An Empirical Study of the Impact of Product ... - Semantic Scholar

IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART A: SYSTEMS AND HUMANS, VOL. 36, NO. 6, NOVEMBER 2006

1185

An Empirical Study of the Impact of Product Characteristics and Electronic Commerce Interface Richness on Consumer Attitude and Purchase Intentions Jungjoo Jahng, Hemant K. Jain, Member, IEEE, and K. Ramamurthy

Abstract—Although there has been a pickup in the growth of business-to-consumer electronic commerce (EC) recently, the overall growth appears to have failed to live up to the various forecasts. Notwithstanding the concerns of security and privacy, this paper posits that for complex products and services offered on the web, existing EC interfaces (ECIs) lack the necessary “richness” to overcome the virtual nature that consumers face compared with a physical store. We empirically validate a conceptual framework that exists in the literature to address these issues of ECI design by drawing on insights from contingency research in information system (IS) design and media choice behavior. This paper finds that a fit or congruence between the product type and the ECI richness (in product information representation, and product-related interaction and communication between the consumer and vendor) leads to superior outcomes. Specifically, buying a complex product (e.g., digital camera) on the Web requires a greater ECI richness in the form of experiencing higher social presence and/or higher product presence. On the other hand, a lean ECI is sufficient for a simple product (e.g., diskette). The role of tolerance for ambiguity (TA), which is an individual difference variable, in this relationship is also examined. A significant threeway fit of TA with ECI richness and product types on consumer behavior is found. A number of implications and future research extensions are discussed. Index Terms—Consumer attitude, consumer behavior, EC, EC interface (ECI) richness, electronic commerce (EC) environment richness, fit, product information representation richness, product presence (PP) experience, social interaction richness, social presence (SP) experience.

I. I NTRODUCTION

E

LECTRONIC commerce (EC) has gained enormous attention from a number of researchers as well as practitioners because of its significant potential for and impact on both businesses and consumers. However, the tremendous growth forecasts, especially of business-to-consumer (B2C) EC, have

Manuscript received January 21, 2004; revised December 29, 2004, May 20, 2005, and June 9, 2005. This work was supported in part by the University of Wisconsin–Milwaukee, and in part by the Institute of Management Research at Seoul National University, Korea. This paper was recommended by Associate Editor N. Cassaigne. J. Jahng is with the College of Business Administration, Seoul National University, Seoul 151-742, Korea (e-mail: [email protected]). H. K. Jain is with the Wisconsin Distinguished and Tata Consultancy Services, School of Business, University of Wisconsin–Milwaukee, Milwaukee, WI 53201 USA (e-mail: [email protected]). K. Ramamurthy is with the School of Business, University of Wisconsin– Milwaukee, Milwaukee, WI 53201 USA (e-mail: [email protected]). Digital Object Identifier 10.1109/TSMCA.2006.878977

not been realized, which is aggravated further by the burst of the so-called Internet bubble. One reason advanced for the underrealization of the forecasted growth in B2C EC is that the EC environments/interfaces [EC interfaces (ECIs)] generally lack the richness of the product-purchase experience available in a physical store [20]. Unlike in a physical-commerce environment where consumers can, if required, touch and feel the products and talk to a “real person” for additional product information or consultation, EC is virtual. However, with further expansion of the EC to mainstream consumers, the virtual nature of the environment invites a greater focus on these interaction difficulties, which is exacerbated even more when the ECI is poorly designed. Previous research emphasizes the need for an improved product experience [21], [22] and proposes the concept of parasocial presence [26]. Given that the ECI is often the only point of contact between consumers and providers of goods/services, an appropriate design of these interfaces to facilitate communication, interaction, and the relationship with consumers becomes a key factor. A deficiency in the interface could trigger a lack of motivation on the part of the consumers and result in lower consumer satisfaction and reduced purchase intention. In addition, the extent of use and success of an EC varies widely from one industry to another [18]. For example, industries such as travel EC have had significant success, while in jewelry and watches, sales are low, indicating that certain types of products/services with certain types of characteristics are more easily able to overcome difficulties posed by the virtual nature of the environment while others may require a much richer EC environment. Thus, we argue that ECI design should consider product characteristics and should be designed to compensate for the deficiency caused by the virtual nature of the environment to have a favorable influence on consumer attitude and behavior. A number of studies [25], [31] have addressed the design of Web-based EC application systems and identified many design features. However, the issues related to the virtual nature of the environment, the relationship among the product type, ECI and consumer characteristics, and their impact on consumer behavior have not been studied. A previous study [20] addresses these issues and proposes a theoretical framework. But, that study attempted to support its model by merely using some personal evaluations of a few real-world websites. While this paper uses a framework similar to the framework used in [20], we have designed and administered a very carefully controlled

1083-4427/$20.00 © 2006 IEEE

Authorized licensed use limited to: California State University Fullerton. Downloaded on October 4, 2009 at 23:26 from IEEE Xplore. Restrictions apply.

1186


experimental study using real subjects; such controls appear to be completely lacking in the above-cited study. Thus, this paper attempts to extend the above-cited study by empirically validating their framework to address issues of ECI design. To do so, we draw on insights from a contingency research in information-systems (IS) design and media choice behavior; specifically, we draw upon task-technology fit (TTF) [15], social presence (SP) [41], and media-richness [6] theories to develop our model and investigate the impact of product characteristics and ECI richness on the consumer attitude and behavior. The contributions of this paper, as would be evident from the results presented later, are to: 1) validate an “EC interface typology” along two dimensions—product-information presentation richness and product-related interaction richness; 2) point out the shortcomings of the existing product-categorization schemes and establish that “product presence” (PP), and “social presence” requirements are reasonable for the emerging EC environments; and 3) demonstrate that a fit/congruence among the ECI, product type and one consumer characteristic, tolerance for ambiguity (TA), leads to favorable consumer outcomes in the form of attitude and purchase intentions. This paper proceeds as follows. The next section presents the theoretical basis for this paper and proposes a research model. Section III describes the research methods used to validate the research model and its propositions. In Section IV, the research findings are presented, and the validity of the model is evaluated. Implications for research and practice, and limitation and future research directions are discussed in Section V. II. T HEORY AND M ODEL D EVELOPMENT Since a reasonable portion of some of the theoretical foundations used here is also present in [20], we will focus more on those aspects of theory that have been underemphasized. In this section, we first present a brief review of the relevant aspects of TTF, SP, and media-richness theories and discuss how these theories serve as the foundations for our study. We then present the relevant “product classification schemes” that exist in marketing and EC literatures. This is followed by the presentation of the research model of this paper and propositions. TTF is defined as the degree to which a technology assists an individual in performing his/her tasks [15]. The essence of TTF theory is that positive (or negative) impacts of technology on individual performance depend on whether that technology has (or does not have) a fit with the task that it supports. The application of TTF in individual ISs [15], data and information representation [30], and group-support-system [51] contexts shows that the impact of these technologies on the user outcomes depends on the level of fit with the tasks. The application of the key tenets of TTF theory, which stated that task characteristics should be carefully examined so that appropriate technology support is provided to the design of ECIs, appears to be logical. Furthermore, the TTF theory (and its findings of relationship with performance) would appear to be a key foundation for explaining the impact of fit between the product characteristics and ECI on consumer outcomes. SP theory (SPT) attempts to explain why certain communication media are chosen for specific types of interaction between

people and how well the media fit the information requirements of the task. Short et al. [41] propose that communication media capable of transmitting greater amounts of varied information simultaneously (such as facial expression, direction of looking, posture, dress, and other nonverbal information—cues) possess greater SP. Such media are friendlier, more personal, evoke emotion, and permit users to perceive others as being psychologically present. Media such as face-to-face meetings and telephone have been noted to be high in SP, whereas media such as (postal) mail services, e-mail, and fax have been found to be relatively low in SP. Low SP can lead to problems in interpreting communicated information, which is a phenomenon related to nonverbal cues getting filtered out. SPT suggests that media with high SP (e.g., face-to-face interaction) should be selected for tasks requiring higher levels of interaction (e.g., explaining features of complex products/services). On the other hand, for tasks such as simple information exchange, the medium’s SP is not as important as its efficiency [43]. Although SP is regarded as a quality of the communication medium in [41], they conceive it as both an objective and a subjective quality of the medium. They point out that users of any given communication medium are, in some sense, aware of the degree of SP offered by the medium; when they perceive an inadequacy in the SP of the medium for the task at hand, they would be dissatisfied and tend to avoid taking the next set of actions (purchase intention in our case). The study in [26] reconceptualizes SP as para-SP to capture the relationship between a website and her visitor. According to this paper, para-SP symbolizes the distance between a customer and the company the site represents. A B2C website can support features such as e-mail, real-time chat, and real-time one-way or two-way video-conferencing capabilities. Thus, the SP constructs can be used to represent the richness of the communication provided by the ECI. Media (information)-richness theory (MRT) posits that individual performance can be improved by matching media characteristics to the needs of information processing tasks [6]; this also ties in with both TTF and SP theories. MRT argues that depending upon the level of uncertainty and equivocality present in the task, certain media are better able to communicate information [6], [8]. Uncertainty in a task exists when a framework for interpreting a message is available, but there is a lack of adequate information to process [7]. Equivocality of tasks, on the other hand, refers to the ambiguity and the existence of multiple and potentially conflicting interpretation of the situation from the available information. In such instances, richer media are preferred [6]. A recent research [30], employing students within a laboratory setting engaged in decision making, reports that while text-based as well as multimedia representations are equally effective in reducing the equivocality levels for analyzable tasks, only multimedia (a richer) representation is capable of reducing equivocality for less-analyzable tasks. In the context of B2C ECI design, multimedia and virtual reality (VR) capabilities can be used to effectively represent the product information in order to improve the customers’ understanding of the product features and to provide a virtual product experience. This richer representation can better convey information, especially nonverbal messages, which can


