The Mediating Role of Perceived Interactivity in the Effect of Actual Interactivity on Attitude toward the Website
Guohua Wu
San Jose State University
Abstract
The purpose of this research is to reconcile seemingly inconsistent findings regarding the effects of interactivity on communication outcomes such as attitude toward the website in two different interactivity research streams (i.e., actual interactivity vs. perceived interactivity) by proposing that perceived interactivity mediates the effect of actual interactivity on attitude toward the website. Empirical evidence supports the mediating role of perceived interactivity. Implications of the findings are discussed in terms of building a theory of interactivity and providing practical insights into interactive advertising and marketing communications.
Introduction
The issue of actual versus perceived interactivity (Williams, Rice, and Rogers 1988), feature-based versus perception-based interactivity (e.g., McMillan, 2000, 2002; Wu 1999), or structural versus experiential interactivity (Liu and Shrum 2002) has generated considerable interest in the interactivity research stream since the construct of interactivity was associated with new media (Rafaeli 1988). Although the distinction between perception and reality of interactivity appears to be philosophical, empirical evidence indicates that perception and reality of interactivity are different. For example, using a web-based content analysis and web-assisted personal interviews, Lee et al. (2004) found that perceptions of interactivity (i.e., perceived interactivity) of three computer manufacturers' websites (i.e., apple.com, dell.com, and hp.com) were different, while the objectively-assessed interactivity (i.e., actual interactivity) was the same among the three websites. It appears that a clear understanding about the subtle distinction between these two similar and interrelated constructs would help explain seemingly inconsistent findings regarding the effects of interactivity on users of new media. Here new media are broadly referred to as "technologies of telecommunication and computing, new user devices (e.g., videodisc machines), and their practical applications in office, home, business, health or educational environments" (Williams, Rice, and Rogers 1988, p. xi).
The World Wide Web is the most recent and popular addition to the new media repertoire, which fascinates both businesses and consumers because of its enormous potential for interaction (i.e., business-with-business, business-with-consumer, and consumer-with-consumers) that otherwise would be impossible or too costly in traditional media. Indeed, as Deighton (1996) put it, interactivity of the Internet has the potential to translate good marketing into good conversation by "putting a more human face on marketplace exchanges without losing the scale of economies of mass marketing" (p. 151). Thus, it seems that interactivity benefits both businesses and consumers.
However, academic research has questioned whether interactivity is always beneficial or desirable (e.g., Ariely 1998; Liu and Shrum 2002). Empirical evidence regarding the effects of interactivity on consumers shows seemingly inconsistent or inconclusive findings. Some researchers have found that interactivity had a positive impact on attitude toward the website (Cho and Leckenby 1999; Hwang and McMillan 2002; Jee and Lee 2002; McMillan 2000; Wu 1999; Yoo and Stout 2001), while others did not find such a positive relationship between interactivity and attitude toward the website (Bezjian-Avery, Calder, and Iacobucci 1998; Coyle and Thorson 2001). One possible reason is that, in these studies, interactivity may have been defined and operationalized differently (Liu and Shrum 2002). Some studies defined interactivity from an objective viewpoint (actual interactivity) and experimentally manipulated the levels of interactivity embodied in a stimulus (Bezjian-Avery, Calder, and Iacobucci 1998; Coyle and Thorson 2001; Fiore and Jin 2003; Haseman, Nuipolatoglu, and Ramamurthy 2001; Raney et al. 2003; Sundar, Kalyanaraman, and Brown 2003). Others conceptualized interactivity from a subjective or perception perspective and measured interactivity using a scale (Cho and Leckenby 1999; Jee and Lee 2002; Hwang and McMillan 2002; Schlosser 2003; Wu 1999; Yoo and Stout). Given that both research streams in interactivity appear to have contributed to our understanding of interactivity, it is time to reconcile the seemingly conflicting findings regarding interactivity's effects. However, no empirical research has been conducted to integrate these two research streams in the literature to achieve a clear understanding about interactivity and its effects. Thus, the purpose of this research is to achieve such a clear understanding by proposing that perceived interactivity mediates the effect of actual interactivity on communication outcomes such as attitude toward the website.
Conceptual Framework and Hypothesis
Actual Interactivity versus Perceived Interactivity
Researchers from a multitude of disciplines ranging from instructional technology, computer-mediated communication, computer science, and information science, to advertising and marketing define interactivity differently. Some scholars lamented that "interactivity is a widely used term with an intuitive appeal, but it is an underdefined concept" (Rafaeli 1988, p. 110), whereas others have avoided using the term "interactivity" by narrowing the meaning of interactivity to user control over information (e.g., Ariely 2000; Klein 2003). It appears that more research needs to be done to cast away misunderstandings of and confusion about interactivity.
