JCMC 9 (2) January 2004
Collab-U  CMC Play  E-Commerce Symposium Net Law  InfoSpaces Usenet
  NetStudy VEs VOs O-Journ HigherEd Conversation Cyberspace Web Commerce
Vol. 6 No. 1  Vol. 6 No. 2  Vol. 6 No. 3  Vol. 6 No. 4  Vol. 7 No. 1  Vol. 7 No. 2  Vol. 7 No. 3
 Vol. 7 No. 4  Vol. 8 No. 1  Vol. 8 No. 2   Vol. 8 No. 3  Vol. 8 No. 4   Vol. 9 No. 1

How Attitude Toward the Web Site Influences Consumer Brand Choice and Confidence While Shopping Online

Byung-Kwan Lee, Ji-Young Hong, and Wei-Na Lee
University of Texas at Austin

Abstract

As more and more consumers spend more money on the Internet, traditional retailers and manufacturers as well as entrepreneurial dot-coms are jousting to explore and shape this new business opportunity. Their long-term survival and profitability may be determined by how well the Web site helps form and sustain positive attitudes toward the site and, eventually, toward the product or the company. The purpose of this study is two-fold: (1) to examine if and how attitude toward the Web site (Ast) affects consumer brand choice; and (2) to examine the association between Ast and consumers' confidence in choice, and the moderating effect of consumer product knowledge in its relationship. The study asked participants to choose a laptop brand after visiting three laptop manufacturer Web sites for a total of 30 minutes. Their product knowledge and attitude toward the three Web sites were also measured. The study found that attitude toward the Web site is a good predictor of consumer brand choice. In addition, confidence in choice seemed to be affected by Ast, depending on product knowledge. For a group with low product knowledge (novices), Ast was likely to influence confidence in choice. For a group with high product knowledge (experts), however, Ast did not seem to affect confidence in choice.

Introduction

Despite the recent economic downturn and slumping PC sales, adoption of the Internet continues to grow and there seems to be little doubt that E-commerce will become mainstream in the future, not just a fad. According to comScore Media Matrix, as of September 2003, the total number of U.S. Internet users reached 150 million for the first time ever (Center for media research, 2003). More users means more opportunities for E-commerce. According to a survey by Goldman Sachs, Harris Interactive and Nielsen/NetRatings, consumers spent $7.4 billion in November, 2003 (Nielsen/NetRatings 2003). This represents an increase of 34 percent over the same month last year.

Internet use in the coming years is projected to be widespread geographically. In 2000, the U.S. accounted for 34 percent of worldwide Internet users. However, by 2005, the balance of Internet users is predicted to be completely shifted and the U.S. is likely to slip to third place following Europe and Asia. Similar shifts are predicted to occur for E-commerce. Based on a report from International Data Corp., Pastore (2001) expects that nearly 1 billion people, about 15 percent of the worldwide population will be using the Internet by 2005, and their use will result in more than $5 trillion in E-commerce.

As more and more consumers become part of the net population and spend more money on the Internet, traditional retailers and manufacturers as well as entrepreneurial dot-coms are competing to explore and shape this business opportunity. The question is not whether to have a Web site but how to be profitable in this intense competition. Even though the immediate tangible measure of success of a firm's Web site may be short-term revenues, its long-term survival and profitability may be determined by how well the Web site helps form and sustain positive attitude toward it and, eventually, toward the product or the company and the purchase intent (Poh & Adam, 2002). The proposition that, if a Web site is well liked, visitors to the Web site may be more receptive to the Web site's content has been supported by studies on attitude toward the Web site (Ast) (Bruner II & Kumar 2000; Poh & Adam 2002; Stevenson & Bruner II 2000).

The central goal of the study reported here was to examine empirically whether consumer attitude toward a manufacturer's Web site (Ast) is a good indicator of its success. Specifically, this study examines if and how Ast influences brand choice. To date, few research studies have dealt with Ast as an indicator of Web site effectiveness. Among the few studies, the focus has been on examining the Web site as an advertising medium and its impact on consumer judgment such as attitudes and purchase intent. However, with the increasing importance of Web sites as a business transaction tool and differences in consumer information processing between making a judgment and a choice, it would be worthwhile to investigate specifically the role of Ast on consumer choice.

The secondary goal of this study is to understand the impact of Ast on consumer confidence in choice. As a way of measuring subjective decision quality (Haubl & Trifts, 1999), consumer confidence in choice may play an influential role in facilitating their current and subsequent purchase activity, especially in the online shopping environment where products cannot be directly examined.

Furthermore, we suggest that consumer product knowledge plays a moderating role in the relationship between Ast and confidence in choice. Interactive information systems on the Internet allow consumers to be appropriately selective in their information search. In such an environment, consumers have to simultaneously manage the information flow as well as understand the information itself. With a finite amount of resources, this dual task may force consumers in a highly interactive environment such as the Web to have a lesser potential for successful information processing (Ariely, 2000). Consequently, resources such as product knowledge may have a unique and influential role in online information search and processing, subsequently affecting the way consumers make a choice and develop confidence in that choice. Therefore, this study also aims to examine how product knowledge moderates the effect of Ast on consumer confidence in choice.

