the importance of such systems providing a satisfactory experience, as per our knowledge no effort has been made to systematically examine the underlying causal structure among important decision related variables such as perceived effort, perceived information accuracy, perceived effectiveness, and satisfaction using Web-based decision support systems (Web-based DSS) that provide information to potential customers. In this study, we develop a conceptual model for investigating cognitive antecedents to Web users’ satisfaction in the context of Web-based DSS. The empirical examination of the research model using structural equations modeling indicates that perceived effectiveness is influenced by perceived information accuracy and effort, and in turn, has positive impact on satisfaction in using Web-based DSS. The implications of this study suggest to information managers that providing customers with highly accurate information from web sites requiring lower effort is perceived as an effective system and such perception leads to increased satisfaction.

The Internet has enabled retailers to exploit opportunities provided by the World Wide Web, such as rapid dissemination of up-to-date information to customers, and it has given retailers control over the formatting and aggregation of information the customers receive (Shim et al., 2002). In the rush to establish an online presence, however, many e-commerce Web sites were developed with little attention to the features that would enhance consumers’ satisfaction with their Web-informed decision making process. Over 80% of Web shoppers have left electronic markets without knowing what they want, and 23% of attempted transactions ended up in failure, and, furthermore, only about 2% of the users who visit a Web site end up making a purchase (Silverman, Bachann, & Al-Akharas, 2001). Nonetheless, the growth of Web retailing is compelling the Web retailers to be competitive in the way they provide product information to consumers. Therefore, designing Web sites that facilitate consumers’ decision making processes so that they are effective and satisfying is becoming an important concern for Web retailers. Most prior studies of decision support systems (DSS) examined the role of DSS as a contributor to decision performance (Silver, 1991 and Todd and Benbasat, 1992). More precisely, some studies investigated decision strategies for processing information and found the trade-off relationship between accuracy and effort (Chenoweth et al., 2004 and Johnson and Payne, 1985). Sharda, Steve, & McDonnel (1988) and Lilien, Ragaswamy, Van Bruggen, & Starke (2004) studied DSS effectiveness. Vessey and Galletta (1991) proposed the cognitive fit perspective for understanding the use of information displays in decision making. Parikh, Fazlollahi, & Verma (2001) and Silver (1991) examined how DSS functionalities enlighten users’ decision-making processes. Todd and Benbasat (1991) used cognitive fit model of decision making to show that decision makers’ goal to maximize decision performance depends on the effort required to implement various decision strategies. Such studies that emphasized decision performance typically downplayed or ignored the level satisfaction of the DSS users. With the growth of competitive Web retailing, however, understanding how to satisfy customers who obtain their information from web sites will be critical for establishing long-term client relationships, which consequently increases profitability (McKinney, Yoon, & Zahedi, 2002). Therefore, understanding how Web users formulate satisfaction when obtaining information is of great importance to Internet business. However, based on our knowledge, there are very few studies that systematically examined cognitive antecedents to satisfaction of users obtaining their information from Web-based DSS. In this study, we advance a model of the decision making process while obtaining information using DSS in the Internet environment to address the following research question: What are the antecedents of Web-based DSS users’ satisfaction? To answer this question, we propose and test a research model which hypothesizes that perceived information accuracy and effort influence Web users’ perceived effectiveness, which in turn affects Web users’ overall satisfaction. This study was conducted in three phases to test our research model. In the first phase, from representative decision makers we identified important decision variables in our decision context (apartment selection) and obtained actual data to create realistic decision tasks. In the second phase, we developed three different Web sites, which varied in their effort demands and potential for perceived information accuracy vis-à-vis their user's preferences. In the third phase, we developed an instrument based on prior literature to measure important constructs, such as Web users’ satisfaction with Web-based DSS and its antecedents. The model was tested in a controlled laboratory experiment. Structural equation modeling was used for data analyses. In what follows, the next two sections provide background and propose the research model and hypotheses for the study. The subsequent two sections report research design and data analyses. In the final section, we provide the conclusion of this study.

