SPOKES-CHARACTERS Creating Character Trust and Positive Brand Attitudes Judith A. Garretson and Ronald W. Niedrich ABSTRACT: Spokes-characters have appeared in numerous promotion campaigns throughout the past century. While interest in these characters has evolved over time, a substantial portion of the research concerning characters has emerged during this past decade. Yet there exists a research gap concerning the impact of specific spokes-character features on brand-related outcomes. In this study, we address this gap and extend this stream of literature with the first empirical examination of the relationship between specific spokes-character features and brand attitude. Based on the tested moderated mediation model, our results show that spokes-character trust is an important mediator of the effects of spokes-character features on brand attitude. The model also indicates that brand experience significantly moderates the effect of spokescharacter trust on brand attitude. Thus, spokes-character features result in more favorable brand attitudes for consumers with less brand experience, whereas spokes-character features have little effect on brand attitudes for consumers with more experience with the brand.
The use of spokes-characters has received increased media attention, with hour-long television programs devoted to addressing their potential benefits to brands. This attention is not surprising, given their prevalence both in historical and modern promotion campaigns. Consistent with prior research (Callcott and Alvey 1991; Callcott and Lee 1995; Phillips 1996c), we consider spokes-characters to be nonhuman characters used to promote a product or a brand. Here, we examine a common type of spokes-characters, noncelebrity characters. These characters aren’t cartoons originally created for animated movies, cartoon programs, and/or comic strips and then licensed by brands to appear in promotions. Rather, they are created for the sole purpose of promoting a product or brand. Since the late 1800s, hundreds of these noncelebrity characters have promoted brands, including the Snuggle Bear, the Poppin’ Fresh Pillsbury Doughboy, and Tony the Tiger. As suggested in television programs, the popular press, and academic research, these characters appear to benefit brands by establishing brand identity and favorable brand associations (Dotz, Morton, and Lund 1996; Fournier 1998; Thompson 2002). Several studies in the spokes-character literature have documented the proliferation of these characters in promotion campaigns and their symbolic role in promoting products (e.g.,
Judith A. Garretson (Ph.D., University of Arkansas) is an assistant professor in the Department of Marketing, E. J. Ourso College of Business Administration, Louisiana State University. Ronald W. Niedrich (Ph.D., University of South Carolina) is an assistant professor in the Department of Marketing, E. J. Ourso College of Business Administration, Louisiana State University.
Callcott and Lee 1994; Neeley et al. 2000; Phillips 1996a) or identified the potential public policy implications of their use (e.g., Garretson and Burton 1998; Mizerski 1995). Outside the policy stream of research, work demonstrates how practitioners use characters in promotion campaigns. The results of content analyses of spokes-character commercials and print advertisements reveal that (1) spokes-characters are used to promote numerous types of products and services, (2) specific types of spokes-characters are more commonly featured with particular products, and (3) practitioners appear to consider the nostalgic qualities and relevancy of characters to advertised products (Callcott and Lee 1994; Neeley et al. 2000; Spears, Mowen, and Chakraborty 1996). Furthermore, analyses of depth-interviews indicate that consumers notice character factors such as nostalgia, relevance to products, and expertise (Callcott and Phillips 1996). In the policy domain, research has addressed whether spokes-characters promote favorable attitudes toward adult products and their use among vulnerable populations (Garretson and Burton 1998; Henke 1995; Kelly et al. 2000). A few of these policy-related studies point toward a link between the use of characters and improved attitudes. Although there appears to be a positive relationship between the use of spokes-characters and favorable brand attitudes, few studies have begun to specifically and empirically identify the types of character qualities responsible for the relationship. That is, a looming question concerns the degree to which qualities such as expertise, relevancy, and nostalgia identified in relevant content analyses and depth-interviews affect brand perceptions. In this paper, we empirically address this question. In doing so, we argue that these factors affect brand attitudes, but we contend that spokes-character Journal of Advertising, vol. 33, no. 2 (Summer 2004), pp. 25–36. © 2004 American Academy of Advertising. All rights reserved. ISSN 0091-3367 / 2004 $9.50 + 0.00.
26
The Journal of Advertising
FIGURE 1 Conceptual Model Character Expertise
Character
Character
Brand
Relevance
Trust
Attitude
Character Nostalgia
trust is an important mediator of this relationship. Furthermore, we also assert that brand experience represents an important moderator to consider, as campaigns often target user groups with different levels of brand experience and research demonstrates that observable features are more influential to those less experienced with the brand (Garbarino and Johnson 1999). Complementing current spokes-character work, this empirical study demonstrates the importance of considering character qualities when developing and managing spokescharacters for the promotion of brands. BACKGROUND AND CONCEPTUALIZATION For more than a century, marketers and advertisers have utilized spokes-characters in promotional campaigns and on product packages (Callcott 1993). Some of the earliest characters include the Michelin Man and the Cream of Wheat Chef, both of which were created in the late nineteenth century and continue to appear in current campaigns. Exploratory research indicates that consumers like spokes-characters and have even expressed their trust and respect for them (Callcott and Phillips 1996). Consumers seem to evaluate characters and their qualities just as they do people. Given our knowledge on animism (Gilmore 1919) and the advances in artwork and animation, this perception that spokes-characters are like people and have souls is not surprising. Nor should it be unforeseen that consumers infer human-like characteristics, such as spokesperson trust, from observable qualities of the characters. These characteristics can have significant implications. For example, in their book in which they document the use of spokescharacters, Dotz, Morton, and Lund (1996) suggest characters can create perceptions of trust, which appears to affect consumers’ attitudes and behaviors. However, to our knowledge, such relationships have not been addressed empirically.