Literature Review

Attitude toward the Web Site and Brand Choice

Analogous to the well-accepted definition of attitude toward the Ad (Aad), a "predisposition to respond in a favorable or unfavorable manner to a particular advertising stimulus during a particular exposure situation" (MacKenzie, Lutz, & Belch, 1986, p. 130), Chen and Wells (1999, p. 28) defined attitude toward the Web site (Ast) as a "predisposition to respond favorably or unfavorably to Web content in natural exposure situations." Their examination of differences among Web sites that scored high, medium and low on the measurement of Ast found that a high Ast score could be from many different underlying perceptual dimensions such as entertainment, informativeness, and organization.

A review of the literature reveals that research studies examining this relatively new construct, Ast, have focused on the role of Web sites as an advertising medium (Bruner & Kumar, 2000; Poh & Adam, 2002; Stevenson, Bruner, & Kumar, 2000). In an effort to test whether the relationship between Aad, brand attitude and purchase intent holds even in an online context, these studies examined the association of Ast with Aad from the site, attention to the ad, brand attitude and purchase intent. Although more rigorous testing is needed to understand the relationships among these constructs on the Web, past research suggests that, if a Web site is well-liked, some visitors to the Web site may be more receptive to the Web site's contents, including its advertisements (Poh & Adam, 2002).

The Internet has become important for marketing communication not only because it could be an effective communication medium but also because it is a business transaction tool with great potential. With the possibility of the Web site's context effects on consumer reactions to the Web content (Bruner & Kumar, 2000), Ast should be of vital importance to Web direct marketers who aim to obtain direct sales or lead as well as to the advertisers with building brand equity as their primary business objective.

The primary research issue in this study centers on how attitude toward the Web site influences brand choice. Even though judgment (constructing an overall evaluation of an alternative) and choice (choosing one alternative from a set) have not always been clearly distinguished, there are sufficient processing differences between judgment and choice that call for caution in accepting Ast - judgment (e.g., attitudes and intentions) results as indicative of the Ast - choice.

While judgment requires an explicit evaluation of each alternative, typically using a continuous or multilevel scale, choice, in contrast, requires only that one alternative be selected and the rest rejected. Therefore, when compared to a choice task, a judgment task was found to lead to more information search, a less variable pattern of search, and a greater amount of inter-dimensional search (Billings & Scherer, 1988). In terms of decision-making strategies, response mode can be predicted to have effects on the use of compensatory versus non-compensatory decision processes. Choosing one alternative and eliminating the rest encourages the use of simplified, non-compensatory strategies such as elimination by aspects (EBA) or conjunctive strategy. That is, under choice situation, alternatives can be eliminated based on only one or two pieces of information (Billings & Scherer 1988; Biehal, Stephens & Curlo 1992).

With these possible processing differences between judgment and choice, it is likely that consumers could choose without completely processing all brand information or forming overall evaluations (Biehal et al. 1992). That is, they may make choices without differentiating between brands on the basis of brand attitude or even without ever forming an overall brand attitude.

These possible differences between judgment and choice may have important implications for understanding the role of Ast in brand choice. That is, used as a cue in differentiating between brands, Ast may have a direct (separate from the effect of brand attitude on brand choice) effect on brand choice. Moreover, this direct, separate, effect of Ast on brand choice may be expected to increase in the online shopping environment. Shoppers appear to be attracted to the Web because of the ease with which they can find products with detailed information and the variety of choices offered (Ward & Lee, 1999). However, in light of the decision behavior literature, these advantages of online shopping may make choices on the Web more complex than ever before, and the resulting high level of task complexity may induce a distinct, simpler processing strategy (e.g., Swait 2001; Swait & Adamowicz 2001). In other words, consumers may use Ast to bypass the difficult choice task created by the decision-making environment on the Web.

While very few research studies have examined the effects of response mode in the context of marketing research, existing research regarding Aad may provide some indication on what to expect. In general, most research focusing on the role of Aad in determining ad outcomes has dealt with how Aad affects brand cognition, brand attitude, and purchase intention (e.g., Burke & Edell, 1989; MacKenzie, Lutz, & Belch, 1986; Miniard, Bhatla, & Rose, 1990). That is, most of the research interest in Aad as an indicator of advertising effectiveness has focused on how Aad affects consumer judgment, rather than consumer choice.

However, Biehal et al. (1992) pointed out that, even though research supports the mediated effects of Aad on intentions (i.e., Aad has an impact on brand attitude through affect transfer, and brand attitude affects intentions), Aad and brand attitude may have direct, separate, influence on choice. As a consequence of manipulating ad pictures to create differences in Aad, they found a direct, positive effect of Aad on focal brand choice over and above that of brand attitude. They suggested that the direct effect of Aad on brand choice might constitute another example of consumers' using other cues or processing heuristics to help them resolve difficult choices. In cases where consumers have isolated two or more brand possibilities that are perceived to be similar overall, they may use Aad to discriminate between them and make a choice.