JAHNG et al.: EMPIRICAL STUDY OF THE IMPACT OF PRODUCT CHARACTERISTICS AND ECI RICHNESS

facilitate the understanding of the information related to equivocal, ambiguous, and unstructured tasks [22], [30]. In addition, richer communication capability can be provided on the site to allow consumers to interact with sales associate to help answer questions and explain product features. Thus, from the foregoing brief discussion, TTF, MRT, and SPT theories offer a strong foundation for the design of B2C ECI. In a B2C context, a primary objective/function is to effectively present the product information to (each) individual consumers and facilitate their interaction with the product at the EC website and to support effective communication with them. This reflects an effort to reduce or compensate for the virtual nature of EC environment to the extent possible. Therefore, we define degree of SP as the extent to which the individual psychologically feels that the salesperson is present in the information-exchange context. Building upon the same theme of “presence” from SPT, the degree of PP represents the extent to which the “consumer” feels that the “product/artifact” is seemingly physically present in the information-exchange context. In this paper, we do not focus on other activities of shopping such as price negotiation, execution of sales transaction, delivery of products or services, or security of financial/payment information all of which are by no means less important; they may be probably even more important. A. Product Characterization Based on SP and PP Requirements For the purpose of this paper, the product characteristics that impact the representation of product information and providing virtual product experience through the ECI are of interest. Additionally, the product characteristics that impact the communication and consultation support required by the consumers for the product-purchase decision also need to be considered. We draw upon the product-characterization scheme proposed in [20] based on SP and PP. The required levels of these two interactions largely depend on the characteristics of the product being considered. We posit that product characteristics that lead to increased importance of visualization and experientiality, and increased product complexity would influence the levels of PP and SP requirements. In [22], visual control and functional control are used to examine the visualization and experientiality aspect of the product, which plays a key role in determining the level of PP required. Based on the notion of SP requirement and PP requirements, we identify four broad types of products—simple (for which low SP/low PP may be adequate), experiential (low SP/high PP), social (high SP/low PP), and complex (for which high SP/ high PP may be necessary). In the next section, using the concepts of SP and PP, we develop a scheme for characterizing B2C ECI.

1187

medium (LCM)—low in SP [26], [42]. The lack of SP reduces the number of cues as well as the multiple levels of meaning that can and (depending on the task context) may need to be processed. On the other hand, a real-time video-conferencing capability would be considered as a rich communication medium (RCM)—high in SP. Such a medium is capable of transmitting greater amounts of varied information/cues simultaneously. Even body language and voice can be transmitted using a real-time video-conferencing capability. The ability to convey personal feelings and emotions exists, albeit not to the same extent as in face-to-face communication. There are a number of other options that offer varied level of SP between the two extremes, e.g., electronic chat rooms, two-way audio, and one-way video conferencing. In terms of representing product information, web technology can offer a variety of options. It can vary from a simple textual description with static picture to VR technology that can be used to simulate direct product experience for the consumer [29]. These technologies allow consumers to interact with online products and show them the full details of those products [40], which can improve their experience with the products. Some exploratory research [9], [22], [29] has already suggested that VR has a potential to improve customers’ product knowledge and enhance their purchase intentions. Depending on the richness of the product information representation, consumers will perceive different levels of PP in ECI. According to MRT, a richer representation using a multiple set of symbol systems can better convey information, especially nonverbal messages, which can facilitate the understanding of the information related to equivocal, ambiguous, and unstructured tasks [30]. Thus, we expect that as representation richness (sensory breadth and depth) and consumers’ ability to control (the type and extent of interaction and information manipulated) in ECI increase, the PP experienced by the consumer would be higher. For example, textual information with a static picture that is typically observed in most websites will represent a lean product information representation (LPIR)—low PP. At the other extreme, the use of VR to simulate actual product experience will represent a rich product information representation (RPIR)—high PP. Based on the above, B2C ECI can be broadly classified as simple (LCM and LPIR), social (RCM and LPIR), experiential (LCM and RPIR), and rich (RCM and RPIR). However, the design of B2C interface should be aligned with the nature of product-purchase tasks. When the interfaces fit the requirements of the product-purchase task, they can be expected to have a positive impact on the outcomes. Drawing upon the key theories discussed earlier (TTF, SPT, and MRT), we posit that the degree of fit between SP/PP required by the productpurchase tasks and SP/PP provided through the richness of EC system interface influences the consumer outcomes.

B. B2C ECI Characterization In the context of ECI design, there are many technology options to support product-purchase tasks in terms of (product) information representation and interaction and communication between buyers and sellers. For example, e-mail or textonly messages would be considered to be lean communication

C. Research Model and Study Propositions Following the foregoing discussions and extending the theoretical bases, we propose an ECI/product fit model. In line with the media congruence hypothesis in the context of advertising versus direct experience [49], we propose that there must be


1188


Fig. 1. Research model.

congruence among the ECI and the products to generate favorable consumer outcomes. Fig. 1 shows the research model. The model indicates that the effectiveness of the ECI can be evaluated in terms of its impact on both attitudinal and behavioral dimensions. Clearly, the “bottom-line” issue of interest for providers of goods/services on the web is whether or not consumers (move from being “browsers” and) actually make the purchase. However, in the absence of “real” consumers interacting with commercial websites for “real” purchase transactions (difficult to implement in controlled laboratory experiments), the closest proxy is purchase intention. This is analogous to behavioral intention of the “Theory of Reasoned Action” [14] and “Technology Acceptance” models [10] that are popular in the IS literature. These models also argue for the importance of attitude developed toward the final action. User satisfaction (a form of attitude) is an important aspect of final action and one of the most popular dependent variable used in the IS literature especially where use is voluntary [13], [19]. The outcomes are operationalized in terms of consumers’ purchase intention [27], [33] and their satisfaction [13], [19] in Fig. 1. The basic theme in our model is that the ECI should be designed to meet the product-purchase-task requirements of consumers. More specifically, the richness of communication medium and product information representation in the ECI should be aligned with the amount of SP and PP requirements of the particular product. For example, the products that require high SP/PP create more uncertainty and ambiguity for the buyers in EC environment (due to virtual nature) and, thus, trigger a need for a richer ECI. Specifically, the consumers need richer channels of communication with sellers and interactive multisensory interaction with the product to help reduce the uncertainty and ambiguity associated with the product-purchase decision [36]. Additionally, we also know that consumers differ from one another not only on traditional demographics such as age, gender, education, and computer/information literacy but on a number of personality traits such as TA, self-efficacy, risktaking propensity, dogmatism, locus of control, degree of motivation, problem-solving/decision-making style, etc. [1], [35], [38], [46], [47, p. 61]. In light of the fact that much of the focus of MRT and its extensions has been on how best media richness can address the issue of ambiguity, equivocality, and uncertainty, we consider “tolerance for ambiguity” as a personalitytrait difference variable that could play a key role in enabling the appropriate fit between SP/PP required by the product and richness of ECIs. Empirical research in the context of information presentation and processing provides a support for this, theorizing that the decision-making outcome is improved

when problem representation and problem solving tools match the characteristics of the task and the individuals [12], [48]. For products requiring lower levels of SP and PP, consumers in EC environments are likely to feel less ambiguity and less uncertainty due to the relatively simple nature of the task. In this case, the information needed to perform the task can be easily and adequately described and conveyed by lean media (e.g., textual description possibly with static pictures for PP and e-mail communication with sellers for SP). We posit this as a “minimum or requisite level of fit.” On the other hand, (relatively) richer ECIs can also unequivocally provide all the needed information being provided by the lean interfaces. While there is inherently nothing wrong in providing richer interfaces than needed, we do not expect them to lead to any additional stronger influences on consumer attitude and behavior. Further, from an economic perspective, an ECI that enables a richer experience than necessary results in underutilization of various features/resources of EC systems and, thus, does not constitute a good strategy either designwise or economically. Therefore, we propose the following hypothesis. H1 For products with lower SP/PP requirements, there will be no difference between the leaner and richer ECIs in terms of consumer outcomes (satisfaction with EC and purchase intentions). On the other hand, for products requiring higher levels of SP/PP, consumers in EC environments are likely to feel more ambiguity/equivocality due to their inability to experience (e.g., visualize, touch, or feel) the product. They would, therefore, attempt to reduce their ambiguity by seeking out and using richer ECIs. In such situations, leaner ECIs cannot reduce their uncertainty and ambiguity related to the product-purchase tasks, thus leading to unfavorable outcomes. Hence, we propose the following hypothesis. H2 For products with higher SP/PP requirements, richer ECIs will lead to greater satisfaction and favorable purchase intentions as compared to leaner ECIs. 1) Role of Individual Differences—TA: TA is defined as the degree to which an individual feels uncomfortable with ambiguity and uncertainty [2], [17], [32]. The level of SP and PP required when making purchase decision for the same product could vary across individuals. It has been noted that individuals’ preference for SP and, thereby, required richness of communication media differ depending on their level of TA [43]. The inclination of individuals to avoid ambiguity profoundly affects the way they use media for their communication task [43]. Based on the same reasoning, it has been argued that faced with decision-making situations where users perceive higher levels of ambiguity/equivocality, they will choose an interface that can provide them with richer information [30]. This suggests that the consumers’ requirement for SP/PP for products in EC environments would vary depending on their TA. Thus, individual differences on TA are expected to be related to the effectiveness of ECI for the product-purchase tasks. Consumers with lower TA (LTA) tend to feel greater ambiguity when purchasing products with higher requirements for SP/PP and would tend to need richer ECIs. On the other hand, these richer interfaces are relatively less critical to consumers



Fig. 2.