McMillan and Hwang (2002) offered an extensive review of the interactivity literature in which they provided the definitions used by scholars in 29 articles. They further classified the researchers' definitions as focusing on (1) process, (2) feature, (3) perception, and (4) a combination of process, feature, and/or perception. Such a classification sheds some light on various definitions and operationalizations of interactivity, but it still falls short in explaining how and why there are conflicting findings in the literature regarding interactivity's effects. It is proposed in this paper that focusing on the dichotomy of actual interactivity vs. perceived interactivity and examining the relationship between the two would provide a productive way of reconciling such conflicting findings in the literature.
To
begin with, it is first necessary to achieve a clear understanding
about what is actual interactivity and what is perceived interactivity.
Actual interactivity could be defined by focusing on the features
of a medium, or capabilities of creating interactive content
or messages (e.g., Hoffman and Novak 1996; Neuman 1991; Rice
and Williams 1984; Rogers 1986; Steuer 1992), or potential
for interaction in general (Rafaeli 1988).
If actual interactivity represents a potential for interaction
in a medium, then different media obviously vary in terms
of such potential. Traditional one-way mass media such as
newspapers, magazines, radio, and television have a low potential
for interaction because they are designed to deliver messages
cost-effectively to a mass audience that has little motivation
to interact with content creators such as contributors or
editors, or commercial message sponsors (i.e., advertisers)
for technical and economic reasons. For example, a newspaper
could provide a response mechanism like a mailing address,
or a telephone number for readers to write letters to the
editor or to call the newspaper directly, but the cost and
inconvenience associated with such a simple interaction prevent
both the newspaper and its readers from interacting on a large
scale.
In contrast, interactive media such as the Web have a high potential for interaction. For instance, an online newspaper allows for various interactions that would be cost-prohibitive to do in an offline newspaper. In the online newspaper environment, readers can click on a writer's email address hyperlink to send a message easily and quickly with almost zero cost. They can also scroll down a page to find other related content, or interact in real time with other readers on a topic of common interest. In addition, readers can click on an ad (e.g., a banner ad), which leads them to a website, where they could obtain more information, or even complete a transaction such as booking a low-fare flight. Apparently, interactive media have a higher potential for interaction or actual interactivity than traditional one-way mass media.
Nevertheless,
actual interactivity can only provide the potential to allow
for interaction, because "[I]nteractivity is potential
adequacy, but it is up to the communicators to realize it"
(Rafaeli, 1988, p. 117). The degree to which "actual
interactivity" (Williams, Rice, and Rogers 1988, p.
174) is fulfilled depends largely on the perceiver. In other
words, whether and to what extent, such higher potential for
interaction in interactive media can be realized to benefit
both interaction participants is largely determined by how
interactions are perceived. Hence, perceived interactivity
appears to play an important role in shaping actual interactivity's
influence on interaction participants.
Perceived interactivity has been defined and operationalized
differently in various studies (see McMillan and Hwang 2002
for a review). In this study, perceived interactivity of websites
by site-visitors (PIsite) is defined as a psychological state
experienced by a site-visitor during the interaction process.
It manifests in three dimensions: (1) perceived control over
the (a) site navigation; (b) the pace or rhythm of the interaction;
(c) the content being accessed, (2) perceived responsiveness
from (a) the site-owner; (b) from the navigation cues and
signs; (c) the persons online, (3) perceived personalization
of the site with regard to (a) acting as if it were a person;
(b) acting as if it wants to know the site visitor; and (c)
acting as if it understands the site visitor (Wu 2000).
In the context of interactive advertising and marketing communications,
marketers must decide on an optimal level of actual interactivity
by selecting from a wide variety of interactive technologies
with different levels of cost and sophistication (e.g., chat-rooms,
online bulletin boards, personalization technology). The customers'
perceptions of actual interactivity are critical input into
that decision process. More importantly, such perceptions
of actual interactivity will potentially influence the effects
of actual interactivity on consumers' attitudes and
behaviors. That is, the impact of actual interactivity on
attitude toward the site appears to hinge on perceived interactivity,
suggesting that the latter might play a mediating role as
shown in Figure 1. In this research, attitude toward the website
was selected as a focal measure of communication outcomes.
Based on the attitude toward the ad research stream in traditional
media (e.g., MacKenzie, Lutz, and Belch 1986), researchers
have indicated that attitude toward the website represents
a useful effectiveness measure of Internet advertising (Jee
and Lee 2002; McMillan, Hwang, and Lee 2003; Rodgers and Thorson
2000; Wu 1999). Some researchers have adapted Aad to measure
attitudes toward the website (e.g., Cho, Lee and Tharp 2001;
McMillan 2000). Chen and Wells (1999) held that attitude toward
the website will gain a parallel status in evaluating effectiveness
as in traditional mass advertising and developed a scale called
Ast to measure attitude toward the website.