Based on existing research on the effects of response mode on the ways in which information is searched and combined, and the characteristics of the E-commerce environment, we suggest that Ast may be an indicator of a Web site's success, through its use as a cue in differentiating brands in a complex and ambiguous online shopping environment. This will be examined by measuring the association of Ast and consumer brand choice.

H1: The likelihood of choosing a brand is positively related to consumers' attitude toward the Web site of the brand.

Confidence in Choice and Product Knowledge

In a brick and mortar environment, consumers can touch and feel the products and freely communicate with sales representatives about the products of interest. On the other hand, consumers in an electronic commerce environment may find it difficult to deal with the inherent nature of virtuality in their interaction (Jahng, Jain, & Ramamurthy, 2000). Hence, we propose that confidence in choice may be another important function of a Web site, and we seek to understand the impact of Ast on confidence in choice. A Web site's ability to reduce discomfort with the uncertainty and ambiguity caused by the lack of direct interaction with products and sellers may help enhance confidence in choice.

However, a consumer's knowledge of the product category such as an understanding of the nature of the attributes, possible tradeoffs, and appropriate criteria for choice, may be an important construct in understanding the effect of Ast on confidence in choice. Without a certain level of product knowledge, consumers may have difficulty in discriminating among alternatives and processing product information. Consequently, they are more likely to rely on extrinsic cues such as Web sites in making a choice and forming confidence in choice. That is, the type of information that consumers use as the basis for choice decisions may be different according to the extent of their product knowledge. The relationship between Ast and confidence in choice therefore varies dependent upon consumer product knowledge.

Considerable research in consumer behavior has examined the relationship between consumer product knowledge and information search/processing. According to Johnson and Russo (1984), experienced consumers have several advantages over consumers new to a product class. The first and most obvious one is superior knowledge of existing alternatives. Research suggests that the more familiar the decision maker is with the problem domain, the more information for solving the task can be drawn from prior knowledge (Coupey, Irwin, & Payne, 1998). Therefore, consumers with high product knowledge will exert less search effort than those who have low product knowledge (Betty & Smith, 1987; Brucks, 1985). This phenomenon may be more evident in a choice task, where the information selection skills of experienced consumers come into play, than in a judgment task (Johnson & Russo, 1984). On the other hand, consumers who are less familiar with a product category must, to some extent, construct preferences on the spot. Thus, the more ambiguous the experience with a product, the more likely consumer evaluations are susceptible to how the product is described (Levin & Gaeth, 1988). Consequently, the influence of task and context factors on choice evaluations and decisions could be greater for problem domains in which the consumer has less prior knowledge and experience (Coupey, Irwin, & Payne, 1998).

Existing evidence on cue utilization also suggests that product knowledge may affect which information will be used as the basis for choice evaluation and decisions. That is, consumers with low prior knowledge are relatively less able to use and interpret intrinsic information and thus rely on price and other extrinsic information (e.g., Park & Lessig, 1981; Rao & Sieben, 1992). Consumers with a high degree of prior knowledge may also resort to using extrinsic cues such as price and brand name to assess quality, if the diagnostic value of the extrinsic information is justified (Rao & Monroe, 1988). However, in cases where such a heuristic is not justified, the use of extrinsic cues in product quality assessments tends to decrease with familiarity (Rao & Monroe, 1998; Rao & Sieben, 1992).

Based on a review of the literature on product knowledge, we suggest that consumer product knowledge will influence the relationship between Ast and confidence in brand choice. This is because the type of information that consumers use as the basis for choice decisions may be different according to their product knowledge. For consumers with low product knowledge, Ast may be related to confidence in choice because it may function as another cue to differentiate among alternatives and as a mechanism to reduce consumer discomfort in an online environment. However, for consumers with high product knowledge, who are well equipped with a framework for analyzing alternatives, it may be difficult to expect Ast to be used in differentiating alternatives and reducing uncertainty inherent in the virtual environment.

H2: The relationship between attitude toward the Web site and confidence in brand choice is different depending on the level of product knowledge.

H2a: Confidence in choice is related to attitude toward the Web site for consumers with low product knowledge.

H2b: Confidence in choice is not related to attitude toward the Web site for consumers with high product knowledge.


Methodology

Sample

A convenience sample of 39 undergraduate students from a southwestern state university was recruited for this study. The students were offered extra credit points as incentives for their participation in the study.

Students took part in the study by completing an online shopping task in small group sessions consisted of 3 to 9 participants. The sessions were held in a laboratory. The laboratory room was equipped with laptops. All laptops had Ethernet connections. Each session was run by an administrator assisted by a lab technician.