Given the recent suggestions regarding spokes-characters and the importance of trust, logical questions emerge. What are the factors that affect spokes-person trust, and does that trust favorably affect brand attitude? As shown by our conceptual model in Figure 1, we attempt to address these relatively unexplored questions. We predict that spokes-person expertise, the relevancy of the character to the promoted product, and nostalgia elicited by characters will affect spokescharacter trust. We also expect spokes-character trust to mediate the impact of those factors on brand attitude. Furthermore, brand experience is hypothesized to moderate the relationships in the model. Concerning this moderating factor, nostalgia should have a more pronounced effect on spokescharacter trust for those consumers with higher levels of brand experience than for those with less experience. In addition, both the relevancy of the spokes-character to the promoted product and spokes-character expertise should be more influential for those with less brand experience. Factors That Influence Spokes-Character Trust A significant portion of research concerning the use of spokescharacters has explored the attitude-related consequences of featuring characters with brands rather than focusing on the antecedents for such effects. These studies offer evidence as to the public policy implications of using spokes-characters (Fox et al. 1998; Garretson and Burton 1998; Henke 1995; Hoy, Young, and Mowen 1986; Kelly et al. 2000; Mizerski 1995). The focus of many of these policy-related studies was whether spokes-characters represent cues to lure younger consumers to adult products. Other related work has public policy implications, but is more theoretical in nature. For example, Macklin’s work (1994, 1996) focuses on how visual cues, like spokes-characters, improve memory for brand-related information and brand attitudes among children. In essence, much of the work in this area contends that the placement of characters in product advertisements offers benefits to firms targeting children and teenagers. Additional work specifically concerning spokes-characters has been more conceptual and theoretical. These nonpolicy advances represent some of the earliest conceptualizations and classifications of these unique brand identifiers (i.e., Callcott and Alvey 1991; Callcott and Lee 1995; Phillips 1996b). These exploratory studies lend insight as to the plausible factors consumers use to evaluate characters and practitioners take into account when creating favorable spokes-characters and brand impressions. Looking across the results of various content analyses, one can see a consistent pattern of spokescharacter and product category pairings used in promotions (Callcott and Lee 1994; Spears, Mowen, and Chakraborty 1996). For example, animal spokes-characters are often paired with services and beverages. This consistent pattern implies
Summer 2004 27
that some types of spokes-characters might be better suited for and benefit specific product categories. The above observation also suggests a more practical rationale behind marketers’ pairings of particular characters with products that might affect consumers’ receptiveness or trust of spokes-characters. Perhaps marketers and advertising agencies contemplate the relevancy or “matchup” of the spokescharacter with the product at the initial inception of character development. In fact, the agency responsible for the development of the Pillsbury Doughboy created the character specifically to personify the benefits of quality and freshness. It wanted the spokes-character to pop out like the product— like dough from a can (“General Mills History of Innovation: The Pillsbury Doughboy” 2001). Initial explorations concerning this congruency suggest favorable consequences for spokes-character trust. In their study, where in-depth interviews were used to elicit consumers’ responses to spokescharacter advertising, Callcott and Phillips (1996) reported that respondents considered the character and product match to be important. Although not empirically examined, the match appeared to frame their evaluation of both the character and the brand. Drawing from the congruence literatures in consumer psychology and advertising (Goodstein 1993; Heckler and Childers 1992; Kamins 1990; Sengupta, Goodstein, and Boninger 1997), relevance is considered to be the degree to which two stimuli match or fit together. In the context of spokes-characters, it is the degree to which the spokescharacter is perceived to be appropriate for the product with which it is associated in promotion campaigns. For example, it is likely that consumers can easily identify the relevance between the Michelin Man, whose body consists of layers of tires, and Michelin tires, the product promoted by the spokes-character. The benefits of relevant product information are widely cited in promotion-related work and include favorable object evaluations resulting from well-matched stimuli (Lee and Mason 1999; Miniard et al. 1991; Sengupta, Goodstein, and Boninger 1997). We predict a similar pattern; the evaluation of spokes-character trust will be influenced by the perceived degree of fit between characters and products. The hypothesis for this relationship is presented below: H1: Relevance between spokes-characters and products will favorably affect spokes-character trust. As with human spokespeople, these characters appear to endorse brands (Callcott 1993; Callcott and Phillips 1996). With the use of creative artwork and animation, agencies can present them as walking, talking, humanlike creatures touting brand benefits. Spokes-characters are often presented in an ownership role for the brand. The Web site for Mr. Clean encourages consumers to “check out his line of
gloves . . . his line of mops” (www.homemadesimple.com/ mrclean). In addition, Toucan Sam, the spokes-character for Kellogg’s Fruit Loops is considered the “expert” on Kellogg’s Fruit Loops cereal and constantly “follows his nose to find” the cereal according to Kellogg’s Web site (www.kelloggcompany .com/kelloggco/our_company/marketing_characters/ index.html). As such, it is not surprising that, in accordance with prior conceptualizations (Callcott and Alvey 1991; Callcott and Lee 1994; McGinnies and Ward 1980), consumers can perceive spokes-characters as experts. Expertise is the perception that the source is able to make valid claims or has knowledge of the product (Hovland, Janis, and Kelley 1953; McGinnies and Ward 1980; Ohanian 1991). As evidenced by numerous campaigns, spokescharacters make product claims and do so repeatedly from campaign to campaign. The importance of this perceived expertise lies in the fact that it can affect consumers’ perceptions of spokes-character trust. As noted in various literatures concerning trust of individuals, this perceived expertise plays a role in the development of trust (Crosby, Evans, and Cowles 1990; Doney and Cannon 1997; Moorman, Deshpande, and Zaltman 1993). Consumers assess the trustworthiness of an individual based on that individual’s personal abilities, consistent with attribution theory in inference making (Kelley 1972). In their work concerning salespeople, Doney and Cannon (1997) demonstrated that expertise is an individual factor that affects the cognitive trust–building process, directly linking expertise with trust. In concert with the above rationale, we expect this link to be present in the spokes-character context. When consumers perceive spokes-characters as exhibiting product expertise, trust should be favorably affected. Hence, the following hypothesis is posited: H2: Spokes-character expertise will favorably affect spokescharacter trust. Evidence suggests that organizations realize the power of nostalgia and its favorable impact on consumers and brand relationships (Braun, Ellis, and Loftus 2002; Poniewozik 2002; Thompson 1999a). Closely related to previous work concerning nostalgia (Callcott and Phillips 1996; Holbrook and Schindler 1994), we consider nostalgia to be the recollection of personal or episodic events. It has long been known that television programs play on nostalgia, attempting to take people back to earlier and simpler times with shows like the Wonder Years and the new television drama American Dreams (Poniewozik 2002). But brands have utilized this captivating appeal in promotion campaigns as well. Several organizations have reintroduced slogans and characters from the past ( Johnson 2002; Thompson 1999b), while others merely remind consumers that brands have been part of their lives for years. Interestingly, research cites consumers’ interest in nos-
28
The Journal of Advertising