Figure
1: A Proposed Framework of the Mediating Role of Perceived
Interactivity in the
Effect of Actual Interactivity on Attitude toward the Website
The Mediating Role of Perceived Interactivity
Many studies have examined the effects of interactivity, but to ensure a meaningful comparison among different studies regarding interactivity's effects, three criteria were used to select previous studies on interactivity for inclusion. First, the effects of interactivity had to include at least one of the three communication outcomes: (1) Attitude toward the ad or the site; (2) attitude toward the brand or the product, and (3) purchase intent or revisit intention. Second, the studies selected here were experimental. Third, only studies conducted after the birth of the World Wide Web (1994) were selected. Applying the three criteria to a range of interactivity studies yielded 14 studies shown in Table 1.
Table
1: A Summary of Empirical Studies on the Effects of Interactivity
on
Attitude toward the Website
Table 1 indicates that the 14 studies can be put into two groups: (a) half construed interactivity as actual interactivity and operationalized the levels of interactivity by varying potential for interaction embodied in the stimulus or medium and (b) half defined interactivity as perceived interactivity and measured interactivity by using an itemized scale. The data in Table 1 can be summarized in three ways. First, all seven studies in the "b" group found a significantly positive relationship between (perceived) interactivity and communication outcomes (i.e., attitude toward the site, attitude toward the brand, or purchase intent). Second, inconsistencies exist in the seven studies in the "a" group: five of them found a significant relationship between (actual) interactivity and communication outcomes (i.e., attitude toward the site, attitude toward the brand, or purchase intent), while the other two in the "a" group did not find a significant relationship between (actual) interactivity and attitude toward the site. Third, one of the seven studies in the "a" group found a negative relationship between actual interactivity and purchase intent (i.e., Bezjian-Avery, Calder, and Iacobucci 1998). Thus, it can be seen that to achieve a clearer understanding of interactivity and its effects, it is of vital importance to recognize the subtle yet critical distinction between actual and perceived interactivity. In addition, it can be seen that perceived interactivity appears consistently to contribute positively toward communication outcomes such as attitude toward the site, attitude toward the brand, and/or purchase intent. Furthermore, the inconsistent findings regarding actual interactivity's effect on attitude toward the website suggest that perceived interactivity might play a mediating role in affecting actual interactivity's impact on attitude toward the website. In other words, when perceived interactivity affects attitude toward the website significantly, the significant relationship between attitude toward the website and actual interactivity would weaken to the extent that it becomes insignificant. It is unusual for the weakening effect to change from significance to zero (Baron and Kenny 1986). Rather, "it is the degree to which the effect is reduced (e.g., the change in regression coefficients) is an indicator of the potency of the mediator" (Holmbeck 1997, p. 602). If the mediating role of perceived interactivity can be established, this might explain why the group "b" studies failed to find a significant relationship between (actual) interactivity and attitude toward the website. Thus, the following hypothesis is formally stated:
H1: Perceived interactivity mediates the effect of actual interactivity on attitude toward the website.
Method
Experimental Design
To
test the hypothesis, an experiment was conducted with 157
undergraduate students from a large southwestern university
in the United States. They participated in exchange for extra
course credit. Forty-three percent were male and fifty-seven
were female. The variable of the level of actual interactivity
(high vs. low) was manipulated. A website of a fictitious
brand of dietary supplement was professionally designed with
two different versions to reflect the high and low levels
of actual interactivity. The website contained a large amount
of nutritional and herbal information adapted from Balch and
Balch (2000), and featured a brand called nutritiongarden.com.
The category of nutritional supplement was considered appropriate
for the sample of college students. College students are health-conscious,
and they were familiar with this product category. In addition,
a pretest with 30 participants from the same university indicated
that eighty percent found nutritional supplement information
on the website interesting and useful.
Consistent with previous research (e.g., Fortin 1997; Frazer
and McMillan 1999; Van Tassel 1988), the level of actual interactivity
was operationalized by varying the presence or absence of
interactive elements. Specifically, a high level of actual
interactivity was characterized by the presence of all six
interactive elements, and a low level of interactivity website
by the absence of the same six interactive elements (see Table
2 for a list of such interactive elements). An "email
hot link" acts as a feedback mechanism. "Java-script
enabled mouse-over effects" refer to the changing of
colors and/or images once a user's mouse moves onto
the link, thereby giving users an illusion of response as
they click through hyperlinks. An "online chat-room"
allows users to engage in a conversation in real-time. A "searchable
pull-down menu" provides users with menu options to
choose from. A "product image" gives users visual
information. "Dynamic creation of content" refers
to the process in which content is generated in real-time,
thereby creating a sense of perceived personalization.
Table 2: Two Versions of the Website
Procedure
When recruited participants arrived at a reserved computer lab, they were randomly assigned to each of the two versions of the nutritional supplement website. They were instructed to browse the website as they normally would do for a maximum of 30 minutes. After they felt their browsing was over, they filled out an online questionnaire hyper-linked from the site. Thirty sessions were conducted in the computer lab with each session having five to fifteen participants.