Data Collection Procedure

At the beginning of each session, participants were told that this was a study to collect information that would help the Dean of their college make a decision on which laptop to require students to purchase. The purpose of the cover story was to engage participants with the task at hand. Instructions for the study were given on the first page of the online survey. In addition, study administrator explained the procedures to make sure that all participants followed the instructions properly.

First, participants were asked to fill out Part I of an online survey questionnaire that asked about their online experience. These questionnaire items specifically measured their product knowledge.

After completing Part I of the survey, participants were told that they had 30 minutes to go online and shop for a full-sized laptop computer within the range of $1,400 to $2,400 in order to make a recommendation to their Dean. During this phase of the study, participants were given three manufacturer sites (www.apple.com, www.compaq.com, and www.dell.com). They were told to visit only these sites to make a decision, since these were the three manufacturers that their Dean was considering. In order to record their online activities truthfully, two software programs (a PC screen recording application and a software program that monitors computer activities and records the data in Excel-like file) were turned on for each participant's PC.

Upon completion of their shopping task, participants were asked to fill out Part II of the online survey which asked for brand choice as a recommendation to the Dean, confidence in choice, and evaluation of each site. They were also asked to provide their demographic information.

Measurement

Three major theoretical constructs were examined. They were: attitude toward the Web site, product knowledge, and confidence in choice. Product knowledge was operationalized and measured by a multi-dimensional scale of self-perceived consumer expertise developed by Kleiser and Mantel (1994) on the basis of Alba and Hutchinson's (1987) dimensions of expertise (see Appendix A). The scale consisted of four qualitatively distinct dimensions of expertise including cognitive automaticity, analysis, elaboration, and memory. These fifteen items were measured on a 7-point scale ranging from 1 ("Strongly disagree") to 7 ("Strongly agree"). Cronbach's alpha for this measure was .86. Therefore, scores for the 15 items were aggregated for subsequent analysis.

Attitude toward the Web site was measured on eight 5-point scales ranging from 1 ("Definitely disagree") to 5 ("Definitely agree"). It was adapted from Chen and Wells' (1999) attitude toward the site scale. The internal reliability of this measure was proven to be acceptable (Cronbach's alpha = .80) and individual item scores were aggregated for later analysis.

Confidence in choice was assessed with an item asking "How confident do you feel about this recommendation?" on a 5-point scale anchored by "Completely uncertain" (1) and "Completely certain" (5).

Results

Preliminary Analysis

As a first step, descriptive analysis of the data was conducted. As shown in Table 1, the mean score for attitude toward the site was 2.9 (Apple), 3.0 (Compaq), and 3.3 (Dell) respectively. Paired samples t-test results revealed that there were no significant differences among the three scores (t (Apple and Compaq) = .2, t(Apple and Dell) = 1.7, t (Compaq and Dell) = 1.6, n.s.).

Sample Size
Mean*
Standard Deviation
Apple
31
2.9
1.0
Compaq
31
3.0
1.0
Dell
33
3.3
0.9

Table 1. Attitude toward the three Web sites.
* The mean value is the average of eight 5-point scale scores.

Among the 33 participants who made a choice, four participants chose Apple (12%) while 13 participants chose Compaq (39%). Dell was chosen by 16 participants in this study (49%). Table 2 shows the breakdown. Average confidence in choice for each group was 3.7 for Apple, 3.5 for Compaq, and 3.6 for Dell. Independent samples t-tests indicate that there were no statistical differences in confidence in choice scores among the three groups.

As in prior studies, a median split was used to divide participants into high and low product knowledge groups based on the aggregated product knowledge scores. As a result, 16 participants were classified into the low knowledge group and 17 in the high knowledge group. The resulting mean composite product knowledge scores were significantly different between the high and low product knowledge groups (M (high product knowledge) = 3.3, M (low product knowledge) = 1.9, p ≤ .001).



 
Frequency
Share(%)
Choice Confidence
Apple
4
12.1
3.75
Compaq
13
39.4
3.46
Dell
16
48.5
3.56
Total
33
100
 
 
Mean Difference
df
t
Apple & Compaq
0.3
15
0.3
Compaq & Dell
0.1
27
0.2
Dell & Apple
0.2
18
0.3

Table 2. Brand choice and confidence.
* p > .05

Next, an independent samples t-test between high and low product knowledge groups was conducted to examine how they differ in attitude toward the Web site and confidence in choice. Table 3 shows the outcome of the test. Consistent with previous studies, the high product knowledge group was significantly more confident in brand choice than the low product knowledge group (M (high product knowledge) = 3.9, M (low product knowledge) = 3.1, t = 2.35, p ≤ .05). However, it was found that attitude toward the three Web sites and attitude toward the Web site of the chosen brand were not significantly different between high and low product knowledge groups.



