• Home
• Issues
• Search
• Submissions
• Editorial Board
• Editorial Staff
• About
• CFP
• RSS

Vol 2 No 2 Spring 2002PDF Download

The Interactive Advertising Model Tested:
The Role of Internet Motives in Ad Processing

Shelly Rodgers

University of Minnesota-Twin Cities

Abstract

Two, one-factor, within-subjects experiments were conducted to examine the role of Internet motives on responses to four types of banners (communicate, research, shop, and surf). Participants were each exposed to a total of 12 ads, or 3 different ads for each banner type. There were three dependent variables - attitude toward the ad, ability to persuade, and intent to click. The moderating variable was Internet motives, which had four levels (communicate, research, shop, and surf). Study 1 sampled a group of college students (N=106) and Study 2 sampled non-student adults (N=38). Results indicate that Internet motives influenced the strength of banner type on ad responses, but not for every banner ad examined. The hypothesized feature-to-motive association effect was found to some degree with the student sample, but was not found with the non-student sample. Evidence suggests that Internet motives serve different functions for students and non-students. Implications of the findings are discussed.

Introduction

Motives are one of the driving forces of consumer behavior. A motive is a desire to do something, an activated state that contains both energy and direction (Deci and Ryan 1985). Thus, consumer motives are drives to satisfy particular needs and wants. Recent attention has been devoted to examining Internet motives. The majority of these studies have focused on identifying a range of Internet motives, including acquiring information (Eighmey 1997), chatting or emailing (Stafford and Stafford 1998), or shopping (Maignan and Lukas 1997). In fact, several scales have been developed to measure Internet motives (e.g., Korgaronkar and Wolin 1999; Rodgers and Sheldon 2002). The assumption guiding these studies is that better predictions about web-related attitudes and behaviors can be made by understanding the very motives that drive Internet use (Rodgers and Cannon 2000). Consumer motives are believed to govern Internet behavior, subsequently influencing any processing of ads encountered while online (Rodgers and Thorson 2000).

In keeping with this line of inquiry, the present study examines the role of Internet motives in processing one specific type of Internet ad - banners. It is argued that Internet motives serve to structure consumers' choice tasks when they make decisions about which Internet ads to attend to and which ads to ignore. After reviewing the relevant literature, the findings from two experiments are presented to illustrate the role of Internet motives in audience processing of banner ads.

Literature Review

Understanding the role of Internet motives in audience processing of ads is in keeping with an important approach in contemporary motivation theory, namely, functionalism. According to this perspective, the meaning of a behavior can only be understood with reference to its function for the person who behaves (Snyder and Cantor 1998). For example, acts of volunteerism (Clary et al. 1998) and sexual behavior (Cooper, Shapiro, and Powers 1998) have been shown to serve very different needs for different people. The assumption in motivation studies is that humans are inherently active. This assumption translates well to the Internet - a medium where consumers are thought to be actively engaged (Hoffman and Novak 1996). In its simplest form, the Internet requires action on the part of the consumer from the very moment the computer is switched on. While it is recognized that varying levels of involvement, or activity, are possible online, the overall assumption is that at least some level of action is necessary for online behavior to take place (Rodgers and Cannon 1999).

Consumers are further assumed to be rational decision-makers, deciding which objects to attend to and which to ignore, based on their goals and motives. Thus, when consumers decide to attend to an Internet ad, they do so based on rational choice, guided by their goal-directed motives, or "action plans" (Cannon, Richardson, and Yaprak 1998). However, the ability to choose rationally among options is limited by a capacity to process information. Referred to as bounded rationality, such limitations include limited short-term memory and limited computational capabilities (Bettman, Luce, and Payne 1998). This perspective on consumer decision-making acknowledges that, despite whatever motives a consumer might have, behavior is also shaped by an interaction of the human capacity to process information and the properties of the task environment (Simon 1990). Thus, it is believed that Internet behavior is governed by an interaction between the individual's motives, limited cognitive capacity, and properties of the environment -- of which Internet ads are a part.