ECI configurations.

with higher TA (HTA) in the context of the same products with higher requirements for SP/PP. Therefore, the effect of richer interfaces on consumer outcomes is more apparent for consumers with LTA than consumers with higher TA. A study [4] supports this line of discussion by noting that media-use behavior and subsequent effectiveness can be better explained when media traits (access/quality and media richness/SP) are combined with individual difference variables (in their study, “self-oriented” versus “others-oriented” disposition). Hence, we propose the following. H3 For products with higher requirements of SP/PP, the impact of richer ECIs on outcomes will be more for consumers with LTA than for those with HTA. However, for products with lower requirements for SP/PP, the impact of ECI richness will be the same regardless of consumers’ TA. III. M EASUREMENTS AND R ESEARCH D ESIGN A. Methods Given the nature of the “fit” model presented above, we employed a controlled laboratory experiment utilizing a 4 (types of ECIs) X 2 (types of products) factorial design to validate the model. ECIs were manipulated to provide simple (LCM and LPIR), social (RCM and LPIR), experiential (LCM and RPIR), and rich (RCM and RPIR) interface, as shown in Fig. 2. The experiment used a complex product that is expected to require a high SP/PP and a simple product that is expected to require a low SP/PP. Pilot tests (with a total of 19 subjects) were used to choose “appropriate products” (simple and complex) for investigation in this paper. Participants in these pilot tests made use of the criteria (importance of visualization, importance of experientiality, and product complexity) that we outlined earlier to determine the distinction between eight candidate products. Of the products considered (digital camera, personal computer, bicycle helmet, athletic shoes, battery cell, PC diskette, home security system, and car), digital camera and PC diskette were selected. The digital camera was determined to fall into the category of “high complexity” in light of the fact that there are over 25 product attributes to consider many of which are not only interdependent but also take on multiple values, while PC diskette fell in simple category. Another set of pilot tests also confirmed that digital camera would require higher levels of SP and PP compared to PC diskette. Four versions of ECI

1189

for each product category were employed. For instance, for digital camera, LPIR contained a textual description of each attribute accompanied by a static picture of the camera. RPIR consisted of interactive multimedia presentation of the product information accompanied by the ability and opportunity for the consumer to virtually try out and experience the camera’s various features such as zoom in a three-dimensional (3-D) setting (e.g., using shockwave). An ECI with LCM supported an e-mail capability to obtain advice on the product from a simulated sales representative, while RCM supported real-time one-way video and two-way audio communication with the simulated sales representative. One of the eight treatments (created by two product types and four versions of ECIs) was randomly assigned to each subject. Other individual differences such as aptitude [i.e., grade point average (GPA)], computer skills, product knowledge, etc., (captured in a separate questionnaire about ten days prior to the actual study) were statistically controlled in this experiment (found necessary as would be discussed later) to isolate the true effect of fit (between ECI and product characteristics) on the consumer outcomes. Subjects for this experiment were recruited from an undergraduate IS course in the business school of a major midwestern university in the U.S. Assuming a “medium-size effect,” it was determined that at least 45 subjects per group (or 360 subjects in all) would be required to realize an acceptable level (80%) of statistical power of the test β, as ascertained from the tables provided by Cohen [5]. In this paper, 792 subjects participated, providing sufficient statistical power. B. Measures The richness of the product information representation and communication medium of the ECI were assessed with a multiitem scale to validate the perception of participants in the experiments. The SP measure by Short et al. [41] was adapted to develop measures of communication-medium richness and product-information-representation richness of ECI. These measures were carefully developed to capture how richly product information is represented and also how richly a buyer can communicate with a seller instead of directly measuring “psychologically being close with a sales person or a product” in line with what the original measurement of SP in [41] did. Participants in the focus groups responded to these two scales after interacting with all four prototypes; only upon their confirmation that the prototypes were indeed significantly different from one another in terms of richness we proceeded with the full administration of the final experiments. Consumer outcomes were measured in terms of their satisfaction with using the ECI and purchase intention. Satisfaction with using the ECI is an extension of the previous work on user satisfaction and was measured by adapting the end-user computing satisfaction scale in [13]; despite this scale’s vintage, it still captures the key sentiments of the recent EC metrics [44], [45] that focus especially on information content, information representation, and website’s functional quality. Purchase intention was measured using the measure by Laroche et al. [28]. TA was measured with MacDonald’s scale [32] that captures subjects’ responses to each of the 20 indicator items as a


1190


binary—“yes” or “no” one week to ten days before the administration of the actual experiments as part of a prestudy questionnaire that also captured other demographic details as well as prior knowledge of eight different products including the two used in this paper. We used the author’s predefined mode of verifying the correct responses to each indicator. As recommended in [32], correct responses to ten or more of these 20 indicators would classify the respondent to have a “lower tolerance for ambiguity” (LTA) with nine or fewer correct responses qualifying the subject to have a “higher tolerance for ambiguity” (HTA). Furthermore, since we use ANOVA model with the dependent variables being metric measures and all experimental/ antecedent variables being discrete variables, (artificially) treating TA to be a metric variable would not provide us a good test of the three-way fit. All these scales and the indicators used to capture demographic variables—age, gender, intelligence (GPA), web knowledge, web usage, product knowledge, and brand knowledge— are shown in Appendix I. The propositions were tested using an analysis-of-covariance (ANCOVA) technique to detect differences among the treatments. The subjects were, as noted, randomly assigned to one of the treatment conditions. C. Experiment The first wave of experiments was administered for digital camera, and a second wave of experiments (with a time gap of about four months) was administered for PC diskette. To motivate students to participate in the study, each participant received a waiver for an assignment. Participation was voluntary and, as required by the law governing human subjects in experiments, they could withdraw at any time during the experiment. As noted, 792 students participated in these experiments. Each participant was randomly assigned to one of four prototypes in each wave. It was necessary to assess any threats to internal validity of the study in view of about a four-month gap between the two waves of experiments. We tested for differences on relevant key demographics (age, gender, aptitude/intelligence, web knowledge, and web use) and on personality-trait difference variable (TA) across these two waves of administration of the experiments. There were no statistically significant differences across the two waves except on “web use” (p ≤ 0.03) and weakly on “web knowledge” (p ≤ 0.06) in favor of the second administration (i.e., those who participated in interacting with PC-diskette product). Given the rapidity with which the awareness, acceptance, and use of the Internet and web was/is occurring in the real world, it is not too much of a surprise to find some of these differences with the passage of about four months of time in this paper. Overall, these results do not indicate any serious threats to internal validity of the study when the two sets of data are combined during hypotheses testing. As noted above, e-mail as well as real-time (one-way) video and (two-way) audio were used to enable two different levels of communication richness in the ECIs. We interviewed a number of candidates to select a person to serve as the simulated sales representative for our study. We selected one with good interpersonal-communication skills; she was provided training

TABLE I DEMOGRAPHIC CHARACTERISTICS OF SAMPLE

on various features of the products. The experiments were conducted in a computer laboratory with 48 Pentium machines running in a LAN environment to ensure a good performance especially for the multimedia interface. We acknowledge that all consumers may not have access to high-speed lines; however, with improving bandwidth capacity and line speeds, this may not be a serious roadblock in the future. The experiments in each wave were administered over a twoweek period in multiple sessions. A maximum of 40 subjects participated in each session. About a week prior to the experiment sessions, all test subjects filled out a pretest survey (see Appendix I—Part A) that elicited demographic information, computer skills, product knowledge, and the individual personality difference characteristic (TA). All participants were informed about the schedule of the experiment sessions well in advance. Pilot sessions were used before the experiments to ensure that participants did not have any difficulty in working with the system. Based on the feedback from pilot session, minor modifications such as changing character size and wording were made in the prototypes and the instructions. The procedures followed by subjects during the experiment are listed in Appendix II. D. Sample Characteristics Out of the total sample size of 792, 408 subjects participated during the first wave to interact with the system to buy a digital camera (complex product type) and 384 subjects participated during the second wave to buy a 3 1/2-inch PC disk pack (simple product type). Almost equal size of the subsample 97 to 103 in each cell across the four different prototypes and two product type groups was obtained. There were 413 male subjects and 378 female subjects (one missing value). Table I displays some of the key demographic characteristics of the subjects in this paper. E. Random-Assignment Check Although all necessary care was taken to assign subjects randomly across experimental groups, it is important to evaluate if the subjects were indeed randomly distributed across the groups