talgic campaigns (Langer 1997). They long for connections to their childhood. Prior exploratory work suggests that spokes-characters are nostalgic representations (Callcott and Alvey 1991; Callcott and Phillips 1996). Consumers are exposed to the characters at an early age, since several characters endorse children’s products (Callcott and Lee 1994; Mizerski 1995). These spokes-characters survive generations and can maintain a constant presence in consumers’ lives, as they appear in advertisements, on product packages, and in consumer sales promotions. A recent content analysis posits that current campaigns actually attempt to elicit this nostalgia with the use of characters (Neeley et al. 2000). They are used to remind consumers of the brand that they might have used and trusted from childhood. They prime personal memories (Callcott and Alvey 1991; Neeley et al. 2000), including those of earlier times and felt trust with the character. Consistent with this prior work, we expect that the nostalgia will favorably affect spokes-character trust, and thus offer the following hypothesis: H3: Nostalgia will favorably affect spokes-character trust. Spokes-Character Trust and Brand Attitudes As noted earlier, spokes-character trust is important given the contention that consumers tend to like and purchase products from trustworthy characters (Dotz, Morton, and Lund 1996). Similar to prior research, we consider trust to be the perceived reliability of the target of trust (Garbarino and Johnson 1999; Lindskold 1978; Morgan and Hunt 1994). Trust is the expectation that the target object represents integrity, sincerity, and honesty (Crosby, Evens, and Cowles 1990). The relationship literature has investigated the role of trust, consistently citing it as a strong predictor of favorable brand-related outcomes (Garbarino and Johnson 1999; Macintosh and Lockshin 1997). As noted in this stream of research, established trust can improve one’s attitude toward a brand. Furthermore, trust is a key variable that mediates successful relationships (Atuahene-Gima and Li 2002; Garbarino and Johnson 1999; Morgan and Hunt 1994). We argue that the relationships consumers feel they have with spokes-characters may share some of the same characteristics as those conceptualized in the relationship literature. Thus, in the context of spokes-characters, we predict similar outcomes of trust. Rather than focusing on the trust of individuals, our work examines spokes-character trust. It refers to the expectation that characters will be honest, sincere, and reliable in their communication and promotion of products. As shown in Hypotheses 4 and 5, we expect spokes-character trust to favorably affect brand attitudes and mediate the effects of spokes-character qualities on brand attitudes.
H4: Spokes-character trust will favorably affect attitude toward the brand. H5: The effects of spokes-character relevance, nostalgia, and expertise on attitude toward the brand will be mediated by spokes-character trust. Brand Experience as a Moderator Brand experience can be defined as the direct observation or use of a product and represents one process by which brand knowledge is acquired (Folkes and Patrick 2003; Wood and Lynch 2002). Over time, consumers who often use a particular brand develop vivid brand associations and evaluations that are later stored in memory. Because the use of the brand is personal, the experiences with the brand are vivid and more memorable. They are linked and can be accessed with brand associations and evaluations with which they are stored (Kempf and Smith 1998). Typically, the experienced user does not modify the brand-related evaluations once they are created and stored in memory (Forgas 1995). Consistent with work in social psychology and memory, consumers who have experience with a given brand are more likely to refer to their existing evaluations rather than construct new evaluations when asked to assess that brand, as long as no contradictory information about the brand is presented (see Forgas 1995 for a review). We expect that brand experience will moderate the impact of spokes-character features on spokes-character trust. More specifically, we predict the effect of character nostalgia on spokes-character trust will be more pronounced for those individuals with more brand experience than for those with less experience. More experienced users should have a greater collection of memories associated with the brand, including the spokes-character brand association. When thinking of the spokes-character, those more experienced with the brand should recall those character-related memories and use them to form spokes-character trust. Like those who have less experience with the brand, those with more experience will consider the relevancy of the character to the promoted product and its expertise. However, those with less experience should not have nostalgic and brand-related memories. Rather, consistent with the categorization literature (e.g., Fiske and Pavelchak 1986), consumers with less experience can’t rely on these memories and will likely use the more observable features to construct spokes-character trust. Hence, they will give more consideration to the relevancy of the character to the product and its expertise. As such, the effects of spokescharacter relevance and expertise on spokes-character trust should be more pronounced for less experienced users. We therefore offer the following hypotheses:
Summer 2004 29
H6: Brand experience will moderate the effects of spokescharacter qualities on spokes-character trust. More specifically, the effect of nostalgia on spokes-character trust will be significantly more positive for consumers with more brand experience, and the effect of relevancy and expertise on spokescharacter trust will be significantly more positive for consumers with less brand experience. Finally, we expect that more experienced users of the brand have more clearly defined attitudes toward the brand than less frequent users of the brand. Consistent with the schema research (Peracchio and Tybout 1996; Sujan and Bettman 1989), attitude toward the brand for more experienced users is more likely to be based on personal experience and more detailed knowledge of several salient brand attributes. On the other hand, less experienced brand users are less likely to have well-formed knowledge structures to formulate brand attitudes. If attitudes exist, they are likely to be weakly held and not easily accessible. For example, whereas experts are less likely to conform to advertisers’ claims, novices find it more difficult to challenge these claims (Alba and Hutchinson 1987). As a result, less experienced brand users are more likely than those with more experience to use spokes-character trust to formulate brand attitudes. Thus: H7: Brand experience will moderate the effects of spokescharacter trust on brand attitude. More specifically, for consumers who are less experienced with the brand, the effect of spokes-character trust on brand attitude will be significantly more positive than for consumers who are more experienced with the brand. METHOD Data Collection One hundred and forty-four undergraduate students took part in this study. As students are commonly exposed to the promotional use of spokes-characters, they are an appropriate population for study in this case. Participants were provided with a cover sheet that explained the procedures in the study. They were informed that researchers were interested in some of their thoughts and opinions about spokes-characters used in promotions. A brief description of noncelebrity spokescharacters were first shared with participants, although no specific characters were identified. After being given the description, respondents were instructed to recall a spokescharacter they had seen recently in a television commercial, magazine advertisement, or in the grocery store. They were then asked to write down the name of that spokes-character and the name of the brand that used the character in its promotion campaigns. Consistent with prior research (Richins
1994a, 1994b), the respondents were then instructed to refer to that specific spokes-character and brand when answering the questions in the study. As with our conceptualization of noncelebrity spokescharacters, participants were instructed to identify only those characters that were created specifically for a promotion campaign and not licensed cartoon characters (e.g., Snoopy for MetLife and Homer Simpson for Butterfinger). With the exception of four surveys, all spokes-characters qualified for the study. As a result, there were 140 usable surveys where a noncelebrity spokes-character was provided and the brand it endorsed was correctly identified. Respondents identified 37 different types of spokes-characters, showing a wide variance of characters recalled and analyzed in this study (see Table 1). Participants’ ages ranged from 20 to 28, and approximately 62% were female. Pretest Given that no established scale was available to measure nostalgia, a pretest was conducted to develop such a scale. The final scale included four, seven-point Likert-type items anchored by strongly disagree and strongly agree. The four nostalgia items were (1) “This character icon makes me think back to events that occurred when I was younger,” (2) “This character icon reminds me of experiences from the past,” (3) “When I see this character icon, it makes me remember different moments in my life,” and (4) “This character makes me think of memories.”1 Results of an exploratory factor analysis reveal a one-factor solution, with factor loadings ranging from .91 to .95. Measures All items in this study were perceptual measures, and when available, drawn from existing and tested scales. Consistent with Miniard et al. (1991) and Sengupta, Goodstein, and Boninger (1997), spokes-character/product relevancy was measured with four, seven-point items anchored by very appropriate and very inappropriate. The four relevancy items were (1) “It makes sense for this character icon to be featured with this product,” (2) “I think that pairing this character icon with this product is appropriate,” (3) “I think that the character icon is relevant for this product,” and (4) “Together, this character icon and this product represent a very good fit.” Coefficient α for the character relevancy scale was .94. Based on Ohanian (1991), spokes-character expertise was measured with five, seven-point items anchored by not an expert/expert, inexperienced/experienced, unknowledgeable/knowledgeable, unqualified/qualified, and unskilled/skilled. Coefficient α for the character expertise scale was .96.