From a broad perspective, Internet ads are conceptualized as external forces that can cause energy and direction in the person. Thus, the decision to attend to or ignore an Internet ad is guided, at least in part, by the consumer's motives. All Internet motives are goal-directed to some degree or another. Motives therefore influence processing by improving consumers' ability to effectively judge whether an Internet ad will help accomplish or deter from the motive. This does not imply that consumers are always consciously aware of their motives, because some goal-directed behaviors can occur unconsciously (see Bargh 1990). Nevertheless, this perspective illustrates the function of motives in serving to structure choice tasks.

Li and Bukovac (1999) conducted an experiment to examine consumers' cognitive responses as a function of banner ad size (large, small) and type (animated, static) and user mode (information-seeking, web-surfing). Although main effects were found for banner size (larger banners outperformed smaller banners) and type (animated outperformed static), no main effects were found for user-mode. That is, whether the individual had an information-seeking or web-surfing mode made no difference on how that individual responded to the ads. However, user-mode was found to interact with banner size to influence cognitive responses such that web-surfers were more likely to click on large banners than information-seekers. In explanation of this finding, the authors propose that web-surfers were less focused on the task at hand than information-seekers and, consequently, had greater cognitive ability to attend to prominent ads. Consistent with the work of Bagozzi and Dholakia (1999), this explanation recognizes that cognitive processes such as attention and selectivity can be influenced by goal-directed motives, or behaviors. Although the exact mechanisms by which this process occurs are not completely clear, results suggest that motives can interact with ad features to influence cognitive responses.

Based on the foregoing discussion, the following interaction hypothesis was tested:

H1: Internet motives will interact with banner type to influence ad responses.

The general assumption of functionalism is that people are most motivated when situational opportunities "match" their personal needs. For example, Clary and colleagues (1998) found that volunteerism was more likely when the persuasive message prompting volunteerism matched the motive for participating as a volunteer. When volunteerism was motivated by a need to enhance social connections, brochures presenting messages that matched this motive were more persuasive than brochures presenting alternative messages.

Thus, consumer responses to Internet ads are expected to be partly a function of the ad's perceived relevance to the motive. Features that are relevant in terms of a consumer's motives are presumed to be more effective at yielding positive cognitive and behavioral responses than ads that are not goal-congruent. Evidence suggests that relevant banner ads outperform irrelevant banner ads (Cho and Leckenby 1999), and that individuals will direct more effort into examining objects they believe will help them reach their goals (Bettman, Luce, and Payne 1998). In effect, a motive may operate as a frame of reference by helping consumers selectively process motive-congruent ads.

If this "matching hypothesis" holds true for the Internet, then online consumers should be most motivated when situational opportunities "match" their personal needs. Thus, if an online consumer is motivated to use the Internet to shop, banner ads that match this motive will presumably be more persuasive than banner ads that do not (Rodgers and Sheldon 2002). In fact, this is the very premise on which Internet ad models such as key word searches and search engine optimization are based. Due to the limited research on this topic, however, the following research question was posed:

RQ1: To what extent will Internet motives interact with banner type to influence responses when the banner ad matches the corresponding motive?

Method: Study 1 - Student Sample

Participants and Design

A total of 107 undergraduate journalism and advertising students from a large Midwestern university participated in Study 1. One individual left large portions of the survey blank and, consequently, was dropped from the analysis. This left a total sample size of 106 (78 females, 26 males, 2 left blank). Students received extra credit for their participation. The experiment was a one-factor, repeated measures ANOVA, which is often referred to as a within-subjects design (Lomax 2001). This means that each participant responded to each level of message appeal - communicate, research, shop and surf. The strength of this design is its ability to control for individuals differences, thereby increasing the sensitivity of the measures (Lomax 2001).