TABLE II RANDOM ASSIGNMENT OF SUBJECTS ACROSS FOUR ECIs AND T WO P RODUCT T YPE G ROUPS

in terms of demographic and other characteristics such as age, gender, GPA, and computer/Internet skill, etc. An analysis of all demographic variables was done to check for any bias in assigning the subjects. Chi-square (for categorical variables) and one-way ANOVA (for continuous variables across more than two groups)/t-test (for continuous variables across two groups) were performed. Table II shows the results. Except for a significant difference on general PC skill and weak difference on GPA, there is no bias in the assignment of the subjects to the four ECIs. However, some differences were observed across the two product types (incidentally, also the two waves of test administration) on product knowledge, brand knowledge, web knowledge, web usage, and GPA. Overall, at a significance of 0.05, the tests show that the subject assignment to the various ECI groups across these demographic variables was random and unbiased, whereas there may exist some bias across the two product groups (i.e., digital camera and PC diskette). As alluded to earlier, greater awareness, acceptance, and use of the Internet/web in the real world is reflected in the passage of about a four-month time gap between these two sets of experiments. The bias (on product and brand knowledge) could be due to obvious differences in the nature of the product (a very simple product and a relatively complex product). Therefore, in all subsequent data analyses, we controlled for these six variables (GPA, PC skill, web knowledge, web usage, product knowledge, and brand knowledge) by including them as covariates in ANCOVA tests. F. Validity and Reliability This paper examined all major types of validity. First, content validity [11] was established by the in-depth literature analysis while developing the theory, precautions taken in deriving the items, and the various stages of refinement undertaken. Both convergent and discriminant validity were established through factor analysis [24]. A factor analysis using principalcomponent factor extraction and varimax rotation technique was performed to examine the unidimensionality/convergent validity [34] of each of the two predefined multiitem construct, EC satisfaction, and purchase intention. A joint factor analy-

1191

TABLE III (a) RESULTS OF INDIVIDUAL FACTOR ANALYSIS—EC SATISFACTION AND PURCHASE INTENTION. (b) RESULTS OF THE JOINT FACTOR ANALYSIS—DISCRIMINANT VALIDITY

sis on all the indicator variables (representing both outcome variables) employing the same factor extraction and rotation approach was employed to determine discriminant validity [24], [39]. The commonly accepted decision criteria in social-science research: an eigenvalue more than one, at least 50% variance being explained, and a simplicity of factor structure were applied [16]. A single-factor solution with a high factor loading (the lowest being 0.67 for Item # 9 in EC satisfaction) emerged in individual factor analyses for each of the two outcome variables (EC satisfaction and purchase intention) satisfying the decision criteria indicated above, providing evidence for unidimensionality and convergent validity for each of these measures, as shown in Table III(a). The results of the joint factor analysis by including all the 11 indicator variables of the two measures used in this paper (i.e., nine EC satisfaction items and two purchase-intention items) to assess discriminant validity are presented in Table III(b). As predefined, a two-factor solution emerged, as shown in Table III(b). All indicator items of EC satisfaction load on a single factor (factor one) and those of the purchase intention on


1192


TABLE IV RESULTS OF TEST FOR CRITERION VALIDITY AND RELIABILITY

factor two. Overall, the two measures are distinguishable from each other, thus providing evidence for satisfactory discriminant validity. Criterion validity was established by examining the correlation between the multiitem indicators measuring a construct and one or more (different) global items measuring the same construct. As seen in Table IV, each of the two multiitem outcome measures (V1: EC satisfaction and V5: Purchase intention) was significantly correlated (p ≤ 0.001) with the overarching criterion variables (V2 to V4 representing overall EC system’s satisfaction, overall decision satisfaction, and overall process satisfaction, respectively, and V6 referring to self-reported order placement action) for each measure, thus confirming criterion validity. This paper evaluated each measurement scale with Cronbach’s alpha, which is one of the most commonly used reliability coefficients, to assess the internal consistency of multiitem scales [37]. Again, as seen from Table IV, both measures had an Alpha much larger than 0.70, which is a threshold acceptable in exploratory research attesting to satisfactory reliability [34], [50]. Finally, before testing this paper’s propositions, it is necessary to test whether the data are normally distributed. The normal QQ plot of standard residuals versus expected normal values for each of the two outcome variables was satisfactory since there was no major departure of the normal plot from the 45◦ line. Thus, the normality assumptions were supported [23]. IV. D ATA A NALYSIS AND R ESULTS To evaluate if a fit between SP/PP required by the product and those provided by ECIs leads to favorable consumer attitudes and behavioral intentions, detailed analysis including tests for three two-way interaction effects between (the four) ECIs, (the two) product types, and (the two levels of) TA, and one three-way interaction across these main factors was performed using ANCOVA with GPA, Web Knowledge, Web Usage, PC

Skill, Product Knowledge, and Brand Knowledge as the six covariates. However, before conducting these data analyses, it was necessary to verify if the participants’ actual perceptions of ECI richness matched with the designed operationalizations. The results of a one-way ANOVA analyses of the “tests of difference” of participants’ perceptions across the four ECIs confirm that the four ECIs are significantly different from one another on communication-medium richness and productinformation-representation richness, as shown in Table V. Therefore, we conclude that the four ECIs adequately pass the manipulation checks. The results of the ANCOVA tests of the study hypotheses are presented for “EC satisfaction” in Table VI(a) and (b) and for “purchase intention” in Table VII(a) and (b) and plotted in Fig. 3(a) and (b).1 Table VI(a) presents details of the mean values of EC satisfaction (on a 1–5 scale), and Table VII(a) presents the same details for purchase intention (on a 1–6 scale) by main factors, and two-way and three-way combinations of the main factors. The values of each three-way combination are also displayed in Fig. 3(a) and (b). The values within the parentheses are the sample sizes for each of the sliced subgroup and subgroup combinations. These two (sub) tables and the associated figures provide a general sense of the direction and dispersion of the two consumer outcome values under examination in this paper. A. ECIs and Product Types The results from Tables VI(a) and VII(b) indicate that for the overall sample, the interaction of ECIs with the “product 1 Given the possibility of even small effects being found to be statistically significant in view of larger than the minimum sample size (of 45 to provide adequate statistical power of 0.80) and thus create a spurious situation, a random sample of 50 subjects was selected for each of the eight cells and models rerun. Data analyses with the reduced sample sizes provide similar results, and thus, the findings of the study with the full sample can be treated with confidence, and that the effects are not likely to be spurious.



1193

TABLE V RESULTS OF MANIPULATION CHECK

TABLE VI (a) COMPARISON OF RESULTS BY STUDY’S TREATMENT VARIABLES FOR CONSUMERS’ SATISFACTION WITH ECI (EC SATISFACTION). (b) ANCOVA RESULTS FOR CONSUMERS’ SATISFACTION WITH ECI (EC SATISFACTION)

type” (which in this paper represents fit between the two) has a significant influence on both outcome variables (EC satisfaction: F = 5.325, p ≤ 0.001; purchase intention: F = 8.949, p ≤ 0.001). There are no other significant two-way interactions. The results [from Table VII(b)] show that there is also a significant three-way interaction effect (of ECIs with product type and with TA) on consumers’ purchase intentions. The EC system (representing the four different ECIs) has a significant direct effect on both consumer-outcome variables (consumer satisfaction: F = 11.217, p ≤ 0.001; purchase intention: F = 5.504, p ≤ 0.001). These results provide a (partial) support for the research model.

In Table VI(b), the effect of the product type is also seen to be significant. The results in Table VI(a) and Fig. 3(a) suggest that this occurs due to a much bigger impact of all three richer interfaces (experimental, social, and rich interfaces) in the case of complex product type rather than simple product type, whereas the difference of EC satisfaction levels of both product types in simple interface is relatively small. We also see a few covariates (Web use and product knowledge) having significant positive influences on EC satisfaction and purchase intention. This suggests that people who have more Web experiences or more product knowledge are more comfortable with electronic commerce environments (and purchase in cyber


1194


TABLE VII (a) COMPARISON OF RESULTS BY STUDY’S TREATMENT VARIABLES FOR CONSUMERS’ PURCHASE INTENTION. (b) ANCOVA RESULTS FOR CONSUMERS’ PURCHASE INTENTION

space) than people who have less Web experiences and less product knowledge. However, note that even after controlling for these background differences, the effect of ECI richness and fit with the product type prevails. As Tables VI(a) and VII(a) indicate, at an overall level, it would appear that there is a sort of hierarchy in terms of the effect of ECIs on two consumer outcomes—the rich ECI followed by social, experiential, and simple in a descending order. Additional analyses (not reported here for brevity) indicate that for a complex product type (digital camera), ECIs are found to be a significant factor influencing both EC satisfaction (F = 12.83, p ≤ 0.001) and purchase intention (F = 12.47, p ≤ 0.001). As shown in Tables VI(a) and VII(a), the rich ECI has the highest, and simple interface has the lowest effect among the four interfaces. The social and experiential interfaces also have stronger effects than the simple interface; however, among themselves, they appear to have about the same effect. This is somewhat analogous to the observations in [22] of either a visual-control or functional-control effect dominating with little difference between the two. The potential for an interaction of these intermediate levels of richness with product types and possible compensatory effects of one type of presence over the other will be discussed later with the implications. On the other hand, in the case of a simple product type (PC disk), somewhat different results are obtained. Although the rich ECI continues to generate the highest effect and simple as well as experiential ECIs generate the lowest effect on EC

satisfaction (with the social ECI obtaining in-between values), their effects on purchase intentions are less consistent. However, even in the case of the effects on satisfaction, they are not statistically significant. Overall, all four ECIs appear to be equally good for a simple product type. Therefore, from a standpoint of requisite level of fit between the ECI and product type as well as from an economic standpoint, it would appear that a simple ECI is adequate in the context of a simple product type. While higher levels of richness (as depicted by the other three ECIs in this paper) are not harmful per se, they appear to be unnecessary and, perhaps, an overkill. But, in the context of a complex product, it would be critical to move toward higher levels of richness to realize superior consumer outcomes. These results support hypotheses H1 and H2 . B. TA and the Three-Way Effects 1) Effects on Purchase Intention: Deeper examination of the purchase-intention values in Table VII(a) and Fig. 3(b) indicates that LTA consumers are much more reluctant to indicate favorable purchase intentions for a complex product type (digital camera) than for a simple product type (PC diskette) when presented with a simple ECI. Also, it can be noted [from symbols and of Fig. 3(b) or columns five and six of Table VII(a)] that on average, the product-informationrepresentation dimension (experiential) contributes relatively more than communication-medium dimension (social) of ECI richness 0.47 (5.32–4.85) versus 0.17 (5.40–5.23) for LTA