Measures
Attitude toward the Website. Participants were told to treat the website as an advertisement and rate the website using eleven seven-point semantic differential items adapted from Olney, Holbrook, and Batra's (1991) attitude toward the ad scale. This scale is made up of three components: hedonism, (fun to see/not fun to see, pleasant/unpleasant, entertaining/not entertaining, enjoyable/not enjoyable), interestingness (important/not important, helpful/not helpful, informative/not informative, useful/not useful), and utilitarianism (curious/not curious, boring/not boring, interesting/not interesting) (a = .93).
Perceived interactivity. Wu's (2000) nine-item scale for measuring perceived interactivity (a = .74) was used. This scale was also used by some interactivity researchers with acceptable reliability (e.g., Jee and Lee 2002, a = .89; Sohn and Leckenby 2003, a = .80). (see Appendix 1). Participants' responses were rated on a five-point Likert-type scale with "strongly agree" and "strongly disagree" as anchors.
Results
Participants
were asked to indicate their degree of agreement on a five-point
Likert-type scale (1 = strongly disagree and 5 = strongly
agree) with the statement "I perceived the website to
be highly interactive." There was a significant difference
(F(1,153) = 23.02, p < .0001) between participants exposed
to a high level of actual interactivity (M = 3.59) and those
to a low level of actual interactivity (M = 2.77). Thus, the
manipulation of the level of actual interactivity was successful.
This manipulation is somewhat subjective. In fact, the objective
manipulation check was undertaken during the website pre-testing
in this way: two experienced web designers served as judges
to verify the presence or absence of those interactive elements
shown in Table 2 for each version of the website. Both confirmed
that the version with a high level of actual interactivity
had all the interactive elements while the one with a low
level of actual interactivity had none of these same interactive
elements.
Testing of Hypothesis: The Mediating Effect of Perceived Interactivity
There are two possible ways of testing mediational effects: (a) regression; (b) structural equation modeling. Although some suggest the structural equation modeling is the preferred strategy, particularly useful when one has multiple indicators for the latent variables under investigation, the two ways of testing mediational effects are basically similar (Holmbeck 1997). Given that our proposed framework in Figure 1 is a relatively simple model with one binary variable (level of actual interactivity) and two measured variables (perceived interactivity and attitude toward the site), it was decided to use Baron and Kenny (1986)'s framework to test the mediational effects in this study. Regression analyses were conducted specifically to examine (1) whether the independent variable (i.e., actual interactivity) significantly accounted for variance in the hypothesized mediator (i.e., perceived interactivity), (2) whether variance in the mediator (i.e., perceived interactivity) accounted for variance in the dependent variable (i.e., attitude toward the website), and (3) whether the relationship between the independent variable (i.e., actual interactivity) and dependent variable (i.e., attitude toward the website) would no longer be significant once the variance in the dependent variable accounted for by the mediator was partialled out.
First, a model was fitted in which actual interactivity (Ai) was regressed on perceived interactivity (Pi). The effect of the independent variable (Ai) significantly explained the variance in the hypothesized mediator perceived interactivity (Pi) (t = 4.31, F = 18.61, p < .001, r2 = .11). This result suggested that the mediator Pi was related to the independent variable Ai whose effect it supposedly mediated. Another regression model was run with attitude toward the site (Asite) as the dependent variable and the mediator (Pi) as the independent variable. Perceived interactivity significantly accounted for variance in the dependent variable Asite (t = 5.89, F = 34.69, p < .001, r2 = .25). A third regression model was fitted with Asite as the dependent variable and actual interactivity as an independent variable. A significant result was obtained (t = 3.14, F = 9.87, p < .001, r2 = .08). Finally, a fourth model was conducted with Asite as the dependent variable and actual interactivity and perceived interactivity as independent variables. The effect of actual interactivity was not significant (t = 1.00, p = .32) after the significant effect of the hypothesized mediator perceived interactivity (t = 5.15, p < .001) was partialled out. These results support the hypothesis (see Table 3 for a summary of results).
Table
3: Testing of Hypothesis: The Mediating Effect of Perceived
Interactivity on
Attitude toward the Site
Actual interactivity is a binary variable, which has a smaller variance than that of perceived interactivity, which is an interval variable. This may have explained why actual interactivity does not account for as much variance of Asite as does perceived interactivity. In response to this concern, a two-way ANOVA was conducted with actual interactivity (high and low) and perceived interactivity (high and low at median split) as two factors and Asite as a dependent variable. Table 4 shows that perceived interactivity has a significant effect on attitude toward the website (F = 11.19, df = 1, p < .001) while actual interactivity has no significant effect on attitude toward the website (F = 3.37, df = 1, p = .069). These results are very similar to those from the regression analyses for the mediating effects.