 
Knowledge Level
Mean (SD)
Sample Size
t
Confidence
Low
3.1 (1.1)
16
2.35*
 
High
3.9 (0.8)
17
Attitude toward the site
Low
2.6 (1.1)
16
0.78
(Apple)
High
2.9 (1.0)
17
Attitude toward the site
Low
2.9 (0.7)
16
0.75
(Compaq)
High
2.7 (1.1)
17
Attitude toward the site
Low
3.2 (0.8)
16
0.50
(Dell)
High
3.4 (0.5)
17
Attitude toward the site
Low
3.7 (0.7)
16
0.29
(Chosen brand)
High
3.7 (0.5)
17

Table 3. Mean and standard deviation for low and high knowledge groups
* p ≤ .05, df = 31


Hypothesis Testing

Table 4 shows frequency of brand choice and rank order of attitude toward each site. As can be seen, participants were highly likely to choose a brand of the Web site they evaluated most favorably. For those who chose Apple (4), all of them evaluated the Apple site most favorably. For those who chose Compaq (13), 77% (10) of them regarded the Compaq Web site most favorably. For those 16 participants who selected Dell, 81% (13) of them ranked the Dell Web site at the top.

Choice
Attitude Ranked First
Brand
# of Choice
Apple
Compaq
Dell
 
Apple
4
4
100%
0
0%
0
0%
Compaq
13
1
8%
10
77%
2
15%
Dell
16
2
13%
1
6%
13
81%
Total
33

Table 4. Frequency of brand choice and attitude ranked first.

In order to test Hypothesis 1 regarding whether attitude toward the site influences brand choice, we regressed brand choice (dependent variable) on attitude toward the three sites (independent variable) with conditional logistic regression (Table 5). As predicted, the Beta coefficient of attitude was statistically significant (t = 3.53, p ≤ .001). It can be concluded that the probability of choosing brand A is significantly higher than choosing other brands if a participant evaluates the site of brand A more favorably than any other brand's site. The independent variable (attitude) of this model explains 43% of the total variation in the dependent variable (brand choice).

Independent Variable
Beta
Coefficient
Standard
Error
t
Pseudo
R2
Attitude
1.98
.5
3.53*
.43

Table 5. Conditional logistic regression results for the effect of attitude on choice.
*p ≤ .001
Dependent Variable: choice

Since the sample size was relatively small, we conducted power analysis in order to determine whether the significant result is statistically meaningful. In our analysis, the effect size is indicated by f2 = R2 / (1 - R2). Given R2 = .43, f2 = .7543, N = 33, a = 0.05, the power is 0.9979 which is deemed acceptable. R2 of .43 is equivalent to Cohen's d of 1.7 which is large for effect size.

Tests of Hypotheses 2a and 2b were conducted to examine how the relationships between attitude toward the site and confidence in choice differ in high and low product knowledge groups. Hypothesis 2a postulates that in low product knowledge group confidence in choice will be influenced by attitude toward the site. Hypothesis 2b predicts that in the high product knowledge group attitude toward the site will not affect confidence in choice because their confidence is expected to be dependent on product knowledge rather than attitude toward the site. Simple regressions were conducted for low and high product knowledge groups. Tables 6 and 7 provide regression results for low and high product knowledge levels respectively.

For the low product knowledge group, attitude toward the site was found to influence confidence significantly (B = .95, t = 2.79, p ≤ .05). This result indicates that with low product knowledge, the more favorable attitude a consumer has, the higher her/his confidence level is. The correlation coefficient (R) of this model is .59 and the model proved to be statistically significant (F = 7.8, df = (1, 14), p ≤ .05). According to this model, attitude toward the site explains 35% of total variation in confidence (R2 = .35). The statistical power of this analysis is 0.7791 which is acceptable. (R2 of .35 is equivalent to Cohen's d of 1.5 which is large for effect size.)

Independent variables
Unstandardized
Coefficients (B)
Standardized
Coefficients (Beta)
t
p
R2
F
(Constant)
-.38
.30
.76
Attitude
.95
.59
2.79
.01
.35
7.8*

Table 6. Regression result for the low product knowledge group.
*p ≤ .05
Dependent Variable: confidence

However, for the high product knowledge group, the coefficient of attitude toward the site did not reach the statistically significant level (B = .21, t = .50, n.s.). It seems that attitude toward the site does not influence confidence in choice in the high product knowledge group.

Independent variables
Unstandardized
Coefficients (B)
Standardized
Coefficients (Beta)
t
p
R2
F
(Constant)
3.14
1.99
.06
Attitude
.21
.13
.50
.62
.01
.2

Table 7. Regression result for the high product knowledge group.
Dependent Variable: confidence

In order to take into account that type I error is likely to be increased by separate simple regressions and provide a more rigorous test of the moderating effect of product knowledge on the relationship between attitude toward the site and confidence in choice, multiple regression analysis was performed. Dependent variable was confidence in choice and independent variables were as follows: (1) attitude toward the site of the chosen brand (i.e., if a participant chose brand A, his/her attitude score of the brand A was entered), (2) a dummy variable of product knowledge group, which had a value of 1 if product knowledge was low and 0 if product knowledge was high, and (3) two-way interaction between attitude and product knowledge group.