Procedure

The experiment took place in a classroom setting. Participants were asked to give their opinions on 12 different banner ads, as part of a study on audience processing. The participants were shown the 12 ads one at a time in random order. Each ad was projected on a large screen at the front of classroom using PowerPoint. After viewing each ad, participants filled out the items comprising the dependent and moderating measures, as well as the 12 manipulation checks. Participants were debriefed at the end of the experiment.

Ad Stimuli

A total of 16 banner ads were taken directly from the Internet and manipulated so that the four message appeals matched the corresponding four motives. For example, the banner that promoted online shopping corresponded with the shopping motive. Likewise, banners that promoted e-mail or chat rooms corresponded with the communicating motive, and so on. The 16 banner ads were pre-tested using a sample of 47 journalism undergraduates (33 women, 14 men) at a large, Midwestern university. The sample used for the pretest was different from that used for the experiment.

The primary purpose of the pretest was to determine whether the matches between the motive and banner ad were successful. This was accomplished with the statement, "Which of these items does the banner advertise?" followed by four options: shopping, surfing, getting information, and communicating/socializing. The goal was to have a reliability of at least 75% to ensure that each banner corresponded strongly with only one of the four motives, but not the other three. A reliability of 75% is typical of those used to determine agreement in content analyses (Wimmer and Dominick 1997). All but two banner ads met the 75% minimum reliability. These two banners (1 surf, 1 communicate) were dropped from the experiment, as were an additional two banners (1 shop, 1 research) to balance the design, leaving a total of 12 banner ads (three for each of the four types) (see Appendix).

Independent Variable

The independent variable is banner type. There are four levels - research, shop, communicate, and surf. The four levels were selected to correspond with the four Internet motives so that the matching hypothesis could be tested.

Dependent Variables

There are three dependent variables - attitude toward the ad (Aad), ability to persuade, and intent to click. All dependent variable were measured on five-point scales. The scale for Aad had five items: not interesting/interesting; not enjoyable/enjoyable; not effective/effective; not likable/likable; and, not favorable/favorable (Holbrook and Batra 1987; Mitchell and Olson 1981). Ability to persuade had four items: not convincing/convincing; not engaging/engaging; not intriguing/intriguing; and, not compelling/compelling. The nine items comprising Aad and ability to persuade were factor analyzed, followed by Varimax rotation. Two factors emerged. The first factor (Aad) had factor loadings ranging from .81 to .89, and the second factor (ability to persuade) had factor loadings ranging from .78 to .87. Factor one had an eigenvalue of 4.75, and factor two had an eigenvalue of 1.94, accounting for 74% of the total variance explained.

Because Aad and ability to persuade were measured for each of the 12 ads, it was necessary to compute individual alpha coefficients to ensure that the measures were reliable for each ad. Cronbach alphas for Aad ranged from a low of .90 to a high of .95, for an average alpha coefficient of .93. Alphas for ability to persuade ranged from a low of .63 to a high of .94, for an average alpha coefficient of .87.

Intent to click was measured with one, five-point semantic differential scale by asking participants to rate the likelihood that they would click on the banner ad, followed by a scale of 1 (not likely at all) to 5 (very likely).

Moderating Variable

The Web Motivation Inventory (WMI) was used to measure Internet motives (Rodgers and Sheldon 2002), the moderating variable. There were twelve items, three for each of four motives. Participants were instructed to "Please circle the answer that is closest to how you feel about the World Wide Web," followed by the items: e-mail other people, connect with my friends; make a purchase; do research; explore new sites; buy things; communicate with others; get information I need; surf for fun; find interesting web pages; purchase a product I've heard about; and, find out things I need to know. All items were measured with Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). Consistent with Rodgers and Sheldon (2002), four factors emerged from the factor analysis and Varimax rotation. As shown in Table 1, the four factors accounted for 80% of the total variance. The means and standard deviations of participants' Internet motives can be found in Table 2.

