- Email: [email protected]
Animals in advertising: Love dogs? Love the ad!
Available online at www.sciencedirect.com
Journal of Business Research 61 (2008) 384 – 391
Animals in advertising: Love dogs? Love the ad!☆ Karen M. Lancendorfer a,⁎, JoAnn L. Atkin b,1 , Bonnie B. Reece c,2 a
c
Haworth College of Business, 3127 Schneider Hall, Western Michigan University, Kalamazoo, MI 49008-3812, United States b Haworth College of Business, 3243 Schneider Hall, Western Michigan University, Kalamazoo, MI 49008, United States Department of Advertising, Public Relations, & Retailing, Michigan State University, 315 Communication Arts & Sciences, East Lansing, MI 48824, United States Received 1 March 2006; received in revised form 1 July 2006; accepted 1 August 2006
Abstract Advertisers frequently use animals in ads, but little academic research focuses on consumer reactions to their use. This study uses the heuristic–systematic model (HSM) to examine consumer response to animal companions in advertisements. Specifically, HSM serves as the theoretical foundation for testing the effects of animal heuristic cues on the formation of attitude toward the ad, attitude toward the brand, and purchase intention. In the current study, the presence of the dog increases heuristic processing, concurrent processing, and ultimately attitude toward the ad. The article proposes managerial implications and avenues for future research. © 2007 Elsevier Inc. All rights reserved. Keywords: Heuristic–systematic model; Animals in advertising; Advertising effects
1. Introduction Animals often appear in advertisements, either in real or cartoon form. Past winners in USA Today's Ad Meter research include Pepsi commercials featuring flying geese and dancing bears, as well as Budweiser ads with ferrets, frogs, and lizards (Kim, Lim, and Bhargava, 1998). In the last four years, the top Ad Meter commercials feature animals, with almost one in five Super Bowl commercials in the past decade featuring an animal: Clydesdale horses, sheep, a zebra, a donkey, baboons, monkeys, a cat, and several varieties of dogs (Horovitz, 2003a,b, 2004, 2005, 2006). The use of animals in advertising implies that, by associating a brand with an attractive cue/stimulus, advertisers can favorably influence consumers' attitudes, even if the stimulus does not relate to the product and provides no product information (Kim, Lim, and Bhargava, 1998). Many advertising ☆ The authors gratefully acknowledge the comments of Nora J. Rifon, Sandi Smith, and Teresa Mastin, as well as the comments and helpful suggestions provided by the Journal of Business Research Editors, Arch G. Woodside, and Morris B. Holbrook, and the anonymous Journal of Business Research reviewers. ⁎ Corresponding author. Tel.: +1 269 387 5996. E-mail addresses: [email protected] (K.M. Lancendorfer), [email protected] (J.L. Atkin), [email protected] (B.B. Reece). 1 Tel.: +1 269 387 6238. 2 Tel.: +1 517 432 3586.
0148-2963/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2006.08.011
executives appear to believe that animals get attention, and they do not need research to confirm this proposal (Croke, 1992). In Super Bowl commercials, the animals generally have a leading role in the storyline; these commercials often involve humor. This article investigates whether or not animals (specifically, animal companions) are successful in print ads. The article explores the effect that animal companions have on consumers' information processing styles and on their attitudes toward the ad and the brand. The study also evaluates the impact that attitude toward pets has on ad processing and ad effects. 2. Background 2.1. The human–animal bond Pets are found in over 69 million U.S. homes; love, companionship, company, and affection are primary benefits people derive from sharing their lives with their pets, and 92% of pet owners view their pets as family members (American Pet Products Manufacturers Association, 2006). With almost twothirds of U.S. households having at least one pet, pets represent a large and growing market (Dale, 2005). Spending by U.S. consumers on their 164 million dogs and cats has doubled in the past decade to $36 billion in 2005 and an estimated $38.4 billion in 2006 (American Pet Products Manufacturers
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
Association, 2006). With these trends, marketers recognize the importance of animal companions in the lives and experiences of consumers (Aylesworth, Chapman, and Dobscha, 1999). Animal companions often serve as extensions of a consumer's self (Belk, 1988), and animal companions affect the social identity of the owner, as well as the owner's self perceptions (Sanders, 1990). Hirschman's (1994) seminal article suggests that animals may also act as friends and family members. Research shows that animal companions enhance consumers' well-being through the medical and psychological benefits of pet-relating consumption experiences (American Pet Products Manufacturers Association, 2006). Holbrook et al. (2001) conclude that pets provide opportunities for deeply involving experiences; for example, to appreciate nature and wildlife, for inspiration, to be playful, to be altruistic and nurturing. Using a photo-essay approach, Holbrook et al. (2001) illustrate the warm and enduring companionship pets offer and the ways in which humans love pets, treat pets as family members, and deeply mourn pets when their lives end. Advertisers often attempt to tap this human–animal bond when they use animals in persuasive messages. 