Overall, the regression model was statistically significant (F = 5.3, df = (3, 29), p ≤ .01), indicating that confidence in choice is influenced by attitude toward the site, knowledge, and their interaction (Table 8). The statistical power of this analysis is 0.9259 which is acceptable. (R2 of .35 is equivalent to Cohen's d of 1.5 which is large for effect size.) The moderating effect of product knowledge on attitude and confidence was examined by looking at the interaction term. It did not achieve statistical significance (B =-.74, t = 1.35, p = .19). However, the relationship pattern of attitude toward the site and confidence across high and low knowledge groups was directionally as hypothesized (Figure 1).

Independent variables
Unstandardized
Coefficients (B)
Standardized
Coefficients (Beta)
t
p
R2
F
(Constant)
-3.92
1.33
.19
Knowledge
(dummy)
3.53
1.69
1.71
.09
Attitude
1.70
.97
2.16
.04
.35
5.3*
Knowledge X
Attitude
-.74
-1.48
1.35
.19

Table 8. Regression result of Ast and product knowledge on confidence in choice.
*p ≤ .01



Figure 1. Relationship between attitude toward the site and confidence.

An additional test (Chow-test)1 was conducted to examine whether the pattern of relationship between attitude toward the site and confidence as shown in Figure 1 is stable across the two knowledge groups. The result was significant (F = 3.9, df = (2, 29), p ≤ .05), which suggests that the effect of attitude toward the site on confidence in choice is significantly different between high and low product knowledge groups. In summary, the regression and Chow test results largely support H2a and H2b. For the low knowledge group, attitude toward the site influences confidence in choice whereas attitude toward the site does not affect choice confidence in the high knowledge group.

Summary and Discussion

This study sought to expand from traditional Aad research to Ast in the online arena. Specifically, we proposed that Ast could positively influence brand choice (Hypothesis 1). Results from the study of Ast provide evidence that favorable Ast increases likelihood for brand choice. That is, the probability of choosing brand A is significantly higher than choosing other brands if a participant evaluates the site of brand A more favorably than any other brand's site. This result is consistent with Biehal, Stephens, and Curlo's (1992) idea that consumers use Aad as a cue to differentiate between brands for choice.

Tests of Hypotheses 2a and 2b were conducted to examine how the relationship between attitude toward the site and confidence in choice differ in high and low product knowledge groups. For the low product knowledge group, we indeed found the effect of attitude toward the site on confidence in choice, but for the high product knowledge group we did not find such an effect. Additional analyses largely supported the moderating role of product knowledge on the effect of attitude toward the site on confidence in choice.

This study found that attitude toward the Web site appears to be a good predictor of consumer brand choice and confidence in that choice. In other words, the extent to which a Web site is able to help form and sustain positive attitude toward the site has a distinct impact on purchase decisions regarding the brand. While it is important to understand the attitude toward the Web site, it is also important to know the underlying perceptual dimensions that contribute to and account for that evaluation. Chen and Wells (1999) suggested entertainment, informativeness, and organization as the underlying dimensions of attitude toward a Web site. They found that the weights users put on each dimension are determined by the types of the Web sites (e.g., dell.com scores high on the informativeness dimension whereas disney.com and pillsbury.com receive high entertainment scores). Therefore, companies and organizations should consider which underlying dimension of attitude toward the Web site is more important and relevant to their prospective users and make building up that dimension of their Web sites the primary goal.

As indicated by the results of H2a and H2b testing, consumer product knowledge seems to moderate the influence of attitude toward the Web site on confidence in choice. In other words, attitude toward the Web site positively affects confidence in choice in the low product knowledge group but it does not influence confidence in high product knowledge group. Therefore, when prospective consumers of a Web site have mostly low product knowledge in areas such as technology products or infrequently purchased items, marketers should pay special attention to helping consumers form positive attitude toward the Web site.

A consumer's confidence in online transactions may be a critical indicator of his/her repeat purchase likelihood and therefore success for online retailers. According to Yahoo/ACNielsen's Internet Confidence Index survey (September 24, 2001), consumer confidence in online purchase has increased seven points from 118 in June, 2001 to 125 in September, 2001. This study suggests that increasing consumer online confidence must come from the online retailer's ability to provide, often through their Web sites, a wide variety of products, useful product information, an appropriate user interface and easy navigation.

Limitations and Suggestions for Future Research

In an attempt to increase our understanding of how attitude toward a Web site influences consumer brand choice and confidence in the choice, an exploratory study was carried out. Useful results that can facilitate future research in this area have been presented. However, caution need to be observed when interpreting the outcome from this study.

Although a statistical power analysis indicates that the sample size used in this study is sufficient, caution should be exercised in generalizing the results to the population. Also, since the sample was gathered conveniently from college students, the ecological validity of the study may be limited.