Table
4: Univariate F Test: The Effects of Actual Interactivity and
Perceived Interactivity
on Attitude toward the Site
Furthermore, a path analysis was conducted to assess the fit of the proposed framework in Figure 1 to the resulting data. The model has a perfect fit with the data (Chi-square = 0 and df = 0). The second model has the path coefficient from actual interactivity to attitude toward the website fixed at zero. The second model fit indices are: Chi-square = 2.243, df = 1. Then, the Chi-square difference = 2.243 with the difference of degree of freedom = 1 is not significant, suggesting that no improvement of constraining the path from actual interactivity to attitude toward the website. Therefore, the mediational effect of perceived interactivity is established (Holmbeck 1997). This result is identical to Baron and Kenny's (1986) mediational effects procedure via regression analyses. Figure 2 shows the values of path coefficients and their significance levels under two conditions: (a) when the path coefficient from actual interactivity to attitude toward the website (Lamda-one) is constrained to zero; (b) and when is Lamda-one non-constrained.
Figure 2: The Resulting Path Analysis for the Framework of the Mediating Role of Perceived Interactivity in the Effect of Actual Interactivity on Attitude toward the Website
Discussion
This study demonstrated that the effect of actual interactivity on attitude toward the site is mediated by perceived interactivity. Specifically, it was shown that perceived interactivity is positively related to actual interactivity (Beta = 3.66, t = 4.31, p < .001), that attitude toward the website is positively related to perceived interactivity (Beta = .10, t = 5.89, p < .001), and that attitude toward the website is positively related to actual interactivity (Beta = .64, t = 3.14, p < .001). However, when perceived interactivity plays a mediating role in the relationship between actual interactivity and attitude toward the website, the significant relationship between attitude toward the website and actual interactivity (Beta = .64, t = 3.14, p < .001) becomes insignificant (Beta = .20, t = 1.00, p = .32).
Given such a complex set of relationships among actual interactivity, perceived interactivity, and attitude toward the website, it is not surprising that the current literature on interactivity's effects seem to be inconsistent and confusing. At the same time, with the clear specification of relationships among them, several conclusions can be drawn.
First, the positive relationship between actual interactivity and perceived interactivity indicates that both actual interactivity and perceived interactivity should be considered simultaneously in order to obtain a complete picture of what is interactivity. This suggests that an integrative approach should be used when designing a study on interactivity. In particular, the actual interactivity research stream should strive to incorporate the perspective of the perceived interactivity research stream by measuring perceived interactivity using an itemized scale (e.g., McMillan and Hwang 2002; Wu 2000).
Second, it is of vital importance to recognize the subtle yet critical distinction between actual interactivity and perceived interactivity in order to gain a clearer understanding of interactivity. Such a distinction allows researchers to recognize whether an interactivity study falls under the "actual interactivity" or "perceived interactivity research stream. The actual interactivity research stream usually conceptualizes interactivity as the levels of potential for interaction as embodied in a stimulus (e.g., a website) and manipulates the levels accordingly, while the perceived interactivity research stream defines interactivity as a perceptual variable and measures it using an itemized scale. Clearly, our dichotomous perspective on interactivity helps sort out different studies that have conflicting results. For example, this research has put 14 studies listed in Table 1 into two categories: (a) actual interactivity and (b) perceived interactivity. Those which fall under the perceived interactivity research stream have shown a consistent positive relationship between perceived interactivity and communication outcomes such as attitude toward the website, attitude toward the brand, or purchase intention. However, the results of this study have shown that it is possible for both actual interactivity and perceived interactivity to have an effect on attitude toward the website if both effects are examined independently of each other. Nevertheless, as the mediational effects have revealed, the effect of actual interactivity on attitude toward the website becomes insignificant once both the effects of actual interactivity and perceived interactivity are considered simultaneously.
Finally,
this research has proven the mediating role of perceived interactivity
in affecting the effect of actual interactivity on attitude toward
the website. This represents the most important contribution toward
our understanding of interactivity's effects because it
helps explain the inconsistencies in the interactivity research
streams.
Among the seven studies in the a group of Table 1, five of them
(Fiore and Jin 2003; Haseman, Nuipolatoglu, and Ramamurthy 2001;
Marcias 2003; Raney et al. 2003, Sundar, Kalyanaraman, and Brown
2003) found a positive relationship between actual interactivity
and attitude toward the website, whereas the other two (Bezjian-Avery,
Calder and Iacobucci 1998; Coyle and Thorson 2001) did not find
such a relationship. This seems contradictory, but according to
the conceptual framework in Figure 1, both results are possible,
as revealed in our study, depending on whether the mediating role
of perceived interactivity is taken into account. Let us take
a look at two studies in Table 1 that did not find a significant
effect of actual interactivity on attitude toward the website.