Table 1. Factor Analysis Results of the Web Motivation Inventory

N=106

Table 2. Means and Standard Deviations of Internet Motives - Students

Manipulation Check

Consistent with the pretest, the manipulation check consisted of asking participants to respond to the statement, "Which of these items does the banner advertise?" followed by four options: shopping, surfing, getting information, and communicating/socializing. As shown in Table 2, the .75 agreement was met for every banner ad except one. Surf 3 fell just below the .75 reliability level (.74). However, Surf 3 was included in the analyses because the .75 level has typically referred to the agreement level of two individuals. The fact that 78 out of 106 individuals agreed that the surf banner ad had a surfing appeal is pretty strong evidence of a successful manipulation.

Table 3. Manipulation Check Study 1

Statistical Analysis

A total of 12 repeated measures ANOVAs were conducted. This enabled examination of the banner type x Internet motive effect for each of the dependent measures. This was accomplished by entering three banners of the same type (e.g., surf 1, surf 2, surf 3) simultaneously with all four Internet motives as the covariates. The procedure was repeated for each of the banner types on all three dependent variables. The findings were assessed with a p-value of .10 or less. Higher levels of significance than the .05 or .01 typically used in social scientific research have been recommended for studies that have practical or managerial importance (Kinnear and Taylor 1991).

Results: Study 1 - Student Sample

Test of Hypothesis and Research Question

Results from the repeated measures ANOVAs indicate a significant banner type x Internet motive interaction for every banner type examined on at least two of the dependent measures used here (H1). For example, the communication motive interacted with the surf banner ad to influence Aad and ability to persuade.

The research question was posed to examine the extent to which Internet motives would influence ad responses when the banner ad corresponded with that motive. Results indicate a match-up effect, but not for every motive. For example, a surf motive and surf banner influenced two of the dependent measures (Aad and ability to persuade) and a shopping motive and shopping banner influenced one of the dependent measures (ability to persuade). However, no interaction effects were found for communication or research banners matching a communicate or research motive. The results for both the hypothesis test and research question are shown in Tables 4-7.

Table 4. Repeated Measures ANOVA Results of the Message Appeal x Communicate Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 5. Repeated Measures ANOVA Results of the Message Appeal x Surf Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 6 Repeated Measures ANOVA Results of the Message Appeal x Research Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 7 Repeated Measures ANOVA Results of the Message Appeal x Shop Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Discussion: Study

As much as Study 1 demonstrates the moderating role of motives in processing banner ads, it is important to recognize that the findings were based on a group of respondents whose motives may differ from the general population of adults. Thus, the purpose of Study 2 was to replicate Study 1 using a sample whose Internet motives may differ, and who represent greater diversity than college students do in terms of age, income, education, and occupation.

Method: Study 2 - Non-student Sample

Participants

A total of 38 non-student adults (15 females, 23 males) participated in Study 2. Participants were recruited using a snowball technique. This entailed having a group of undergraduate advertising students at a large Midwestern university invite non-students, 18 and older, to participate in the experiment. Students received extra credit for recruiting the participants.

Participant ages ranged from 18 to 56, with a mean age of 32 (SD=12.00). The majority of the participants (47%) had a college degree, 23% had at least some graduate school, and 11% had either a high school diploma or some college work. Occupations included white collar (40%), management (18%), professional (13%), blue collar (11%), other (16%), and one individual (3%) left this item blank. Incomes ranged from $10,000-20,000 to $101,000 or more, with a median income of $31,000-40,000. Thus, the goal of having a more diverse sample than college students was met to some extent, at least in terms of age, education, occupation, and income.

Ad Stimuli

The same 12 banner ads used in Study 1 were printed off in four-color and collated into a packet. For consistency, the ads were presented in the same order as in Study 1.