Second, the product knowledge scales used in this study were based on Alba and Hutchinson's (1987) four different expertise dimensions. The measure, however, does not involve product familiarity and experience which constitute important portions of consumer knowledge. Therefore, care must be taken in interpreting the results of the study concerning product knowledge. Product familiarity and experience should be included to construct a complete indicator of product knowledge in future studies. In general, there are two major approaches to operationalizing and measuring product knowledge: objective knowledge and self-assessed or subjective knowledge (Brucks, 1985; Park, Mothersbaugh, & Feick, 1994). While the former approach contributes to the understanding of the impact of memory contents on a decision maker's evaluation and choice decisions, the latter provides information about a decision maker's systematic biases and heuristics in choice evaluations and decisions (Park & Lessig, 1981). This study dealt with subjective knowledge only. Further research will be needed to investigate how objective and subjective knowledge influence the relationship between attitude, brand choice and confidence in choice. In addition, the role of product knowledge could be dependent upon product types. While it was beyond the scope of this study, an important future research direction would be to examine how product knowledge influences attitude formation in an online environment.

Third, this study did not take into account the effect of brand attitude on brand choice and confidence in choice. It is probable that previously formed brand attitude influences attitude toward the Web site and/or brand choice. Future studies should consider the role of brand attitude in investigating the impact of attitude toward the Web site on consumer online brand choice and confidence in choice. It is strongly suggested that future research incorporate the major constructs used in this study and additional constructs suggested here to form a conceptual model that can be empirically tested through a multi-method approach.

Footnotes

1. The Chow test result was not presented in a table. We employed the Chow test in order to test the stability of regression coefficients between two groups. That is, the question asked is "Do high and low product knowledge groups react in the same way?" Essentially, the how test combines data from both groups by creating an extra predictor that is coded 1 if the data point comes from one group and 0 for the other group. This extra predictor then is multiplied by every predictor in the study, thereby creating additional interaction terms.

References

Aaker, D. A., & Stayman, D. M. (1990). Measuring audience perceptions of commercials and relating them to ad impact. Journal of Advertising Research, 30(4), 7-17.

Alba, J. W. & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal of Consumer Research, 13, 411-452.

Ariely, D. (2000). Controlling the information flow: Effects on consumers' decision making and preferences. Journal of Consumer Research, 27, 233-248.

Bettman, J. R. (1982). A functional analysis of the role of overall evaluation of alternatives in choice processes. Advances in Consumer Research, 9, 87-93.

Betty, S. E. & Smith, S. M. (1987). External search effort: An investigation across several product categories. Journal of Consumer Research, 14(1), 83-92.

Biehal, G., Stephens, D., & Curlo, E. (1992). Attitude toward the ad and brand choice. Journal of Advertising, 21(3), 19-36.

Billings, R. S. & Scherer, L. L. (1988). The effects of response mode and importance on decision-making strategies: Judgment versus choice. Organizational Behavior and Human Performance, 41, 1-19.

Brucks, M. (1985). The effects of product class knowledge on information search behavior. Journal of Consumer Research, 12, 1-16.

Bruner II, G. C. & Kumar, A. (2000). Web commercials and advertising hierarchy-of-effects. Journal of Advertising Research, 40(1/2), 35-42.

Burke, M. C. & Edell, J. A. (1989). The impact of feelings on ad-based affect and cognition. Journal of Marketing Research, 26, 69-83.

Center for Media Research (2003, October). 150 million users on the Internet. Retrieved January 20, 2004 from http://www.centerformediaresearch.com/cfmr_brief.cfm?fnl=031023.

Chen, Q. & Wells, W. D. (1999). Attitude toward the site. Journal of Advertising Research, 39(5), 27-37.

Coupey, E., Irwin, J. R., & Payne, J. W. (1998). Product category familiarity and preference construction. Journal of Consumer Research, 24, 459-468.

Haley, R. I. & Baldinger, A. L. (1991). The ARF copy research validity project. Journal of Advertising Research, 31, 11-32.

Haubl, G. & Trifts, V. (2000). Consumer decision making in online shopping environments: The effects of interactive decision aids. Marketing Science, 19(1), 4-21.

Holbrook, M. B. (1978). Beyond attitude structure: Toward the information determinants of attitude. Journal of Marketing Research, 15, 545-556.

Jahng, J., Jain, H., & Ramamurthy, K. (2000). Effective design of electronic commerce environments: A proposed theory of congruence and an illustration. IEEE Transactions on Systems, Man, and Cybernetics, 30(4), 456.

Johnson, E. J. & Russo, J. E. (1984). Product familiarity and learning new information. Journal of Consumer Research, 11, 542-550.

Kleiser, S. B. & Mantel, S. P. (1994). The dimensions of consumer expertise: A scale development. American Marketing Association, Summer, 20-26.

Levin, I. P. & Gaeth, G. J. (1988). How consumers are affected by the framing of attribute information before and after consuming the product. Journal of Consumer Research, 15, 374-378.

MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23, 130-143.