Fig. 3. (a) Consumers’ satisfaction by ECIs, product type (simple—SP; complex—CP), and TA (low—LTA; high—HTA). (b) Consumers’ purchase Intention by ECIs, product type (simple—SP; complex—CP), and TA (low— LTA; high—HTA).

consumers across different product types—complex vis-à-vis simple product type. Furthermore, it may be noted that there appears to be a significant synergistic effect of product information representation and communication medium, which is an improvement of 1.26 (5.69–4.43) on a 1–6 range when a rich ECI is presented for LTA consumers dealing with a complex rather than a simple product type. For a complex product type [as noted from a symbol of Fig. 3(b) or column six of Table VII(a)], there is a synergistic influence noted for a rich ECI, which is an improvement of 1.96 (5.69–3.73) on a 1–6 range relative to a simple ECI as compared to only richer product information representation being provided (1.59) or only richer communication medium being provided (1.67). But these LTA consumers do not exhibit a similar pattern of behavior when provided with progressively richer ECIs in dealing with a simple product type [see up and down swings in Fig. 3(b)]. In fact, they may be displaying a sense of confusion/uncertainty when provided with a very rich ECI to deal with a simple product type. In the case of HTA consumers [as noted from symbols and x of Fig. 3(b) or columns five and six of Table VII(a)], the contribution of richer product-information-representation dimension is again very substantive vis-à-vis communicationmedium dimension (1.17 versus 0.46) when examining their purchase intentions across the two product types. However, the extent of impact of each dimension of richness and the synergistic effects are lesser for HTA consumers within the complex product type. Finally, their behavior also appears to be

1195

less systematic when presented with (progressively) richer ECIs in dealing with a simple product type, but they may be more consistent than LTA consumers as noted earlier. Additional data analyses (not presented here in the interests of journal space) also suggest that in the LTA consumer group, the two-way fit between ECIs and product type is stronger (p ≤ 0.001) than for the HTA group (p ≤ 0.047). 2) Effects on EC Satisfaction: Analysis for EC satisfaction [from details provided in Table VI(a)/Fig. 3(a)] similar to that presented in the previous few paragraphs suggests that both LTA and HTA consumers express greater satisfaction when either richer communication medium or product information representation is available in the ECI and a synergistic improvement when both are present in the ECIs. The synergistic-improvement effects, however, appear to be slightly greater in the case of HTA as compared to LTA consumers. Closer examination of the EC satisfaction values in Table VI(a) and Fig. 3(a) (symbols and ) indicates that LTA consumers are extremely dissatisfied when presented with a simple ECI to interact with a complex product type (digital camera) as compared to simple product type (PC diskette). When dealing with complex product types, it can also be noted that on average, both product-information-representation and communication-medium richnesses contribute equally (0.50 increase over a base value of 3.66 on a 1–5 range obtained with a simple ECI); there is a marginal synergistic influence as well for rich ECI, which is an improvement of 0.58 as compared to an improvement of 0.50 with only richer product information representation or communication medium being provided. In the case of HTA consumers, the pattern is similar with each dimension contributing an improvement of 0.35 to 0.33, respectively, and a marginal synergistic improvement of 0.43 when both are provided. The strength of these influences is lower for HTA consumers than for LTA consumers within the complex product type. When these consumers deal with a simple product type, both LTA and HTA consumers are less consistent in their expression of satisfaction with the various EC systems they use [see Fig. 3(a)]. The richer product information representation seems to result in a drop in their level of satisfaction relative to the LPIR. They appear to feel that rich representation in the case of such a simple product as PC diskette is more an annoyance than a value-added functionality, with the HTA group expressing greater displeasure. But, both LTA and HTA consumers overcome the negative attitude if richer communication medium is provided in their ECI—either alone in social ECI or together with richer product information representation in rich ECI. Although the three-way interaction is not statistically significant in its effect on EC satisfaction outcome, additional analysis (not reported here in the interests of journal space) indicates that the two-way fit between ECIs and product type is stronger (and, in fact, significant only) for LTA consumers (F = 4.553, p ≤ 0.004) as compared to HTA consumers (F = 2.275, p ≤ 0.080). Overall, the results imply that individual difference on TA does indeed play an important role. TA has a significant threeway-fit effect in the case of purchase intentions as well as differential effects on the two-way fit between ECIs and product


1196


types. TA also produces different types of behavioral effects within each category of product type when presented with different ECIs. Although TA has no significant three-way-fit effect in the case of consumer satisfaction with their ECIs, it does generate differential effects on the two-way fit between ECIs and product types as well as differential attitudes on the part of consumers within each of the product types when ECIs with different levels of richness are provided. These findings generally support hypothesis H3 . V. D ISCUSSIONS AND C ONCLUSION A. Discussion of Research Findings As can be seen from the research findings presented above, some results (e.g., fit between ECIs and product types) emerge as posited while others (e.g., no significant difference among experiential versus social forms of ECIs and a potential compensatory effect) are somewhat unexpected. 1) Congruence Between ECI Richness and Product Types: This paper’s main proposition was that a fit between SP/PP required by the product and richness of ECIs would lead to favorable consumer outcomes. More specifically, we posited that the influence of this fit would be important in the case of consumers interacting with the website to acquire complex products, and that a notion of requisite level of richness or fit would operate in the case of consumers wanting to acquire simple products. The results support this proposition unambiguously. In the case of a complex product type (digital camera), simple ECI is the worst, and rich ECI is the best among the four interfaces in terms of the resulting attitude and purchase intention of the consumers. Although the influence of the fit (between the ECI and product types) is in the right direction, and the first two propositions of this paper are totally supported, the realization of superior attitude and behavioral intentions across all three richer categories of ECI, especially the experiential and social, is not much different from one another. Several possible interpretations of this result follow next. First, it is possible that there may exist some compensatory effects between product-information-representation and communication-medium richnesses. Participants in this study, consciously or otherwise, appear to be making tradeoffs between the use of richer communication medium (e.g., real-time communication) and richer product information representation (e.g., using interactive 3-D-multimedia technology capabilities). Either of these functionality and experience appears to help them to, first, obtain and then process the needed information and be ready to make the decision. Providing any one of these two types of richness is able to reduce the need for and possibly nullify the effect of the other, and makes it unnecessary to generate more favorable consumer outcomes. To realize favorable consumer outcomes, either richer product information representation or richer communication medium may be adequate in ECIs that deal with the complex product types investigated here. Second, although the results during the pilot-test stages of this research showed that the product selected as a complex product type in this paper (digital camera) required high SP and high PP, it may not be complex enough to require highest levels

of both SP and PP in the main study. Thus, either the social or experiential ECI might be sufficient to satisfy the level of SP/PP requirements of digital camera. Third, even if the digital camera as a product is perceived to be complex enough, the rich ECI provided in this paper might not have been rich enough in a true sense to satisfy both high-SP and high-PP requirements of digital camera. Although the rich ECI provided the richest product information representation and communication medium of the four interfaces, its interactive multimedia features were limited; for example, consumers could not backtrack to a previous stage while testing a productfeature, freely zoom in/zoom out the picture, etc. Instead of combining each technology involved in experiential ECI (3-D multimedia) and social ECI (real-time audio/video conferencing), we could have employed more sophisticated interactive multimedia (e.g., VR) capable of offering much higher richness in ECI to distinguish it from the two other rich ECIs. Such functionality, in turn, might have lead to significantly superior consumer outcomes than the experiential and social ECI used in this paper. But, for the purpose of better control in this paper and to identify the true influences of each of these variables, we deliberately designed the rich ECI in such way that it contained the additive combination of feature of experiential and feature of social ECI. Thus, while the results seem to show that combining RCM and RPIR results in a synergistic effect, the extent of synergy may not have been large enough. Overall, the first two main propositions of this paper are fairly well supported. For a simple product, simple ECI is found to be sufficient. This is what we have posited and labeled as a “minimum, requisite level of richness and fit.” For a complex product, rich ECI is needed. 2) Individual Differences on TA: Results presented earlier indicate the importance of individual difference on TA (low LTA and high HTA). Besides having a direct role expressed in a significant three-way interaction effect on purchase intentions, TA also plays an important role in shaping the relationship of the fit between the ECIs and product types with consumer outcomes. The results show the influence of congruence between ECIs and product types on consumer outcomes to be more pronounced in LTA than in HTA consumers. When interacting with a complex product type, LTA consumers are more likely than HTA consumers to avoid the ambiguity and uncertainty associated with buying a complex product. Thus, when a simple ECI is provided, LTA consumers appear to be more uncomfortable than HTA group and exhibit more unfavorable attitude/behavioral intention. On the contrary, both HTA and LTA consumers appear to almost equally appreciate richer interfaces. The differences across these two consumer groups seem to be more apparent for the complex-product category. When interacting with a complex product, LTA consumers appear to express significant differences between the simple and each of the three other richer ECIs. HTA consumers do not seem to express as strong or consistent a difference. However, as noted earlier, an evidence for compensatory effects appears to exist. These findings confirm some recent observations [3], [4] that media-use behavior and subsequent effectiveness can be better explained when media traits (e.g., media richness/SP) are combined with individual difference variables.