30
The Journal of Advertising TABLE 1 Most Frequently Identified Spokes-Characters
Spokes-character Poppin’ Fresh Pillsbury Doughboy Tony the Tiger Energizer Bunny M&M’s Characters Snuggle Bear Keebler Elves Mr. Mini-Wheat The Geico Gecko Toucan Sam
Endorsed brand
Respondents
Pillsbury Products Kellogg’s Frosted Flakes Cereal Energizer Batteries M&M’s Chocolate Candy Snuggle Fabric Softener Keebler Snacks Kellogg’s Frosted Mini-Wheats Cereal Geico Insurance Fruit Loops Cereal
30 26 17 9 7 6 5 4 4
Notes: Spokes-characters identified by three respondents or less include: Lucky Charms Leprechaun; Cingular X-Man; Aflac Duck; Joe Camel; Aunt Jemima; Kool-Aid Man; Trix Bunny; Mr. Clean; Chester Cheetah; Snap, Crackle, and Pop; Captain Crunch; Sony’s Blue Alien; Honeycomb Cereal Kid; Cheerios Bee; Jack ’n’ the Box; Charlie Tuna; Buddy Lee; Morton Salt Girl; Elsie the Cow; Chevron Car; Michelin Man; Mrs. Butterworth; Hamburger Helper Glove; and Scrubbing Bubbles.
The scale developed in the pretest was used to measure spokes-character nostalgia. Coefficient α for the nostalgia scale was .95. Based on Ohanian (1991), spokes-character trust was measured with five, seven-point items anchored by undependable/dependable, dishonest/honest, unreliable/reliable, insincere/ sincere, and untrustworthy/trustworthy. Coefficient α for the character trust scale was .91. Consistent with Garretson and Burton (1998), Goodstein (1993), and Miniard et al. (1991), attitude toward the brand was measured using four, seven-point items anchored by like/dislike, favorable/unfavorable, good/bad, and positive/negative. Coefficient α for the brand attitude scale was .97. Finally, to measure brand experience, respondents were asked, “How often do you use this brand?” Three, sevenpoint items of never chosen/always chosen, unpreferred/preferred, and rarely used/always used were used to assess this measure (Lane 2000). Coefficient α for the scale was .92. RESULTS The Results section is organized as follows. First, using confirmatory factor analysis, we assess the discriminant and convergent validity of the measured constructs. Next, using path analysis, we test the model for the moderating effect of brand experience and the mediating effect of spokes-character trust. Finally, after the final path model has been established, we estimate the model parameters and test the remaining hypotheses for direct and indirect effects. Measurement Model Using confirmatory factor analysis, a 25-item, fully correlated six-factor (i.e., all six factors are intercorrelated) measurement model was fit to the data to test the convergent and discriminant validity of the constructs. Based on a number of indices,
the model fit was adequate ( χ2 = 471.4, df = 260; p value = 0; GFI [goodness-of-fit index] = .803; TLI [Tucker-Lewis index] = .938; CFI [comparative fit index] = .947; RMSEA [root mean square error of approximation] = .075). As shown in Table 2, the variance-extracted estimates are greater than the squared correlations between all constructs. Thus, based on the test recommended by Fornell and Larcker (1981), all constructs displayed evidence of discriminant validity. Furthermore, all factor loadings were significant ( p < .001), thus providing evidence of convergent validity (Anderson and Gerbing 1988). Tests of Moderation and Mediation To test for the moderating effect of brand experience, we employed a multigroup approach, which requires the use of the covariance matrix rather than the correlation matrix (Maruyama 1998, p. 258). The multigroup structural equation model did not converge, however, and we therefore elected to use multigroup path analysis to test the hypotheses.2 As previously noted, the measurement model had acceptable psychometric properties and all construct reliabilities are above .9. The major impact of using path analysis is a loss of statistical power because path analysis does not account for measurement error. Thus, the tests using path analysis are more conservative than the use of structural equation modeling. The statistical procedures require that tests of moderation first be conducted, resulting in first analyzing predictions made in H6 and H7. We tested the model for the moderating effect of brand experience using a median split of the brand experience measure. Respondents were classified as either more experienced users of the brand (M = 6.146, SD = 1.35) or less experienced users of the brand (M = 3.97, SD = 1.35). For comparison, we define the baseline model as the fully unconstrained model where all paths are free to vary across both
Summer 2004 31 TABLE 2 Construct Reliability, Variance Extracted, and Correlations Coefficient α
Variance extracted
Character expertise
.957
.758
Character relevance
.938
.849
Character nostalgia
.949
.824
Character trust
.907
.672
Brand attitude
.971
.902
Brand experience
.919
.827
Constructs
Character expertise
Character relevance
.501 (7.614) .312 (3.926) .653 (12.174) .435 (6.089) .435 (4.507)
.293 (3.691) .296 (3.664) .273 (3.441) .333 (4.325)
Character nostalgia
Character trust
Brand attitude
.561 (9.085) .391 (5.162)
.478 (7.115)
.290 (3.546) .222 (2.695) .197 (2.361)
Notes: Correlation t values are provided in parentheses. All correlations are significant at p < .01. In all cases, the variance-extracted estimates are greater than the squared correlation.