The growth of online retailing on the Internet has been phenomenal over the last few years. In the context of heightened competition, it is important for Web retailers to provide provide information in a way that supports consumers’ decision making, and futher, leads to the web-site user's satisfaction. This research is motivated by the assumption that billions of dollars are at stake in Web retailing, and that Web retailers run the risk of inadvertently losing customers if they do not understand the antecedents to the user satisfaction. Web-site customers must confront the situation in which they face a barrage of information of various types, sources, and formats, highlighting the need to help the user address the prospect of information overload. Ostensibly Web-based DSS could help. Generalizing from prior research on offline DSS may fall short, however, because it, does not sufficiently explain the situation of web-site users. Most prior research on offline DSS focused on whether or how the DSS could improve performance relative to not using the DSS, but was not concerned with the users’ satisfaction. That research appears to have been predicated on the assumption that either managers could force the use of DSS if the managers thought they were effective or that users would automatically recognize the advantage of using an effective DSS (versus not using it). Thus, the offline DSS research did not specifically address the users’ satisfaction, nor its antecedents. For these reasons we see the merit of extending prior DSS research and for treating Web-based DSS as a special entity worthy of study in its own right. We contribute to the literature of electronic commerce by shifting focus from researching the effects of users’ cognitive processing of strategies while using a DSS, to researching users’ perceptions of the experience of using Web delivered DSS to obtain information. Our research contributes to the IS literature by proposing and testing a model that explicitly includes perceived effectiveness as an intervening variable between perceived effort and perceived information accuracy and user satisfaction. This insight was gained because of the DSS context, which has to deal with the user's trade-off between perceived effort and perceived information accuracy, implying that the user must reach a combined assessment of the two, and that this combined assessment is, in essence, the perceived effectiveness of the system. We find that perceived effectiveness and perceived information accuracy influence Web DSS user satisfaction. Interestingly, our hypothesized relationship between perceived effort and perceived information accuracy was not supported. Our hypothesis was derived from prior research findings that making an accurate feature easier to use increased the likelihood that it would be used. This causal linkage might not have been salient to our participants, who were asked to use the Web DSS feature whether it was easy or not. Given the importance of having satisfied Web customers, this study suggests an opportunity for Web retailers to provide features to assist consumers in obtaining information to make choices. Since such decision aiding features may have differential effects on perceived effort and perceived information accuracy it is worthwhile for Web retailers to measure the impact of new features on these constructs. As a starting point, we recommend assigning to the computer the things the computer is best at (Jones & Brown, 2002) such as executing built-in algorithms to incorporate user preferences by eliciting weights from users and converting them into a final score for each product. Every consumer has preferences concerning the attributes that go into making a choice and the effectiveness evaluations and satisfaction formulated out of using Web DSS will be influenced by the extent to which a Web DSS provides recommendations that consider and integrate consumers’ preference structure. Our study found that 73% of variation in the satisfaction construct is explained by the variation in the effectiveness and accuracy constructs. Our findings suggest a warning for the retailers that provide customer web sites. The retailers should not venture to build Web-DSS features without finding out if consumers are actually satisfied with the features. If the level of satisfaction turns out too low, this study provides theoretical constructs that could be used to derive and test features for possible changes in the way information is delivered on the web site that would increase the level of customer satisfaction. Web shoppers can use features provided on a web site to formulate search criteria, do the actual search and finally design a set of alternatives to choose from. The web site can then provide different levels of support to enable the decision maker to make the final choice. This could range from providing just the attribute-by-alternative comparison matrix to eliciting user preferences to recommend products. Making decisions on the Web, however, can be a frustrating experience because of the huge amounts of information available about any product. For a decision maker, cognitively processing information of numerous brands and attributes might be a difficult task. Given the role of information load on the decision maker to come up with relevant alternatives and attributes, determining how several cognitive variables shape satisfaction with the Web-based DSS becomes an important question. Our investigation lays out valuable findings, but we experienced some limitations, which could lead to further studies. First, this study did not investigate the relationship between satisfaction with a Web DSS and intention to purchase. Such type of research would be of great interest to web retailers but was beyond the scope of this study which was focused on the more direct effect of obtaining information on the web site. Second, the psychological factors investigated in this study need to be decomposed to the next level of detail to attain greater understanding of the underlying cognitive processes. For example, decision makers’ level of satisfaction can be altered depending upon the tasks they want to accomplish. Therefore, different levels of tasks and associated psychological constructs need to be investigated to provide detailed theoretical frameworks. Third, future study needs to examine the effects of different Web-based DSS functionalities on the psychological constructs. Generalizing the categories of Web-based DSS functionalities and their impacts would be of great contributions to practitioners.