Bezjian-Avery, Calder, and Iacobucci (1998) found no significant
relationship between (actual) interactivity and attitude toward
the ad. They held that "interactivity is fundamentally the
ability to control information" (p. 24). Though they did
not conduct a manipulation check on the independent variable of
(actual) interactivity, it could be inferred that in their study
subjects' perceived control over information might not have
been successfully manipulated because "perceptions of control
may be broader and easier to influence than actual control"
(Klein 2003, p. 58). In fact, when investigating the role of telepresence,
Klein (2003) manipulated user control in a very similar way to
what Bezjian-Avery, Calder, and Iacobucci (1998) did with (actual)
interactivity. Thus, it is possible that Bezjian-Avery, Calder,
and Iacobucci's study (1998) might have suffered from lack
of consideration of the confounding factor of "perceived
control over information," which is very close to one of
three dimensions of our conceptualization of perceived interactivity
- perceived control. Accordingly, it is not surprising to
see their non-significant finding regard the effect of (actual)
interactivity on attitude toward the ad.
Coyle and Thorson's (2001) operationalized levels of interactivity
as the number of choices and the presence or absence of a clickable
image. That is, they also manipulated the ability to control,
just as Bezjian-Avery, Calder, and Iacobucci (1998) and Klein
(2003) did. Thus, a similar case could be made for Coyle and Thorson's
(2001) study with regard to the insignificant relationship between
(actual) interactivity and attitude toward the website.
As for the five studies in the "a" group that found
a positive relationship between actual interactivity and attitude
toward the website, it is very possible that these positive effects
might not have been significant if the mediating role of perceived
interactivity had been considered, as revealed in our study.
Implications and Conclusions
This research has several important implications for building a theory of interactivity and offering practical insights for interactive advertising and marketing communications practice. It has proposed and tested a framework for integrating actual interactivity and perceived interactivity research streams by clearly specifying a network of relationships among actual interactivity, perceived interactivity, and attitude toward the website. In particular, the specification of the mediating role of perceived interactivity in the effect of actual interactivity in the attitude toward the website has illuminated the conflicting findings in the actual interactivity research stream. A theory of interactivity is incomplete without considering both actual and perceived interactivity. Similarly, a study on interactivity will likely suffer from producing inconclusive results regarding interactivity's effects if the mediating role of perceived interactivity in the effect of actual interactivity on attitude toward the website is not considered. We recommend that future researchers in interactivity research seriously consider the mediating role of perceived interactivity in the effect of actual interactivity on communication outcomes.
In recognizing the subtle yet critical distinction between actual interactivity and perceived interactivity, interactive advertising and marketing practitioners can see more clearly that it is up to both the target audience and the marketer to fulfill the actual interactivity or potential for interaction enabled by interactive technologies. Therefore, the marketer should constantly keep in touch with the perceptions of actual interactivity by the audience. These perceptions will help determine an optimal level of actual interactivity that would be adequately realized by the audience. In doing so, return on investment in interactive technologies can be maximized.
Despite the important implications for both theory and practice of interactivity, it should be noted that this study suffers from several limitations of its own. First, this study only tested one website, and future research needs to use more websites in different categories to see whether the same results will hold. Second, given the ever-changing nature of interactive and computer technologies, what was perceived to be an interactive element signifying a high level of actual interactivity in the past may not be perceived interactive in the present or in the future. The level of actual interactivity manipulated in this study was useful at the time the study was conducted. Future research may need to revise types of interactive elements incorporated in a website to operationalize the levels of actual interactivity.
In conclusion, this research has successfully achieved its objective of reconciling the seemingly inconsistent findings in the literature regarding interactivity's effects by having proposed and tested a framework for a network of relationships among actual interactivity, perceived interactivity, and attitude toward the website. This framework effectively integrates the actual interactivity and perceived interactivity research streams, thereby pointing to a new direction in research on the construct of interactivity and its effects.
References
Ariely, Dan (1998), Controlling the Information Flow: Effects on Consumers' Decision-Making and Preferences, unpublished doctoral dissertation, Duke University.
--- (2000), "Controlling the Information Flow: Effects on Consumers' Decision-Making and Preferences," Journal of Consumer Research, 27 (September), 233-248.
Balch, Phyllis A. and James F. Balch (2000), Prescription for Nutritional Healing: A Practical A-to-Z Reference to Drug-free Remedies Using Vitamins, Minerals, Herbs and Food Supplements, New York: Avery Publishing Group.
Baron, Reuben M. and David A. Kenny (1986), "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations," Journal of Personality and Social Psychology, 51 (December), 1173-1182.
Bezjian-Avery, Alexa, Bobby Calder, and Dawn Iacobucci (1998), "New Media Interactive Advertising vs. Traditional Advertising," Journal of Advertising Research, 38 (July/August), 23-32.
Chen, Qimei and William D. Wells (1999), "Attitude toward the Site," Journal of Advertising Research, September/October, 11-37.
Cho, Chang-Hoan and John D. Leckenby (1999), "Interactivity as a Measure of Advertising Effectiveness," in Proceedings of 1999 Conference of the American Academy of Advertising, Marilyn S. Roberts, ed., Gainesville, FL: American Academy of Advertising, 162-179.