B. Implications for Research and Practice As noted earlier, “fit-” related theories (e.g., TTF, MRT, SPT) suggest that technologies that realize fit with tasks outperform technologies that do not. We extend this argument a step further in that along with misfit, the notion of requisite fit may need to be recognized, which these theories may have implicitly or explicitly assumed. The notion of requisite fit assumes relevance in some product interaction situations where technologies that are richer than needed (what we would label as overfit) are provided without any commensurate beneficial outcome. In other instances of interaction with complex products, technologies that are leaner than needed (what we label as underfit or misfit) lead to significantly negative consequences. Therefore, this paper suggests a new generation of fit theory that each task requires a minimum or requisite level of richness of technologies, and as long as technologies meet the minimum requirement, such technologies can generate favorable outcomes, and that additional richness while not detrimental neither leads to more beneficial consumer outcomes nor optimal utilization of resources. This research offers practitioners an opportunity to evaluate and (re)design their EC system interfaces to realize enhanced effectiveness (e.g., potentially increased sales). Our framework is easy to understand and has high practical applicability. Providers of products/services need to design their EC system interfaces to fit with the characteristics of products/services they provide and also carefully consider consumers they deal with. The study stresses the importance of consumer interaction and provides practical guidelines on how to best design the interaction in terms of product information representation and communication media with a wide variety of currently available technologies. We would speculate that without appropriate interface design, companies desiring to conduct online commerce would not be able to survive the fierce competition from the exploding number of competing commercial websites. Many merchants (may) provide both simple and complex products. In such instances, it is even more important for them to recognize that a distinction be drawn in terms of ECI richness that they should offer. Their websites should be designed to be adaptive depending on the product types chosen by consumers. In terms of interaction, this paper suggests that adequately rich interface must be provided for complex products, but for simple products, rich interface is optional. Also, results of this paper suggest existence of potential compensatory effects between richer communication media and product information representation. Consequentially, firms that sell complex products can choose to either provide highly interactive multimedia features on their commercial websites or maintain online help desk similar to toll-free call centers; they do not necessarily have to provide both. Consumers should be able to try out various features of products through interactive multimedia technologies (including VR) or communicate with real persons (e.g., seller or expert) for additional product information so that they can develop a full sense of the products they want to buy. Providing sophisticated interactive multimedia capabilities or maintaining this human pool is likely to be expensive. Since either capability can increase effectiveness of EC environments,

1197

it should be considered as a path to gaining competitive advantage. If firms want to operate their businesses dealing with complex products 24 h a day and seven days a week without any real-time human contact, interactive multimedia technologies with the requisite level of richness needed for those complex products could be adequate. This is especially true in light of the fact that the cost of making available human pool 24 × 7 for real-time interaction is enormous. For simple products, this paper shows that richer interfaces do not significantly increase consumer outcomes. Consumers may not need or use high level of SP provided by audio/video communication. In fact, they may feel overwhelmed. Thus, designing in rich-interface features could be a poor allocation of resources. However, such features could enhance satisfaction (because richer interfaces might be inherently more enjoyable than lean interfaces), and some consumers (first time buyers or buyers with very low TA) might want to have interaction with those features. One possible option of ECI design for simple products is to provide RCM for a limited time (e.g., busiest shopping date or time) or RPIR for a limited number of items (e.g., best selling items or promotional items). But, decision on optimal combinations of both of these features should be based on the economic analysis of cost and advantage of each with due consideration of the nature of products/services. Another factor in need of consideration when designing ECIs is the individual differences. Consumers differ in various aspects. The results of this paper suggest that the same interface for a given product generates different influences on consumers depending upon their TA. This also has another implication for practitioners. Consumer preferences cannot be tracked unless they are identified when accessing the website. Companies can first try to figure out the type or information usage style of their target consumers using various methods (in the form of profile registration, sample survey, etc., that many do at this time). It is not too difficult to develop a profile of the consumers when they access the website on the first occasion through the application of artificial intelligence. These profiles can be stored in the database for future reference whenever the users access the website again (similar to what Amazon.com and others do). We suggest that these firms can go to the next logical step; using the profile of target consumers can provide multiple interfaces for consumers to select from or provide the most desirable interface using intelligent techniques. For LTA consumers buying complex products, richer ECIs should be provided; for HTA consumers, richer interface may help while it is not necessary. If we do not know whether a customer is LTA or HTA, it is perhaps better to provide a richer interface. However, e-tailers need to understand that this comes at an additional cost. On the other hand, trying to assess the TA level of (potential) customers also comes at a cost. Thus, the firm needs to assess the cost and benefits of either alternative before finalizing their decision. C. Limitations and Future Research Directions We should point out that “interface design” to obtain a “fit” between the product types and EC systems is just one among the many other equally important variables that may be a


1198


key to the success of B2C EC on the Web. A few other key variables may be price negotiation and competitive pricing (for price-sensitive consumers), ease of negotiating and exchanging resources between the buyer and seller, ability to configure the product to personal preferences (e.g., building one’s own PC such as on Dell or Gateway Computer’s website), reliable and timely delivery of quality products or services, or security of financial/payment information, etc. In this paper, we did not focus on these other activities of e-shopping; these other variables may be probably even more important depending on the transaction environment. The ECI/product characterization presented in this paper is subjective in nature. Also, this paper only used two levels (low and high) of each dimension in ECI design. Since SP and PP required by products are continuums in nature, different levels of each presence can be achieved using various technologies. Instead of using only two levels, future studies should examine multiple levels (e.g., low, medium, and high) of each dimension to obtain more insight on ECI design. As intended by the original proponent of the measure, TA, which is the individual difference variable in this paper, was dichotomized to classify consumers into two categories (low and high TA groups). It is possible that such dichotomization could result in some loss of informational content/variability of this variable. Therefore, readers should be careful when interpreting the results of testing the third proposition of this paper. Furthermore, this paper focused on just one aspect of individual difference—TA in sort of isolation. This measure was originally designed to classify people in general: not specifically consumers. It is important to also recognize that many other facets of individual differences (e.g., psychological type, cognitive processing skills, life style, etc.) may be candidates for consideration. This paper used PC diskette as a simple product and digital camera as a complex product. These two products are really not comparable in terms of price. These are also not comparable in terms of risk (of financial exposure) in the event of an incorrect decision. Although we had provided other confidence-building features (e.g., no questions asked, money-back guarantee, quality assurance, etc.), consumers buying digital camera are likely to be more careful in obtaining product information and more serious in their decision making than consumers buying PC diskette. Future studies should replicate these experiments using other simple products with comparable price to a digital camera to examine whether similar results are obtained. Also, note that this paper only looked at a single product in each category of complex and simple products. No generalizations can be made based on a single study of just one product in each category. It is necessary to not only replicate such studies on other similar simple and complex products but also examine the two other categories of product belonging to high SP/low PP (e.g., retirement investment plan), and low SP/high PP (e.g., music CD, movie video tapes, DVD). Also, despite our depiction of the B2C commerce environment where the subjects were asked to assume that they were in the “market for these products and that they had enough money,” no real distinctions were made to clarify the need for these goods as those that are “necessary/critical items,” “routine purchase,” or “nice to

have,” etc. Furthermore, we cannot unequivocally claim that the products chosen in this paper (especially digital camera) are “personally relevant” to the subjects. Such distinctions are likely to mold the attitude and purchase intentions as well. These limitations must be recognized when interpreting the findings, and future research should attempt to factor these as additional variables. This paper used self-reported consumer perceptions (satisfaction, purchase intention) as measures of effectiveness of ECIs and also employed students as subjects. Given that the subjects are not “real consumers” having a “real need” to purchase the products presented in this paper, this research, as is true of most other laboratory studies, suffers from a lack of real-world realism. Future research could consider a field experiment to examine whether congruent environments enable greater effectiveness and whether such effectiveness applies equitably to both providers and consumers of the product/service. In conclusion, this paper has validated an “ECI typology” along two dimensions (product information representation and communication medium), established that “product presence” and “social presence” requirements are reasonable for product characterization in the emerging EC environments, and demonstrated that a fit/congruence among the ECI, product type and consumer characteristic, TA, leads to favorable consumer outcomes in the form of attitude and purchase intentions. A PPENDIX I Pretest measurements (indicated below in Part A) were performed about a week to ten days prior to the experiments, and posttest measurements (Part B) were obtained after the subjects’ participation. All indicator variables were suitably worded and presented in the form of a questionnaire. Pretest Measurements 1) Age: _________ 2) Gender: M F 3) G.P.A.: High School _________ Current College Cumulative _________ 4) General PC skills, web knowledge, and web usage: measures based on responses to the following statements on a 1–5 range (1 = very low and 5 = very high): a) word processing (e.g., WordPerfect, Word, MacWrite, etc.); b) spread sheet programs (e.g., Lotus 1-2-3, Excel, Quatropro, etc.); c) data base packages (e.g., Paradox, Access, dBaseIV, etc.); d) graphics packages (e.g., PowerPoint, Harvard Graphics, Corel Draw.); e) level of your knowledge about the Internet/Worldwide-web (Web); f) use of the Internet/Web to access information or surf just for fun; g) use of the Internet/Web to participate in online chat sessions; h) use of the Internet/Web to buy or sell; and