FIGURE 2 Partially Constrained Multigroup Path Model Character Expertise .371*
Character
M (−.134)
Character
Relevance
L (.076)
Trust
1
M (.181*) L (.642*)
Brand Attitude
1
.080*
Character
e1
e2
Nostalgia
experience. Although the path from relevance to trust was moderated, the paths from expertise and nostalgia to trust were not moderated by brand experience. Thus, H6 was partially supported. H7 was supported, as the path from trust to brand attitude is moderated by brand experience. Path analysis is particularly well suited for testing mediation (Wegener and Fabrigar 2000). Tests of mediation were conducted by independently adding a direct path to brand attitude from each exogenous variable. Again, we use the fully unconstrained model as the baseline for model comparisons. As shown in Table 3, none of the direct paths were significant based on χ2 difference tests. Thus, any significant effects of the exogenous variables on brand attitude are fully mediated by spokes-character trust, thus providing support for H5. Tests of Direct and Indirect Effects
Notes: Unstandardized parameter estimates are provided for both more (M) and less (L) experienced users. A single parameter estimate shown for a path indicates no significant differences between the two groups. Two parameter estimates shown for a path indicates significant differences between the two groups at p < .10. Asterisks indicate parameters that are significantly different from zero at p < .10.
groups. As shown in Table 3, the unconstrained model fit the data very well (χ2 = 1.730, df = 6; p value = .943; GFI = .995; CFI = 1.000; RMSEA = 0). Consistent with the procedures outlined by Jöreskog and Sörbom (1988), tests of moderation were conducted by independently constraining each of the seven direct and correlated paths in the model. As shown in Table 3 and in Figure 2, two of the predicted parameter estimates are significantly moderated by brand experience based on χ2 difference tests.3 Hypothesis 6 predicted that the paths from the three exogenous variables to trust were moderated by brand
Based on the previous results, we define the partially constrained model where the four parameter estimates not significantly moderated by brand experience were constrained while allowing the remaining paths to freely vary across the two groups. As shown in Table 3, the partially constrained model fit the data very well (χ 2 = 3.712, df = 10; p value = .959; GFI = .989; CFI = 1.000; RMSEA = 0). The unconstrained and partially constrained models are not statistically different from each other, ∆χ 2(4) = 1.982; p value > .5. However, based on our theoretical expectation of a moderated model and model parsimony, we conclude in favor of the partially constrained model. Thus, H1 through H4 are based on the direct effects of the partially constrained model shown in Figure 2. For simplicity, the unstandardized parameter estimates are provided in Figure 2. Table 4 provides the standardized parameter estimates and t values for the partially constrained model. Although H1 was not supported, H2, H3,
32
The Journal of Advertising TABLE 3 Tests of Moderation and Mediation Model χ 2 (df )
Model
Model p value
χ 2 (df ) Difference
Unconstrained
1.730 (6)
.943
n.a.
Partially constrained
3.712 (10)
.959
1.982 (4)
1.818 (7) 5.275 (7) 2.827 (7) 25.29 (7)
.969 .626 .901 .001
.088 (1) 3.545a (1) 1.097 (1) 23.56b (1)
1.294 (4) .329 (4) 1.696 (4)
.862 .988 .792
.436 (2) 1.401 (2) .034 (2)
Moderation tests Expertise → Trust Relevance → Trust Nostalgia → Trust Trust → Attitude Mediation tests Expertise → Attitude Relevance → Attitude Nostalgia → Attitude
Conclusions
More parsimonious Does not support H6 Supports H6 Does not support H6 Supports H7 Supports H5 Supports H5 Supports H5
Notes: Difference at p < .10 and p < .001, respectively. All tests compare the rival model to the unconstrained model. Tests of moderation are conducted by independently constraining each path in the model. Tests of mediation are conducted by independently adding a direct path to attitude for each exogenous variable. n.a. = not available a and b indicate a significant χ2
and H4 were supported based on the t values of the parameter estimates reported in Table 4. In the previous analysis, the direct effects were moderated by brand experience. To provide further insight into these effects, we test for moderated mediation using the procedures outlined by Wegener and Fabrigar (2000, p. 437) where the mediation analysis is conducted at each level of the moderator (more and less experienced users). As previously established, any significant effects of the exogenous variables on brand attitude are fully mediated by spokes-character trust. Thus, moderated mediation can be evaluated by comparison of the indirect effects across groups. As seen in Table 4, the exogenous variables have no significant effect on brand attitude for the more experienced users. However, spokescharacter expertise and nostalgia have significant indirect effects on brand attitude for less experienced users. GENERAL DISCUSSION The results of this current study yield insight into the factors accounting for the favorable impact of spokes-characters on brand attitudes. In sum, five of our seven hypotheses were supported, with H6 partially supported and H1 not supported. Our findings complement prior research illustrating the prolific use of characters in promotion campaigns. As shown by various content analyses and other qualitative work (Callcott and Alvey 1991; Callcott and Phillips 1996; Neeley et al. 2000), these characters are commonly featured in promotion campaigns and appear to embody various humanlike qualities. We extend the current literature by demonstrating empirically the importance of specific spokes-character qualities and trust. Understanding how such character qualities affect spokes-character trust and brand attitudes across various mar-
ket segments, in general, should provide greater insight as to the influential role of characters in promotion campaigns. To the best of our knowledge, this study represents one of the first studies to empirically examine the relationship between specific spokes-character features and brand attitude. Contributions and Implications The current research offers three distinct contributions. First, this study demonstrates that spokes-character trust can be influenced by spokes-character characteristics. Prior exploratory research and the popular press have suggested that spokes-characters embody qualities that ought to signal trust (Callcott and Phillips 1996; Dotz, Morton, and Lund 1996). Our findings offer empirical evidence for these propositions. We found perceived expertise and nostalgia to engender spokes-character trust. Consumers appear to rely on spokescharacter features as signals of this honesty and sincerity, as greater levels of perceived character expertise and nostalgic memories produced greater levels of trust. Regardless of consumers’ brand experience, these character-specific features influence consumers’ overall impressions of spokes-character trustworthiness. Interestingly, though, the relevancy of the spokes-character to the advertised product did not directly affect spokescharacter trust. Given the theoretical rationale and empirical support for relevancy in prior promotion-related research (Gwinner and Eaton 1999), this finding is rather unexpected, especially relative to spokesperson research where human endorsers are viewed more favorably when they endorse products more relevant to their image (Sengupta, Goodstein, and Boninger 1997). Content analyses concerning nonhuman spokes-characters (Callcott and Lee 1994) suggest that simi-
Summer 2004 33 TABLE 4 Standardized Parameter Estimates for Partially Constrained Model Parameter More brand experience Direct effects Expertise → Trust Relevance → Trust Nostalgia → Trust Trust → Attitude Indirect effects Expertise → Trust → Attitude Relevance → Trust → Attitude Nostalgia → Trust → Attitude Less brand experience Direct effects Expertise → Trust Relevance → Trust Nostalgia → Trust Trust → Attitude Indirect effects Expertise → Trust → Attitude Relevance → Trust → Attitude Nostalgia → Trust → Attitude
Estimates (t values)
Conclusions
.540 (7.046)c –.134 (–1.357) .169 (2.159)b .181 (1.566)a
Strongly supports H2 Does not support H1 Supports H3 Marginally supports H4
.098 (1.118) –.024 (–.713) .031 (1.000)
Not significant Not significant Not significant
.519 (7.046)c .093 (.924) .134 (2.159)b .676 (7.725)c
Strongly supports H2 Does not support H1 Supports H3 Strongly supports H4
.351 (4.760)c .063 (.891) .091 (2.040)b
Significant Not significant Significant
Notes: Boldface estimates indicate parameters that were constrained across the two groups. Between-group differences in the standardized estimates for constrained parameters reflect differences in the factor variances across groups. All hypotheses predicted positive values, and thus one-tailed tests are employed. a, b, and c indicate parameters that are significantly different from zero at p < .10, p < .05, and p < .001, respectively.