Cho, Chang-Hoan, Jung-Gyo Lee and Marye Tharp (2001), "Different Forced-Exposure Levels to Banner Advertisements," Journal of Advertising Research, 41 (Jul/Aug), 45-56.
Coyle, James E. and Esther Thorson (2001), "The Effects of Progressive Levels of Interactivity and Vividness in Web Marketing Sites," Journal of Advertising, 30 (Fall), 65-77.
Deighton, John (1996), "The Future of the Interactive Marketing," Harvard Business Review, 74 (6), 151-62.
Fiore, Ann Marie and Hyun-Jeong Jin (2003), "Influence of Image Interactivity on Approach Responses towards an Online Retailer," Internet Research, 13 (1), 38-48.
Fortin, David R. (1997), The Impact of Interactivity on Advertising Effectiveness in the New Media, unpublished doctoral dissertation, University of Rhode Island.
Frazer, Charles and Sally McMillan (1999), "Sophistication on the World Wide Web: Evaluating Structure, Function and Commercial Goals of Web Sites," in Advertising and the World Wide Web, David W. Schumann and Esther Thorson, eds., Mahwah, NJ: Erlbaum Associates.
Haseman,
William D., Vichuda Nuipolatoglu, and K. Ramamurthy (2002), "An
Empirical Investigation of the Influences of the Degree of Interactivity
on User-outcomes in a Multimedia Environment," Information
Resources Management Journal, 15 (2), 31-48.
Hoffman, Donna and Thomas P. Novak (1996), "Marketing in
Hypermedia Computer-Mediated Environments: Conceptual Foundations,"
Journal of Marketing, 60 (3), 50-58.
Holmbeck,
Grayson N. (1997), "Toward Terminological, Conceptual, and
Statistical Clarity in the Study of Mediators and Moderators:
Examples from the Child-Clinical and Pediatric Psychology Literatures,"
Journal of Consulting and Clinical Psychology, 64 (4),
599-610.
Hwang, Jang-Sun and Sally J. McMillan (2002), "The Role
of Interactivity and Involvement in Attitude toward the Web Site,"
in Proceedings of the 2002 Conference of the American Academy
of Advertising, Avery. M. Abernethy, ed., Auburn, AL: American
Academy of Advertising, 10-17.
Jee, Joonhyung
and Wei-Na Lee (2002), "Antecedents and Consequences of
Perceived Interactivity: An Exploratory Study," Journal
of Interactive Advertising, 3 (Fall), <http://www.jiad.org/vol3/no1/jee/index.htm>.
Lee, Se-Jin, Wei-Na Lee, Hyojin Kim, and Patricia A. Stout (2002),
"A Comparison of Objective Characteristics and User Perception
of Web Sites," Journal of Interactive Advertising,
4 (2), <http://www.jiad.org/vol4/no2/lee/index.htm>.
Liu, Yuping and L. J. Shrum (2002), "What Is Interactivity
and Is It Always Such a Good Thing? Implications of Definition,
Person, and Situation for the Influence of Interactivity on Advertising
Effectiveness," Journal of Advertising, 31 (Winter),
53-64.
Klein,
Lisa R. (2003), "Creating Virtual Product Experiences: The
Role of Telepresence," Journal of Interactive Marketing,
17 (Winter), 41-55.
Macias, Wendy (2003), "A Preliminary Structural Equation
Model of Comprehension and Persuasion of Interactive Advertising
Brand Web Sites," Journal of Interactive Advertising, 3
(2), .
MacKenzie, Scott B. and Richard J. Lutz, and George E. Belch (1986),
"The Role of Attitude toward the Ad as a Mediator of Advertising
Effectiveness: A Test of Competing Explanations," Journal
of Marketing Research, 23 (May), 130-143.
McMillan, Sally J. (2000), "Interactivity Is in the Eye
of the Beholder: Function, Perception, Involvement, and Attitude
toward the Web Site," in Proceedings of the 2000 Conference
of the American Academy of Advertising, Mary A.
Shaver, ed., East Lansing, MI: American Academy of Advertising,
71-78.
--- (2002), "A Four-part Model of Cyber-Interactivity:
Some Cyber-spaces are More Interactive than Others," New
Media and Society, 4 (2), 271-91.
--- and Jang-Sun Hwang (2002), "Measures
of Perceived Interactivity: An Exploration of the Role of Direction
of Communication, User Control, and Time in Shaping Perceptions
of Interactivity," Journal of Advertising, 31 (Fall), 29-42.
---, ---, and Guiohk Lee (2003),
"Effects of Structural and Perceptual Factors on Attitudes
toward the Website," Journal of Advertising Research, 43
(December), 400-409.
Neuman, W. Russell (1991), The Future of the Mass Audience. Cambridge, MA: Cambridge University Press.