i) overall, how would you characterize your computer skills? j) overall, how would you characterize your Internet skills? 5) Product Knowledge and Brand Knowledge: measures based on responses to the following statements on a 1–5 range (1 = very low and 5 = very high): a) knowledge about digital camera and its various features; b) familiarity with various brands of digital cameras; and c) extent of preference for Kodak products (camera, film, etc.). 6) TA: measure based on “true” (T) or “false” (F) responses to each of following 20 statements. A score of ten or more correct responses (as specified in MacDonald [32]) is coded as LTA and the remaining as HTA. a) A problem has a little attraction for me if I do not think it has a solution. b) I am just a little uncomfortable with people unless I feel that I can understand their behavior. c) There is a right way and a wrong way to do almost everything. d) I would rather bet one to six on a long shot than three to one on a probable winner. e) The way to understand complex problems is to be concerned with their larger aspects instead of breaking them into smaller pieces. f) I get pretty anxious when I am in a social situation over which I have no control. g) Practically, every problem has a solution. h) It bothers me when I am unable to follow another person’s train of thought. i) I have always felt that there is a clear difference between right and wrong. j) It bothers me when I do not know how other people react to me. k) Nothing gets accomplished in this world unless you stick to some basic rules. l) If I were a doctor, I would prefer the uncertainties of a psychiatrist to the clear and definite work of someone like a surgeon or X-ray specialist. m) Vague and impressionistic pictures really have little appeal for me. n) If I were a scientist, it would bother me if my work would never be completed (because science will always make new discoveries). o) Before an examination, I feel much less anxious if I know how many questions there will be. p) The best part of working a jigsaw puzzle is putting in that last piece. q) Sometimes, I rather enjoy going against the rules and doing things I am not supposed to do. r) I do not like to work on a problem unless there is a possibility of coming out with a clear-cut and unambiguous answer. s) I like to fool around with new ideas, even if they turn out later to be a total waste of time. t) Perfect balance is the essence of all good composition.

1199

Posttest Measurements 1) SP—Richness of Communication Medium: measure based on responses to the following statements on a 1–6 scale (1 = strongly disagree; 2 = disagree; 3 = somewhat disagree; 4 = somewhat agree; 5 = agree; and 6 = strongly agree). a) I could easily communicate with the human expert when I needed to. b) I received the needed clarification from the expert without difficulty. c) I felt comfortable in evaluating the product information because I felt I could always ask the human expert linked to the application if I had any questions. d) The overall experience was as if I was purchasing in a real store where I can approach a sales/service representative whenever I have any questions. 2) PP—Richness of Product Information Representation: based on responses to the following statements on a 1–6 scale (1 = strongly disagree; 2 = disagree; 3 = somewhat disagree; 4 = somewhat agree; 5 = agree; and 6 = strongly agree). a) I was able to fairly easily visualize the product and all its features. b) All necessary features/specifications of the product were vividly represented. c) I was able to obtain/understand all necessary information about the product. d) It was easy to understand/process all product information that was presented. e) The overall experience was as if I was purchasing the product in a real store. 3) EC Satisfaction: based on the responses to the following statements on a 1–5 scale (1 = almost never, 2 = some of the time, 3 = about half of the time, 4 = most of the time, and 5 = almost always). a) Did the system provide the precise information you need? b) Did the information content meet your needs? c) Did the system provide information displays that seem to be just about exactly what you need? d) Did the system provide sufficient information? e) Do you think the output was presented in a useful format? f) Was the information clear? g) Was the system user friendly? h) Was the system easy to use? i) Did you get the information you needed in time? 4) Purchase Intent: based on intent to buy the product now given that sufficient money is available. Likely Improbable

1 2 3 4 5 6 7 1 2 3 4 5 6 7

Unlikely Probable

A PPENDIX II P ROCEDURES D URING THE E XPERIMENT S ESSIONS 1) Volunteers signed their names on the sign-up sheet for a session convenient to them. On the actual day of the


1200


experiment, they were assigned randomly to a workstation and received brief instructions from the coordinator that explained what kind of role he/she was supposed to play. 2) A case scenario described a buying situation, asked subjects to assume that they do have a need for the product, and that they have enough money if they wished to buy the product. This was done to minimize the criticism that laboratory studies may not simulate a real-world consumer purchase. However, given that the two products (digital camera and PC diskette) selected here are quite different in terms of “value” and “risk of an error/incorrect decision,” we took precautions in indicating on the web interfaces as well as in the written instructions that the quality of the product was guaranteed and that the buyers could return back the products for full refund if they were not satisfied after delivery. The same case scenario was provided to the subjects in each of the four interface categories. Once they had read the case scenario, they interacted with the prototype they had been assigned to. They looked at all the details of product information and tried out different features available in their respective prototype. Subjects interacting with “social” ECI or “rich” ECI provided put headsets. 3) Subjects belonging to these two categories were limited to 20 per session. Despite this restriction on number, it was very likely (and, in fact, expected) that (with one person serving as the seller’s representative) there could be a delay in fulfilling participants’ real-time video/audio requests for service. Subjects were made aware of the possibility of such delays (depending on queue length) in the case scenario write-up. The results, however, indicated that there were no more than four–six people waiting in queue at any one time and that the average servicing time was about 1–2 min. Such a phenomenon is quite typical of real-world scenarios of call centers. 4) Once the subjects finished interacting with prototypes, they filled out a posttest survey (see Appendix I) that was used to measure their satisfaction with the ECI and purchase intention. R EFERENCES [1] I. Benbasat and A. A. Dexter, “Individual differences in the use of decision support aids,” J. Account. Res., vol. 20, no. 1, pp. 1–11, 1982. [2] S. Budner, “Intolerance of ambiguity as a personality variable,” J. Person., vol. 30, no. 1, pp. 29–50, 1962. [3] J. R. Carlson and R. W. Zmud, “Channel expansion theory and the experiential nature of media richness perceptions,” Acad. Manage. J., vol. 42, no. 1, pp. 153–170, 1999. [4] P. J. Carlson and G. B. Davis, “An investigation of media selection among directors and managers: From “Self” to “Other” orientation,” MIS Q., vol. 22, no. 3, pp. 335–362, 1998. [5] J. Cohen, Statistical Power Analysis for the Behavioral Science. New York: Academic, 1988. [6] R. L. Daft and R. H. Lengel, “Organizational information requirements, media richness and structural design,” Manage. Sci., vol. 32, no. 5, pp. 554–571, May 1986. [7] R. L. Daft and K. Weick, “Toward a model of organizations as interpretation systems,” Acad. Manage. Rev., vol. 9, no. 2, pp. 284–295, Apr. 1984. [8] R. L. Daft, R. H. Lengel, and L. K. Trevino, “The relationship among message equivocality, media selection, and manager performance,” MIS Q., vol. 11, no. 3, pp. 355–366, 1987.