Miniard, P. W., Bhatla, S., & Rose, R. L. (1990). On the formation and relationship of ad and brand attitudes: An Experimental and Causal Analysis. Journal of Marketing Research, 27, 290-303.

Nielsen/NetRatings (2003, December). Online consumers spent $7.4 billion in November, up 34 percent from 2002.. News Release. Retrieved January 20, 2004 from http://www.nielsen-netratings.com/pr/pr_031209.pdf.

Park, C. W. & Lessig, V. P. (1981). Familiarity and its impact on consumer decision biases and heuristics. Journal of Consumer Research, 8, 223-230.

Park, C. W., Mothersbaugh, D. L., & Feick, L. (1994). Consumer knowledge assessment. Journal of Consumer Research, 21(1), 71-82.

Pastore, M. (2001). Net users worldwide taking commerce online. Retrieved June 13, 2003 from http://cyberatlas.internet.com/big_picture/geographics/article/0,,5911_783851,00.html.

Poh, D. M. H. & Adam, S. (2002). An exploratory investigation of attitude toward the Website and the advertising hierarchy of effects. Retrieved June 1, 2003 from http://ausweb.scu.edu.au/aw02/papers/refereed/poh/paper.html.

Rao, A. R. & Monroe, K. B. (1988). The moderating effect of prior knowledge on cue utilization in product evaluations. Journal of Consumer Research, 15, 253-264.

Rao, A. R. & Sieben, W. A. (1992). The effect of prior knowledge on price acceptability and the type of information examined. Journal of Consumer Research, 19, 256-270.

Russo, J. E. & Johnson, E. J. (1980). What do consumers know about familiar products? Advances in Consumer Research, 7, 417-423.

Spence, M. T. & Brucks, M. (1997). The moderating effects of problem characteristics on experts' and novices' judgment. Journal of Marketing Research, 34, 233-247.

Stevenson, J. C., Bruner II, G. C., & Kumar, A. (2000). Webpage background and viewer attitudes. Journal of Advertising Research, 40(1/2), 29-34.

Swait, J. (2001). The influence of task complexity on consumer choice: A latent class model of decision strategy. Journal of Consumer Research, 28(1), 135-148.

Swait, J. & Adamowicz, W. (2001). Choice environment, market complexity, and consumer behavior: A theoretical and empirical approach for incorporating decision complexity into models of consumer choice. Organizational Behavior and Human Decision Processes, 86(2), 141-167.

Ward, M. R. & Lee, M. J. (2000). Internet shopping, consumer search and product branding. Journal of Product and Brand Management, 9(1), 20-25.

Yahoo/ACNielsen's Internet confidence index survey (2001, September 24). Retrieved June 13, 2003 from http://docs.yahoo.com/docs/pr/release809.html.

Appendix

Constructs and Items Used


Construct
Items
  

Product Knowledge
(Kleiser & Mantel, 1994)

 (7-point scale anchored by "Strongly disagree" and "Strongly agree")

1. I automatically know which brand of laptop computers to buy.
2. I am loyal to one brand of laptop computers.
3. At the place of purchase, I can visually detect my preferred brand without much effort.
4. I can immediately identify my preferred brand even if it is located with other brands of laptop computers.
5. When I purchase my preferred brand, I do not pay attention to the other brands of laptop computers.
6. I enjoy learning about laptop computers.
7. I will search for the latest information on laptop computers before I purchase a brand.
8. I keep current on the most recent developments in laptop computers.
9. I consider myself knowledgeable on laptop computers.
10. My knowledge of laptop computers helps me to understand
very technical information about this product.
11. I use my knowledge on laptop computers to verify that
advertising claims are in fact true.
12. I can recall almost all existing brands of laptop computers
from memory.
13. I can recognize almost all brand names of laptop computers.
14. I can recall product-specific attributes of laptop computers.
15. I can recall brand-specific attributes of the various brands of
laptop computers.

Attitude toward the Web site
(Chen & Wells, 1999)

(5-point scale anchored by "Definitely Disagree" and "Definitely Agree")

1. This Web site makes it easy for me to build a relationship with this company.
2. I would like to visit this Web site again in the future.
3. I am satisfied with the service provided by this Web site.
4. I feel comfortable in surfing this Web site.
5. I feel surfing this Web site is a good way to spend my time.
6. This site is for a brand that I am familiar with.
7. I've visited this Web site before.
8. Compared with other Web sites, I would rate this one as (5-point scale anchored by "One of the worst" and "One of the best")

Confidence in Brand Choice (5-point scale anchored by "Completely uncertain" and "Completely certain")

How confident do you feel about this recommendation?


About the Authors

Byung-Kwan Lee and Ji-Young Hong are doctoral students and Wei-Na Lee is Associate Professor in the Department of Advertising, University of Texas at Austin, CMA 7.142, 1 University Station A1200, Austin, TX 78712-1092.

©Copyright 2004 Journal of Computer-Mediated Communication