Olney, Thomas, Morris Holbrook, and Rajeev Batra (1991), "Consumer Responses to Advertising: The Effects of Ad Content, Emotions and Attitude toward the Ad on Viewing Time," Journal of Consumer Research, 17 (March), 440-453.
Rafaeli, Sheizaf (1988), "Interactivity from New Media to Communication," in Advancing Communication Science: Merging Mass and Interpersonal Processes, Robert P. Hawkins, John M. Wiemann, and Suzanne Pingree, eds., Beverly Hills, CA: Sage, 110-134.
Raney, Arthur A., Laura M. Arpan, Kartik Pashupati, and Dale A. Brill (2003), "At the Movies, on the Web: An Investigation of the Effects of Entertaining and Interactive Web Content on Site and Brand Evaluations, Journal of Interactive Marketing, 17 (4), 38-53
Rice,
Ronald and F. Williams (1984), "Theories Old and New: The
Study of New Media," in The New Media: Communication, Research
and Technology, R. Rice & Associates, eds., Beverly Hills,
CA: Sage.
Rodgers, Shelly and Esther Thorson (2000), "The Interactive
Advertising Model: How Users Perceive and Process Online Ads,"
Journal of Interactive Advertising, 1 (1), .
Rogers, Everett (1986), Communication Technology: The New Media in Society, New York: Free Press.
Schlosser, Ann E. (2003), "Experiencing Products in the Virtual World: The Role of Goal and Imagery in Influencing Attitudes versus Purchase Intentions," Journal of Consumer Research, 30 (September), 184-198.
Sohn, Dongyoung and John Leckenby (2003), "The Role of Expected Interactivity in Interactive Ad Processing," in Proceedings of the 2003 Conference of the American Academy of Advertising, Les Carlson, ed., Clemson: SC, American Academy of Advertising, 115-124.
Steuer, Jonathan (1992), "Defining Virtual Reality: Dimensions Determining Telepresence," Journal of Communication, 42 (Fall), 73-93.
Sundar, S. Shyam, Sriram Kalyanaraman, and Justin Brown (2003), "Explicating Web site interactivity: Impression formation effects in political campaign sites," Communication Research, 30 (February), 30-59.
Van Tassel, Joan-Marie (1987), "Affect, Involvement, and Interactivity: Modifying Health Behavior Using the Interactive Videodisc," unpublished doctoral dissertation, University of Southern California.
Williams, Frederick, Ronald Rice, and Everett Rogers (1988), Research Methods and the New Media, New York, NY: The Free Press.
Wu, Guohua (1999), "Perceived Interactivity and Attitudes toward Web Sites," in Proceedings of the 1999 Conference of the American Academy of Advertising, Marilyn S. Roberts, ed., Gainesville, FL: American Academy of Advertising, 254-262.
--- (2000), "The Role of Perceived Interactivity in Interactive Ad Processing," unpublished doctoral dissertation, The University of Texas at Austin.
Yoo, Chan Y. and Patricia A. Stout (2001), "Factors Affecting Users' Interactivity with the website and the Consequences of Users' Interactivity," in Proceedings of the 2001 Conference of the American Academy of Advertising, Charles. R. Taylor, ed., Villanova, PA: American Academy of Advertising, 53-60.
Appendix 1
The nine-item Likert scale for measuring perceived interactivity of websites (Wu 2000):
1. I was
in control of my navigation through this Web site.
2. I had some control over the content of this Web site that I
wanted to see.
3. I was in total control over the pace of my visit to this Web
site.
4. I could communicate with the company directly for further questions
about the company or its products if I wanted to.
5. The site had the ability to respond to my specific questions
quickly and efficiently.
6. I could communicate in real time with other customers who shared
my interest in this product category.
7. I just had a personal conversation with a sociable, knowledgeable
and warm representative from the company.
8. The Web site was like talking back to me while I clicked through
the website.
9. I perceive the Web site to be sensitive to my nutritional information
needs.
Note: The scale ranged from Strongly Agree to Strongly Disagree.
Acknowledgments
This research is part of the author's doctoral dissertation. The author would like to acknowledge financial support from the 1999-2000 Continuing Fellowship of The University of Texas at Austin. The author wishes to thank his dissertation committee members -- John Leckenby, Patricia Stout, Wei-Na Lee, Frederick Williams, and Andrew Whinston for their support and encouragement. In addition, he wishes to thank Sally McMillan, the three anonymous JIAD reviewers, Diana Stover, and Therese Louie for their helpful comments.
About the Author
Guohua Wu (Ph.D., The University of Texas at Austin), Assistant Professor of Advertising, School of Journalism & Mass Communication, San Jose State University. His research interests include interactive advertising and marketing, cross-cultural consumer behavior, and e-commerce. He has published in International Journal of Advertising, Journal of Consumer Marketing, and Journal of Current Issues and Research in Advertising. Email: gwu@casa.sjsu.edu