[9] T. Daugherty, H. Li, and F. Biocca, “Experiential ecommerce: A summary of research investigating the impact of virtual experience on consumer learning,” in Online Consumer Psychology: Understanding and Inﬂuencing Consumer Behavior in the Virtual World, C. Haugtvedt, K. Machleit, and R. Yalch, Eds. Mahwah, NJ: Lawrence Erlbaum, 2002. [10] F. D. Davis, “Perceived usefulness, perceived ease of use, and user acceptance of information technology,” MIS Q., vol. 13, no. 3, pp. 319–340, Sep. 1989. [11] D. Davis and R. M. Cosenza, Business Research For Decision Making. Boston, MA: PWS-Kent, 1985. [12] A. R. Dennis and T. A. Carte, “Using geographical information systems for decision making: Extending cognitive fit theory to map-based presentation,” Inf. Syst. Res., vol. 9, no. 2, pp. 194–203, 1998. [13] W. J. Doll and G. Torkzadeh, “A discrepancy model of end-user computing involvement,” Manage. Sci., vol. 35, no. 10, pp. 1151–1171, Oct. 1989. [14] M. Fishbein and I. Ajzen, Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley, 1975. [15] D. L. Goodhue and R. L. Thompson, “Task-technology fit and individual performance,” MIS Q., vol. 19, no. 2, pp. 213–236, Jun. 1995. [16] J. F. Hair, Jr., R. E. Anderson, R. L. Tatham, and W. C. Black, Multivariate Data Analysis with Readings. Englewood Cliffs, NJ: Prentice-Hall, 1995. [17] G. Hofstede, “Cultural dimensions in management and planning,” Asia Pac. J. Manage., vol. 1, no. 2, pp. 81–99, Jan. 1984. [18] “Online retailers score big gains for holidays,” Technology and Business Daily Infoworld.com, Jan. 14, 2003. [19] B. Ives, M. H. Olson, and J. J. Baroudi, “The measurement of user information satisfaction,” Commun. ACM, vol. 26, no. 10, pp. 785–793, Oct. 1983. [20] J. Jahng, H. Jain, and K. Ramamurthy, “Effective design of electronic commerce environments: A proposed theory of congruence and an illustration,” IEEE Trans. Syst., Man, Cybern. A, Syst., Humans, vol. 30, no. 4, pp. 456–471, Jul. 2000. [21] S. L. Jarvenpaa and P. A. Todd, “Consumer reactions to electronic shopping on the world wide web,” Int. J. Electron. Commer., vol. 1, no. 2, pp. 59–88, 1996/1997. [22] Z. Jiang and I. Benbasat, “Virtual product experience: Effects of visual and functional control of products on perceived diagnosticity and flow in electronic shopping,” J. Manage Inf. Syst., vol. 21, no. 3, pp. 111–147, 2005. [23] R. A. Johnson and D. W. Wichern, Applied Multivariate Statistical Analysis, 3rd ed. Englewood Cliffs, NJ: Prentice-Hall, 1992. [24] F. N. Kerlinger, Foundations of Behavioral Research. New York: McGraw-Hill, 1978. [25] L. R. Klein, “Creating virtual experiences in the new media,” Doctoral Dissertation, Harvard Univ., Cambridge, MA, 1999. [26] N. Kumar and I. Benbasat, “Para-social presence: A re-conceptualization of ‘social presence’ to capture the relationship between and web site and her visitors,” in Proc. 35th HICSS, Jan. 2002, pp. 106–112. [27] P. A. LaBarbera, P. Weingard, and E. A. Yorkston, “Matching the message to the mind: Advertising imagery and consumer processing style,” J. Advert. Res., vol. 38, no. 5, pp. 29–43, 1998. [28] M. Laroche, C. Kim, and L. Zhou, “Brand familiarity and confidence as determinants of purchase intention: An empirical test in a multiple brand context,” J. Bus. Res., vol. 37, no. 2, pp. 115–120, Oct. 1996. [29] H. Li, T. Daugherty, and T. Biocca, “The role of virtual experience in consumer learning,” J. Consum. Psychol., vol. 13, no. 4, pp. 395–407, 2003. [30] K. H. Lim and I. Benbasat, “The effect of multimedia on perceived equivocality and perceived usefulness of information systems,” MIS Q., vol. 24, no. 3, pp. 449–471, Sep. 2000. [31] C. Liu, “Critical success factors for designing an electronic market on the World Wide Web: An exploratory study,” Ph.D. dissertation, Mississippi State Univ., Starkville, 1997. [32] A. P. MacDonald, Jr., “Revised scale for ambiguity tolerance: Reliability and validity,” Psychol. Rep., vol. 26, no. 3, pp. 791–798, 1970. [33] A. Mehta, “How advertising response modeling (ARM) can increase ad effectiveness,” J. Advert. Res., vol. 34, no. 3, pp. 62–74, 1994. [34] J. C. Nunnally, Psychometric Theory. New York: McGraw-Hill, 1978. [35] P. C. Nutt, “Evaluating MIS design principles,” MIS Q., vol. 10, no. 1, pp. 139–156, 1986. [36] J. W. Palmer and D. A. Griffith, “An emerging model of web site design for marketing,” Commun. ACM, vol. 41, no. 3, pp. 44–51, Mar. 1998. [37] R. A. Peterson, “A meta-analysis of Cronbach’s coefficient alpha,” J. Consum. Res., vol. 21, no. 2, pp. 381–391, 1994.



[38] E. J. Phares, Locus of Control in Personality. Morristown, NJ: General Learning, 1976. [39] J. L. Price and C. W. Mueller, Handbook of Organizational Measurement. Marshfield, MA: Pitman, 1986. [40] C. Ryan, “Virtual reality in marketing,” Direct Market, vol. 63, no. 12, pp. 57–62, 2001. [41] J. Short, E. Williams, and B. Christie, The Social Psychology of Telecommunications. London, U.K.: Wiley, 1976. [42] C. Stohl and C. W. Redding, “Messages and message exchange processes,” in Handbook of Organizational Communication—An Interdisciplinary Perspective, J. Frederick, L. L. Putnam, K. H. Roberts, and L. W. Porter, Eds. Beverly Hills, CA: SAGE Publications, 1987. [43] D. W. Straub, “The effect of culture on IT diffusion: E-Mail and FAX in Japan and the U.S.,” Inf. Syst. Res., vol. 5, no. 1, pp. 23–47, 1994. [44] D. W. Straub, D. L. Hoffman, B. W. Weber, and C. Steinfeld, “Toward new metrics for net-enabled organizations,” Inf. Syst. Res., vol. 13, no. 3, pp. 227–238, 2002. [45] ——, “Measuring e-commerce in net-enabled organizations: An introduction to the special issue,” Inf. Syst. Res., vol. 13, no. 2, pp. 115–124, 2002. [46] R. N. Taylor and M. D. Dunnette, “Relative contribution of decision maker attributes to decision processes,” Organ. Behav. Hum. Perf., vol. 12, no. 2, pp. 286–298, 1974. [47] E. M. Trauth and E. Cole, “The organizational interface: A method for supporting end users of packaged software,” MIS Q., vol. 16, no. 1, pp. 35–53, 1992. [48] I. Vessey and D. Galletta, “Cognitive fit: An empirical study of information acquisition,” Inf. Syst. Res., vol. 2, no. 1, pp. 63–84, 1991. [49] A. A. Wright and J. G. Lynch, Jr., “Communication effects of advertising versus direct experience when search and experience attributes are present,” J. Consum. Res., vol. 21, no. 1, pp. 708–718, 1995. [50] Y. Yoon, T. Guimaraes, and Q. O’Neal, “Exploring the factor associated with expert systems success,” MIS Q., vol. 19, no. 1, pp. 83–106, 1995. [51] I. Zigurs and B. K. Buckland, “A theory of task/technology fit and group support systems effectiveness,” MIS Q., vol. 22, no. 3, pp. 313–334, 1998.

Jungjoo Jahng received the Ph.D. degree in management information systems from the University of Wisconsin–Milwaukee, in 2000. He is an Associate Professor in management information systems in the College of Business Administration, Seoul National University, Seoul, Korea. His research interests are in the domains of electronic commerce, IS strategy, and strategic IS planning. His research has appeared in a number of journals such as IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS, European Journal of Information Systems, E-Services Journal, and in refereed conference proceedings such as the Association for Information Systems Conference (AMCIS). He had the privilege of being one of the few accepted nationally as a doctoral consortium candidate in both AIS 1999 Conference as well as International Conference on Information Systems (ICIS-99), which is the premier conference for IS researchers and professionals. He also has a number of years of real-world systems experience in South Korea. Dr. Jahng won the first Stafford Beer Award in 2004 for his 2002 paper that appeared in the European Journal of Information Systems. One of his research papers was also nominated as one of the best papers in the AMCIS1999 Conference.

1201

Hemant K. Jain (M’06) received the B.S. degree in mechanical engineering from the University of Indore, Indore, India, the M.Tech. degree in industrial engineering from the Indian Institute of Technology, Kharagpur, India, and the Ph.D. degree in information systems from Lehigh University, Bethlehem, PA, in 1981. He is a Wisconsin Distinguished and Tata Consultancy Services Professor of management information system in the School of Business Administration, University of Wisconsin–Milwaukee. His research interests are in the area of electronic commerce, systems development using reusable components and web services, distributed and cooperative computing systems, architecture design, database management, and data warehousing. He serves as a Consultant for a number of Fortune 500 companies. He has published over 50 articles in leading journals like Information Systems Research, MIS Quarterly, IEEE TRANSACTIONS ON SOFTWARE ENGINEERING, Journal of MIS, IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS, Naval Research Quarterly, Decision Sciences, Decision Support Systems, Communications of the ACM, and Information and Management. Additionally, he has published over 40 papers in refereed conference proceedings. He is an Associate Editor of Information Systems Research, which is a flagship journal of INFORMS. He also serves on the editorial board of Information Technology and Management, International Journal of Web Services Research, Information Management, and the International Journal of Information Technology and Decision Making. Dr. Jain is on the board and a member of the Steering Committee of the IEEE Technical Community for Services Computing and is a member of the Service, Systems, and Organizations Technical Committee of the IEEE SMC Society. He was the program committee Cochair of the 2004 IEEE conference on Web services. He is a member of the IEEE Computer Society.

K. Ramamurthy received the B.S. degree in mechanical engineering from the University of Madras, India, graduate diploma in statistical quality control and operations research from Indian Statistical Institute, India, Master of Business Administration (MBA) from Concordia University, Montreal, Canada, and the Ph.D. degree in management information systems from the University of Pittsburgh, Pittsburgh, PA. He is a Professor of management information systems in the School of Business Administration, University of Wisconsin–Milwaukee. He has nearly 20 years of industry experience and has held several senior technical and executive positions. His current research interests include electronic commerce including interorganizational systems/electronic data interchange (EDI) and the Internet; adoption, implementation, and diffusion of modern information technologies; business value of IT; strategic IS planning; data warehousing and data resource management; decision and knowledge systems for individual and group support; total quality management (TQM) including software quality; and computer integrated manufacturing technologies. He has published over 35 articles in major journals including MIS Quarterly; Journal of MIS; Decision Sciences; IEEE TRANSACTIONS ON SOFTWARE ENGINEERING; IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS; Decision Support Systems; International Journal of Electronic Commerce; IEEE TRANSACTIONS ON ENGINEERING MANAGEMENT; Journal of Organizational Computing and Electronic Commerce; International Journal of HumanComputer Studies; European Journal of Information Systems; Journal of International Marketing; International Journal of Production Research; International Journal of Man-Machine Studies; OMEGA; Transportation Journal; INFOR; and a number of refereed conference proceedings. He is on the editorial board of MIS Quarterly. Dr. Ramamurthy is Roger L. Fitzsimonds Scholar. He is a charter member of Association for Information Systems (AIS), and elected to Beta Gamma and Sigma honor society.