lar patterns should emerge. These data do not appear to support that pattern, however. Relevancy appears to be an issue of lesser concern for brands using spokes-characters, but it appears to be important for human endorsers, thus adding another distinguishing factor between human and nonhuman endorsers addressed in the endorser literature (Stafford, Stafford, and Day 2002). It should also be noted that although relevancy was not significant with existing, memorable characters, it might still play a role in establishing trust for new spokes-characters introduced by brands. This study also provides evidence that spokes-character trust mediates the relationship between spokes-character qualities on brand attitude. This mediated relationship represents the second contribution of this research. The implications of spokes-character trust are significant. Trust appears to be a vital, yet controllable factor. This controllability might favor their use over other types of spokespeople, as recently suggested in the literature (Stafford, Stafford, and Day 2002). Brands, agencies, and animators can create and control the trust elicited by specific types of character qualities, including perceived expertise, to foster spokes-character trust and, ultimately, to affect brand attitudes. They can also reintroduce the nostalgic qualities of characters to adults, a strategy recently recognized in current campaigns (Neeley et al. 2000). In sum, our findings show that efforts toward creating a trust-
worthy spokes-character can be rewarded with more favorable brand evaluations. The third contribution of this research concerns the moderated relationship between spokes-character trust and brand attitude. As predicted in H7, spokes-character trust favorably influenced brand attitudes, but this relationship was significant for only those less experienced with the brand. It is important to note that, while less significant than that for less experienced users, there is a link between character trust and attitudes for those more experienced with the brand. Regardless, this qualified trust–attitude relationship indicates that improving spokes-character trust poses more pronounced brand-related outcomes for those less experienced with the brand. Given the theoretical work concerning schemas (Peracchio and Tybout 1996; Sujan and Bettman 1989), this finding is not surprising. Those with less brand experience appear to rely more heavily on observable character qualities when forming brand attitudes compared to those with more brand experience. Focusing on the trustworthiness of the character might represent a viable strategy to create and foster positive brand equity for those in this market segment. In contrast to predictions, brand experience did not moderate two of the relationships leading to spokes-character trust. For those with more brand experience, a more pronounced effect of character nostalgia on spokes-character trust was pre-
34
The Journal of Advertising
dicted. In contrast, we expected a more pronounced effect of relevancy and expertise on trust for those with less brand experience. Overall, our results indicate those with varying experience levels consider expertise and nostalgia when forming spokes-character impressions, as both were predictors of spokes-character trust. Perhaps when forming the trustworthiness of the spokes-character, those memories accessed by experienced users were no more diagnostic than those accessed by those less experienced with the brand. The findings support contentions in the literature that experience does not necessarily ensure consumers are apt at assessing diagnostic memories (Hoch 2002). Those with experience might also benefit from observable qualities such as expertise prior to forming evaluations. In this study, those with both less and more experience found it useful to refer to memories as well as observable characteristics when assessing a specific brand association, spokes-character trust. Recommendations for Future Research Continued efforts focusing on additional features not examined here could more fully identify the set of factors that signal spokes-character trust and affect brand attitudes. Related to this idea, research investigating how spokes-characters might foster consumer–brand relationships appears warranted. The notion that characters can legitimize brands as partners has been suggested, yet untested (Fournier 1998). As organizations attempt to differentiate brands in a competitive marketplace, the use of spokes-characters that cultivate friendships between consumers and brands might prove beneficial in establishing commitment to the brand. Future research might also examine the impact of both gender and age on consumer–brand relationships. Doing so might provide an assessment of whether specific segments are more likely than others to feel emotionally bonded with spokes-characters and trust their endorsement. Finally, the results concerning nostalgia illustrate the importance of tapping into consumers’ memories. Given our findings concerning the usefulness of nostalgic memories in forming spokes-character trust and the recent illustrations of nostalgic campaigns in the current marketplace (Neeley et al. 2000), research addressing the role of nostalgia on brand evaluations and consumer–brand relationships appears to represent another opportunity. Using an experimental design, future research might specifically compare the spokes-characters recalled for brands used during childhood to those for brands currently used to further examine the potential nostalgic effects. Other experiments might assess the potential influence of spokescharacter features on character trust and brand attitude across segments with less and more exposure to spokes-characters and the brands they promote. Finally, research using a longitudinal design to gauge the importance of various spokes-character features at different points in time also appears warranted.
NOTES 1. In the survey instructions, we defined the term “character icon” as noncelebrity characters, consistent with the noncelebrity spokescharacter definition. Thus, we use the two terms synonymously. 2. The lack of convergence using structural equation modeling is likely due to the relatively small sample size when testing structural invariance across two groups. 3. Although not predicted, brand experience also moderated the correlation between character expertise and character relevance.