Dependent Variables

The instrument in Study 2 was exactly the same as the instrument used in Study 1. The dependent measures were Aad, ability to persuade, and intent to click. A factor analysis with a Varimax rotation revealed the same two-factor solution as in Study 1. For example, Factor 1 (Aad) had an eigenvalue of 4.46, compared to the eigenvalue of 4.75 reported in Study 1. Factor 2 (ability to persuade) had an eigenvalue of 2.28, compared to the eigenvalue of 1.94 reported in Study 1. Both factors accounted for 75% of the total variance, compared to 74% of total explained variance reported in Study 1. Intent to click was measured on the same, five-point scale as in Study 1.

As in Study 1, individual alpha coefficients were run for the dependent measures to ensure that all measures were reliable for the 12 ads. Cronbach alphas for Aad ranged from a low of .89 to a high of .97, for an average alpha coefficient of .94. Alphas for ability to persuade ranged from a low of .76 to a high of .94, for an average alpha coefficient of .89. Both coefficients are similar to those reported in Study 1, .93 and .87, respectively.

Analysis

The exact analysis reported in Study 1 was used in Study 2. The same two factors emerged from the factor analysis conducted on the dependent measures, and the same four factors emerged for the moderating variable - Internet motives. To limit the number of tables shown here, however, the results of the factor analysis were excluded. Similar to Study 1, all p-values of less than .10 were reported.

Manipulation Check

Study 2 contained the exact items used for the manipulation check in Study 1. However, because the results were identical to those reported in Study 1, as well as a need to limit the number of tables shown, the table showing the results of the manipulation check for Study 2 was excluded.

Results: Study 2 - Non-student Sample

Results indicate that the non-student and student samples differed on three of the Internet motives. For example, non-students were more likely than students to shop and surf, but were less likely to research. Students and non-students did not differ on the communication motive. The means and standard deviations for the Internet motives of the non-student sample are shown in Table 8.

Table 8 Means and Standard Deviations of Internet Motives - Non-students

Similar to Study 1, the results in Study 2 indicate a significant banner type x Internet motive interaction effect on the dependent measures. For example, the communication motive interacted with the research banner to influence intent to click. Likewise, the shopping motive interacted with the research banner to influence Aad and ability to persuade (H1). However, unlike the results of Study 1, this finding was not replicated for all four Internet motives. The surfing motive did not interact with any of the banner ads to predict ad responses. The results of Study 2 also differ from those reported in Study 1 in that an interaction between the matching motive and banner type was not found for any of the motives (RQ1). The results are summarized in Tables 9-12.

Table 9 Repeated Measures ANOVA Results of the Message Appeal x Communicate Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 10 Repeated Measures ANOVA Results of the Message Appeal x Surf Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 11 Repeated Measures ANOVA Results of the Message Appeal x Research Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

Table 12 Repeated Measures ANOVA Results of the Message Appeal x Shop Motive Effect

Aad=Attitude Toward the Ad
AP=Ability to Persuade
IC=Intent to Click

General Discussion

The purpose of this research was to examine the role of Internet motives in ad processing of Internet ads. The results suggest that Internet motives moderate, or interact with, banner type to influence cognitive responses. For example, a communication motive interacted with a surfing banner to influence cognitive attitude toward the ad and ability to persuade for the student sample, and intent to click for the non-student sample. Likewise, the shopping motive interacted with surfing and shopping banners in Study 1, and researching and communicating banners in Study 2, to influence cognitive responses.

The results of this study further suggest that some banner ads will be ineffective presumably because of the Internet motive itself. For example, Study 2 found that the research motive was one of the least effective motives for determining cognitive responses to the banner ads. Whether this is because people in "research mode" are highly focused on the task, as suggested by Li and Bukovac (1999), is unclear. Results from Study 1 demonstrate that individuals with a research mode preference also attend to and process ads that promote shopping. One explanation for this finding is that the research motive serves a different function for students than it does for adults. For example, students presumably use the Internet to research term papers and school projects, whereas non-students presumably use the Internet to research information about products, services, health-related questions and so on. Thus, it may be that Internet motives function differently for different people, thereby differentially affecting ad responses.

Several interesting patterns of feature-to-motive associations emerged as the result of this study. For instance, students with a shopping motive appear to be open to evaluating a variety of ads, including surfing and shopping ads. In contrast, non-students with a shopping motive appear to mostly evaluate banner ads with a research appeal. This finding lends additional support to the argument that some individuals are more goal-directed in their online behavior than others. This point also highlights the importance of accounting for goal-directedness, as proposed in two online processing models (e.g., Hoffman and Novak 1996; Rodgers and Thorson 2001).

Evidence from this research suggests that individuals also differ in terms of their Internet motives as a whole. For example, students were more likely than non-students were to use the Internet for research motives, whereas non-students were more likely to shop and surf. This finding demonstrates the need to replicate research studies on a variety of dimensions, including the sampled population (Hunter 2001; Wells 2001).

Practical Implications

Advertising and marketing practitioners attempt to target the right ad to the right person at the right time. This is accomplished on the Internet through ad strategies such as key word searches and search engine optimization. In this way, practitioners can deliver ads that are at least somewhat relevant to the current interest of the consumer, as indicated by the search term. These findings suggest an alternative way to target consumers, by providing ads that offer feature-to-motive associations. Ads that complement the user's motive may have more success at being noticed and clicked on than ads that do not. Thus, one managerial implication is that advertising effectiveness can be improved to the extent that Internet ads are matched with the motives of the Internet user.

Caveats

As with any study, this one has its limitations. First, participants viewed the stimulus materials through forced exposure. Although forced exposure is an accepted means of conducting an ad effectiveness experiment, it is clearly not as realistic as embedding ads in the context of informational, entertainment, or editorial content. Thus, there is a need to replicate the current study by having individuals spend time online to see how motivation influences the stream of actual Internet behavior - the clicks, scrolls navigation, and movement from one Web site to the next.

In addition, the findings are limited because the banner ads were fictitious. This was done for control purposes. However, steps were taken to ensure that the ads were representative of actual banner ads, including manipulating real banner ads that were taken directly off the Internet. Thus, future studies will want to use real banner ads and, for that matter, different types of Internet ads as well (e.g., sponsorships, pop-ups, interstitials, etc.). Nevertheless, having 12 banner ads enhances generalization of the findings across a variety of banner ads. This at least overcomes some of the limitations of drawing conclusions from an N of 1 (see Wells 2001).

A potential limitation was the fact that motives were measured rather than manipulated. The study by Li and Bukovac (1999) highlights the difficulties involved with manipulating Internet motives (i.e., motive switching). Given this, and the fact that an Internet scale already existed, it was decided to measure rather than manipulate Internet motives. In addition, despite the fact that student and non-student samples were used, the individuals were not selected randomly. This, of course, limits the ability to generalize the findings to a random sample of the population. Nevertheless, this study at least attempted to replicate the study using a sample other than what we typically see in social scientific studies (i.e., college students).

Last, while having the benefit of controlling for individual differences, within-subjects designs risk a carry-over effect from one level of Factor A to another because participants are exposed to all levels of A (Lomax 2001). While there is always a chance that such a design will create demand characteristics, participants did not indicate that they knew the study's purpose when questioned during the debriefing. However, within-subjects designs also tend to produce significantly larger effect sizes for the ad attitude (Brown and Stayman 1992). Thus, it is possible that the design of the study inflated the responses to the dependent measures. Nevertheless, this limitation can be overcome by using between-subjects designs in follow-up studies.

In short, this research provides support for the notion that consumer motives influence ad processing. However, the exact mechanisms by which this process occurs are unclear. It may be that motives serve as a frame of reference, helping consumers to selectively process ads congruent with that motive. Based on the lack of consistent support for this idea, however, these findings should be considered preliminary until additional replications and extensions of the Interactive Advertising Model have been explored.