REFERENCES Alba, Joseph W., and J. Wesley Hutchinson (1987), “Dimensions of Consumer Expertise,” Journal of Consumer Research, 13 (March), 411– 454 Anderson, James C., and David W. Gerbing (1988), “Structural Equation Modeling in Practice: A Review and Recommended Two-Step Approach,” Psychological Bulletin, 103, 411–423. Atuahene-Gima, Kwaku, and Haiyang Li (2002), “When Does Trust Matter? Antecedents and Contingent Effects of Supervisee Trust on Performance in Selling New Products in China and the United States,” Journal of Marketing, 66 ( July), 61–81. Braun, Kathryn A., Rhiannon Ellis, and Elizabeth F. Loftus (2002), “Make My Memory: How Advertising Can Change Our Memories of the Past,” Psychology and Marketing, 19 (1), 1–23. Callcott, Margaret F. (1993), “The Spokes-Character in Advertising: An Historical Survey and Framework for Future Research,” Ph.D. diss., University of Texas. ———, and Patricia A. Alvey (1991), “Toons Sell . . . and Sometimes They Don’t: An Advertising Spokes-Character Typology and Exploratory Study,” in Proceedings of the 1991 Conference of the American Academy of Advertising, Rebecca Holman, ed., New York: D’Arcy Masius Benton and Bowles, 43–52. ———, and Wei-Na Lee (1994), “A Content Analysis of Animation and Animated Spokes-Characters in Television Commercials,” Journal of Advertising, 23 (4), 1–12. ———, and ——— (1995), “Establishing the Spokes-Character in Academic Inquiry: Historical Overview and Framework for Definition,” Advances in Consumer Research, 22, 144–151. ———, and Barbara J. Phillips (1996), “Observations: Elves Make Good Cookies: Creating Likable Spokes-Character Advertising,” Journal of Advertising Research, 36 (September/October), 73–79. Crosby, Lawrence A., Kenneth R. Evans, and Deborah Cowles (1990), “Relationship Quality in Services Selling: An Interpersonal Influence Perspective,” Journal of Marketing, 54 ( July), 68–81. Doney, Patricia M., and Joseph P. Cannon (1997), “An Examination of the Nature of Trust in Buyer–Seller Relationships,” Journal of Marketing, 65 (2), 35–51. Dotz, Warren, Jim Morton, and John William Lund (1996),
Summer 2004 35
What a Character! Twentieth-Century American Advertising Icons, San Francisco: Chronicle Books. Fiske, Susan T., and Mark A. Pavelchak (1986), “Category-Based Versus Piecemeal-Based Affective Responses: Developments in Schema-Triggered Affect,” in Handbook of Motivation and Cognition, Richard N. Sorrentino and E. Troy Higgins, eds., New York: Guilford, 167–203. Folkes, Valerie S., and Vanessa M. Patrick (2003), “The Positivity Effect in Perceptions of Services: Seen One, Seen Them All?” Journal of Consumer Research, 30 (1), 125–137. Forgas, Joseph P. (1995), “Mood and Judgement: The Affect Infusion Model,” Psychological Bulletin, 117 (1), 39–66. Fornell, Claus, and David F. Larcker (1981), “Evaluating Structural Equation Models with Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18, 39–50. Fournier, Susan (1998), “Consumers and Their Brands: Developing Relationship Theory in Consumer Research,” Journal of Consumer Research, 24 (March), 343–373. Fox, Richard J., Dean M. Krugman, James E. Fletcher, and Paul M. Fischer (1998), “Adolescents’ Attention to Beer and Cigarette Print Ads and Associated Product Warnings,” Journal of Advertising, 27 (3), 57–68. Garbarino, Ellen, and Mark S. Johnson (1999), “The Different Roles of Satisfaction, Trust, and Commitment in Customer Relationships,” Journal of Marketing, 63 (April), 70–87. Garretson, Judith A., and Scot Burton (1998), “Alcoholic Beverage Sales Promotion: An Initial Investigation of the Role of Warning Messages and Brand Characters Among Consumers Over and Under the Legal Drinking Age,” Journal of Public Policy and Marketing, 17 (1), 35– 47. “General Mills History of Innovation: The Pillsbury Doughboy” (2001), available at www.general mills.com/corporate/about/ history/hist_doughboy.pdf (accessed April 13, 2004). Gilmore, George W. (1919), Animism, Boston: Marshal Jones. Goodstein, Ronald C. (1993), “Category-Based Applications and Extensions in Advertising: Motivating More Extensive Ad Processing,” Journal of Consumer Research, 20 (1), 87–99. Gwinner, Kevin P., and John Eaton (1999), “Building Brand Image Through Event Sponsorship: The Role of Image Transfer,” Journal of Advertising, 28 (4), 47–57. Heckler, Susan, and Terry L. Childers (1992), “The Role of Expectancy and Relevancy in Memory for Verbal and Visual Information: What Is Incongruency?” Journal of Consumer Research, 18 (4), 475–792. Henke, Lucy L. (1995), “Young Children’s Perceptions of Cigarette Brand Advertising Symbols: Awareness, Affect, and Target Market Identification,” Journal of Advertising, 24 (4), 13–28. Hoch, Steven J. (2002), “Product Experience is Seductive,” Journal of Consumer Research, 29 (3), 448– 454. Holbrook, Morris B., and Robert M. Schindler (1994), “Age, Sex, and Attitude Toward the Past as Predictors of Consumers’ Aesthetic Taste for Cultural Products,” Journal of Marketing Research, 31 (3), 412–422. Hovland, Carl I., Irving L. Janis, and Harold H. Kelley (1953), Communication and Persuasion, New Haven: Yale University Press.