References

Bagozzi, R. P. (1997), "Goal-Directed Behaviors in Marketing: Cognitive and Emotional Perspective," Psychology and Marketing, 14 (6), 539-543.

Bagozzi, R. P. and U. Dholakia (1999), "Goal Setting and Goal Striving in Consumer Behavior," Journal of Marketing, 63 (Special Issue), 19-32.

Bargh, J. A. (1990), "Auto-Motives: Preconscious Determinants of Thought and Behavior," In Handbook of Motivation and Cognition: Foundations of Social Behavior, Vol. 2., E. T. Higgins and R. M. Sorrentino, eds., New York: Guilford, 93-130.

Bettman, J. R., M. F. Luce, and J. W. Payne (1998), "Constructive Consumer Choice Processes," Journal of Consumer Research, 25 (December), 187-217.

Briggs, R. and N. Hollis (1997), "Advertising on the web: Is there Response before Click-through?," Journal of Advertising Research, 37 (2), 33-45.

Brown, S. P. and D. M. Stayman (1992), "Antecedents and Consequences of Attitude toward the Ad: A Meta-Analysis," Journal of Consumer Research, 19 (June), 34-51.

Cannon, H., T. Richardson, and A. Yaprak (1998), "Toward a Framework for Evaluating Internet Advertising Affectiveness," Paper presented at the American Academy of Advertising, Lexington, KY.

Cho, C. H. (1998), "How Advertising works on the WWW: Modified Elaboration Likelihood Model," Paper presented at the American Academy of Advertising, Lexington, KY.

Clary, E. G., M. Snyder, R. D. Ridge, J. Copeland, A. A. Stukas, J. Haugen, and P. Miene (1998), "Understanding and Assessing the Motivations of Volunteers: A Functional Approach," Journal of Personality and Social Psychology, 74 (6),1516-1530.

Cooper, M. L., C. M. Shapiro, and A. M. Powers (1998), "Motivations for Sex and Risky Sexual Behavior among Adolescents and Young Adults: A Functional Perspective," Journal of Personality and Social Psychology, 75 (6),1528-1558.

Deci, E. L., and R. M. Ryan (1985), "The General Causality Orientations Scale: Self-Determination in Personality," Journal of Research in Personality, 19, 109-134.

Eighmey, J. (1997), "Profiling User Responses to Commercial Web Sites," Journal of Advertising Research, (May/June), 59-66.

Ghose, S. and W. Dou (1998), "Interactive Functions and their Impacts on the Appeal of Internet Presence Sites," Journal of Advertising Research, (March/April), 29-43.

Gutman, J. (1997), "Means-end Chains and Hierarchies," Psychology and Marketing, 14 (6), 545-560.

Gutman, J. (1982), "A Means-end Chain Model based on Consumer Categorization Processes," Journal of Marketing, 46 (2), 60-72.

Holbrook, M. B. and R. Batra (1987), "Assessing the Role of Emotions as Mediators of Consumer Responses to Advertising," Journal of Consumer Research, 14 (December), 404-420.

Hoffman, D. L. and T. P. Novak (1996), "Marketing in Hypermedia Computer-Mediated Environments: Conceptual Foundations," Journal of Marketing, 60 (July), 50-68.

Huffman, C. and M. J. Houston (1993), "Goal-oriented Experiences and the Development of Knowledge," Journal of Consumer Research, 20 (September), 190-207.

Hunger, J. E. (2001), "The Desperate Need for Replications," Journal of Consumer Research, 28 (June), 149-158.

Katz, J. E. and P. Aspden (1997), "Motivations for and Barriers to Internet Usage: Results of a National Public Opinion Survey," Internet Research: Electronic Networking Applications Policy, 7 (3), 170-188.

Kinnear, T. C. and J. R. Taylor (1991), Marketing Research: An Applied Approach, (4th ed.), New York: McGraw-Hill Book Company.

Korgaonkar, P. K. and L. D. Wolin (1999), "A Multivariate Analysis of Web Usage," Journal of Advertising Research, (March/April), 53-68.

Lawson, R. (1997), "Consumer Decision Making within a Goal-Driven Framework," Psychology and Marketing, 14 (5), 427-449.

Li, H. and J. L. Bukovac (1999), "Cognitive Impact of Banner Ad Characteristics: An Experimental Study," Journalism and Mass Communication Quarterly, 76 (2), 341-353.

Lomax, R. G. (2001), An Introduction to Statistical Concepts for Educational and Behavioral Sciences, Mahwah, NJ: Lawrence Erlbaum Associates.

Lutz, R. J. (1985), "Affective and Cognitive Antecedents of Attitude toward the Ad: A Conceptual Framework," In Psychological Processes and Advertising Effects, L. F. Alwitt and A. A. Mitchell, eds., Hillsdale, NJ: Lawrence Erlbaum Associates, 45-64.

Maignan, I. and B. A. Lukas (1997), "The Nature and Social Uses of the Internet: A Qualitative Investigation," Journal of Consumer Affairs, 31 (2), 346-371.

Martin, I. M. and D. W. Stewart (2001), "The Differential Impact of Goal Congruency on Attitudes, Intentions, and the Transfer of Brand Equity," Journal of Marketing Research, 38 (November), 471-484.

McClelland, D. C. (1987), Human Motivation, Cambridge, MA: Cambridge University Press.

Mitchell, A. A. and J. C. Olson (1981), "Are Product Attribute Beliefs the only Mediator of Advertising Effects on Brand Attitude?," Journal of Marketing Research, 18 (August), 318-332.

Pavlou, P. A. and D. W. Stewart (2000), "Measuring the Effects and Effectiveness of Interactive Advertising: A Research Agenda," Journal of Interactive Advertising, 1 (1), <http://jiad.org/vol1/no1/pavlou>.

Reeve, J. (1997), Understanding Motivation and Emotion, (2nd ed.), Orlando, Florida: Harcourt Brace and Company.

Rodgers, S. and H. M. Cannon (2000), "The Many Faces of Web Users: An Exploratory Study of Functionally-Based Web-Usage Groups," Paper presented at the American Academy of Advertising conference.

Rodgers, S. and K. M. Sheldon (2002), "The Web Motivation Inventory: An Improved way to Characterize Internet Users," Journal of Advertising Research (forthcoming).

Rodgers, S. and E. Thorson (2000), "The Interactive Advertising Model: How Users Perceive and Process Online Ads," Journal of Interactive Advertising, (1) 1 <http://jiad.org/vol1/no1/rodgers>.

Simon, H. A. (1955), "A Behavioral Model of Rational Choice," Quarterly Journal of Economics, 69 (February), 99-118.

Simon, H. A. (1990), "Invariants of Human Behavior," Annual Review of Psychology, 41, 1-19.

Stafford, T. F. and M. R. Stafford (1998), "Uses and Gratifications of the World Wide Web: A Preliminary Study," In The Proceedings of the 1998 Conference of the American Academy of Advertising, D. D. Muehling, ed., Washington State University, Pullman, Washington, 174-181.

Wells, W. D. (2001), "The Perils of N = 1," Journal of Consumer Research, 28 (December), 494-498.

Wimmer, R. D. and J. R. Dominick (1997), Mass Media Research: An Introduction, Belmont, CA: Wadsworth Publishing Company.

Appendix A Stimulus Materials

Surfing Banners

Research/Information Acquisition Banners

Communicate/Socialize Banners

Shopping Banners

About the Authors

Shelly Rodgers is Assistant Professor in School of Journalism & Mass Communication at the University of Minnesota-Twin Cities.