Hoy, Mariea Grubbs, Clifford E. Young, and John C. Mowen (1986), “Animated Host-Selling Advertisements: Their Impact on Young Children’s Recognition, Attitudes, and Behavior,” Journal of Public Policy and Marketing, 5, 171–184. Johnson, Branwell (2002), “Buried Treasure,” Marketing Week (April 25), 22–25. Jöreskog, Karl, and Dag Sörbom (1988), LISREL 7: A Guide to the Program and Applications, Chicago: SPSS. Kamins, Michael A. (1990), “An Investigation into the ‘MatchUp’ Hypothesis in Celebrity Advertising: When Beauty May Be Only Skin Deep,” Journal of Advertising, 19 (1), 4 –13. Kelley, Harold H. (1972), “Causal Schemata and the Attribution Process,” in Attribution: Perceiving the Causes of Behavior, E. E. Jones, D. E. Kanouse, Harold H. Kelley, R. E. Nisbett, S. Valins, and B. Winer, eds., Morristown, NJ: General Learning Press, 151–174. Kelly, Kathleen J., Michael D. Slater, David Karan, and Liza Hunn (2000), “The Use of Human Models and Cartoon Characters in Magazine Advertisements for Cigarettes, Beer, and Nonalcoholic Beverages,” Journal of Public Policy and Marketing, 19 (2), 189–200. Kempf, Deanna S., and Robert E. Smith (1998), “Consumer Processing of Product Trial and the Influence of Prior Advertising: A Structural Modeling Approach,” Journal of Marketing Research, 35 (3), 325–339. Lane, Vicki R. (2000), “The Impact of Ad Repetition and Ad Content on Consumer Perceptions of Incongruent Extensions,” Journal of Marketing, 64 (2), 80–91. Langer, Judith (1997), “What Consumers Wish Brand Managers Knew,” Journal of Advertising Research, 37 (6), 60–65. Lee, Yih Hwai, and Charlotte Mason (1999), “Responses to Information Incongruency in Advertising: The Role of Expectancy, Relevancy, and Humor,” Journal of Consumer Research, 26 (2), 156–169. Lindskold, Svenn (1978), “Trust Development, the GRIT Proposal and the Effects of Conciliatory Acts on Conflict and Cooperation,” Psychological Bulletin, 85 (4), 772–793. Macintosh, Gerrard, and Lawrence S. Lockshin (1997), “Retail Relationships and Store Loyalty: A Multi-level Perspective,” International Journal of Research in Marketing, 14 (5), 487–497. Macklin, M. Carole (1994), “The Effects of an Advertising Retrieval Cue on Young Children’s Memory and Brand Evaluations,” Psychology and Marketing, 11 (3), 291–311. ——— (1996), “Preschoolers’ Learning of Brand Names from Visual Cues,” Journal of Consumer Research, 23 (December), 251–261. Maruyama, Geoffrey M. (1998), Basics of Structural Equation Modeling, Thousand Oaks, CA: Sage. McGinnies, Elliott, and Charles D. Ward (1980), “Better Liked Than Right: Trustworthiness and Expertise as Factors in Credibility,” Personality and Social Psychology Bulletin, 6 (3), 467– 472. Miniard, Paul W., Sunil Bhatla, Kenneth R. Lord, Peter R. Dickson, and H. Rao Unnava (1991), “Picture-Based Persuasion and the Moderating Role of Involvement,” Journal of Consumer Research, 18 ( June), 92–107.
36
The Journal of Advertising
Mizerski, Richard (1995), “The Relationship Between Cartoon Trade Character Recognition and Attitude Toward Product Category in Young Children,” Journal of Marketing, 50 (October), 58–70. Moorman, Christine, Rohit Deshpande, and Gerald Zaltman (1993), “Factors Affecting Trust in Market Relationships,” Journal of Marketing, 57 ( January), 81–101. Morgan, Robert M., and Shelby D. Hunt (1994), “The Commitment–Trust Theory of Relationship Marketing,” Journal of Marketing, 58 ( July), 20–38. Neeley, Sabrina M., Wendy Macias, Tara M. Clark, and Wei-Na Lee (2000), “Advertising Spokes-Character Attributes and the Use Relationships,” in Proceedings of the 2000 Society of Consumer Psychology Conference, J. Inman, J. Tepper, and T. Whittler, eds., San Antonio, TX: Society of Consumer Psychology, 94–99. Ohanian, Roobina (1991), “The Impact of Celebrity Spokespersons’ Perceived Image on Consumers’ Intention to Purchase,” Journal of Advertising Research, 31 (1), 46–54. Peracchio, Laura, and Alice M. Tybout (1996), “The Moderating Role of Prior Knowledge in Schema-Based Product Evaluation,” Journal of Consumer Research, 23 (December), 177–192. Phillips, Barbara J. (1996a), “Advertising and the Cultural Meaning of Animals,” in Advances in Consumer Research, vol. 23, Kim Corfman and John Lynch, eds., Provo, UT: Association for Consumer Research, 354–360. ——— (1996b), “The Role of Advertising Trade Characters in Forming Product Perceptions,” in Proceedings of the 1996 Conference of the American Academy of Advertising, Gary B. Wilcox, ed., Austin: University of Texas, 171–179. ——— (1996c), “Defining Trade Characters and Their Role in American Popular Culture,” Journal of Popular Culture, 29 (4), 143–158. Poniewozik, James (2002), “Look Back in Angst,” Time, 160 (September 23), 73–74. Richins, Marsha L. (1994a), “Special Possessions and the Ex-
pression of Material Values,” Journal of Consumer Research, 21 (3), 522–533. ——— (1994b), “Valuing Things: The Public and Private Meanings of Possessions,” Journal of Consumer Research, 21 (3), 504–521. Sengupta, Jaideep, Ronald C. Goodstein, and David S. Boninger (1997), “All Cues Are Not Created Equal: Obtaining Attitude Persistence Under Low-Involvement Conditions,” Journal of Consumer Research, 23 (March), 351–361. Spears, Nancy E., John C. Mowen, and Goutam Chakraborty (1996), “Symbolic Role of Animals in Print Advertising: Content Analysis and Conceptual Development,” Journal of Business Research, 37, 87–95. Stafford, Marla Royne, Thomas F. Stafford, and Ellen Day (2002), “A Contingency Approach: The Effects of Spokesperson Type and Service Type on Service Advertising Perceptions,” Journal of Advertising, 31 (2), 17–34. Sujan, Mita, and James Bettman (1989), “The Effects of Brand Positioning Strategies on Consumers’ Brand and Category Perceptions: Some Insights from Schema Research,” Journal of Marketing Research, 26 (November), 454–467. Thompson, Stephanie (1999a), “Cap’n Crunch: ‘Ahoy, Adults!’” Adweek, 40 ( January 11), 6. ——— (1999b), “Jolly Green Giant Returns, with Irony,” Brandweek, 40 (August 9), 4. ——— (2002), “Coldwell Expands Icon from Web to TV Ads, Signs,” Advertising Age, 73 (August 5), 4, 40. Wegener, Duane T., and Leandre R. Fabrigar (2000), “Analysis and Design for Nonexperimental Data: Addressing Causal and Noncausal Hypotheses,” in Handbook of Research Methods in Social and Personality Psychology, Harry T. Reis and Charles M. Judd, eds., New York: Cambridge University Press, 412–450. Wood, Stacy L., and John G. Lynch, Jr. (2002), “Prior Knowledge and Complacency in New Product Learning,” Journal of Consumer Research, 29 (December), 416– 426.