2.2. Animals and advertising Marketing promotions regularly use animals as visual symbols, ranging from the AFLAC duck to the Energizer Bunny to the Taco Bell Chihuahua (Feldhamer et al., 2002). Often, consumers' awareness of the animal link to the brand is higher than that for advertisements starring human celebrities (Hoggan, 1989). Despite their widespread use in advertising, however, little academic research is available on this topic. When it is available, such research focuses only on studies using animals as stimuli to explore hypotheses regarding classical conditioning or the effects of pictures on persuasion (Kim et al., 1998; Miniard et al., 1991). Lerner and Kalof (1999) examine the use of animals as cultural symbols in television commercials. They study six primary themes in the portrayal of animals: as loved ones, as symbols, as tools, as allegories, as nuisances, and as part of nature. Phillips (1996) hypothesizes that animal characters proliferate in advertising because animal characters transfer meanings to the brands. If most consumers associate positive cultural meanings with specific animal characters, advertisers can exploit those meanings (Phillips, 1996). Spears et al. (1996, p. 90) suggest that, when advertising uses animals, “consumers are influenced by both the symbolic meanings that have been culturally assigned to that animal as well as the physical attractiveness and likeability of how the animal is portrayed.” Their analysis of print advertisements featuring animals reveals that advertisers associate particular animals with particular product categories. 2.3. Heuristic–systematic model of persuasive communication (HSM) The heuristic–systematic model (HSM) is a dual-process model that is useful for explaining how individuals process
385
information in persuasive messages and helps clarify the influence of animal companions in advertisements on consumer attitudes toward the ad and the brand. The HSM posits that people can process information systematically or heuristically and that these are qualitatively different modes of processing (Todorov et al., 2002). According to the HSM, individuals are cognitive misers and expend effort on a cognitive task only when they have sufficient motivation and resources to do so (Chaiken and Maheswaran, 1994). Highly motivated individuals engage in systematic processing by making an effort to analyze and understand relevant information. Less motivated individuals base their decisions on simple decision rules (heuristics). These heuristic processors use mental shortcuts in lieu of engaging in issue-relevant thinking to form a judgment. Many people process messages this way, judging their validity and making decisions through the use of superficial cues such as message length, a spokesperson, or the presentation of statistical data (Griffin et al., 2002). Such heuristics often derive from experience and have empirical validity. HSM allows both systematic and heuristic processing to occur. Depending on the level of an individual's involvement with the message content, heuristic and systematic processing can proceed concurrently, or one mode may dominate. Additivity is the concurrent occurrence of heuristic and systematic processing. Additivity shows through the generation of both heuristic cue-related and attribute-related thoughts (Chaiken and Maheswaran, 1994). Empirical research notes the role of both systematic and heuristic processing as a mediator of advertising's effect on brand attitudes (see MacKenzie and Spreng, 1992 for a review). Because the HSM presents two different but intertwined routes to persuasion, HSM is important for understanding the factors that determine when brand attitudes might be influenced by consumers' diligent processing of advertising information and when they might be influenced primarily by cues that trigger heuristic processing (MacKenzie and Spreng, 1992). The study here investigates the use of animal companions as advertising cues that should induce a heuristic processing strategy. Although little research is available for guiding a heuristic view of persuasion, individuals often agree or disagree with a message on the basis of their reactions to cues such as communicator credibility and mood (Chaiken, 1980). While marketers are beginning to study the animals-andadvertising link, this area remains fertile ground for research attention. The majority of published research is qualitative in nature. Quantitative research can also add to understanding in this area. Aylesworth et al. (1999, p. 388) suggest that exploring the “effect of using animals in advertising on constructs such as attitude toward the ad, attitude toward the brand, and memorability are areas in which both researchers and practitioners should have a keen interest.” 3. Theoretical concepts and hypotheses The current study extends research on the use of animals and animal companions in ads by applying the HSM model of persuasion in order to illustrate the effects of heuristic
386
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
processing on typical measures of advertising effectiveness. The principal outcome variables of advertising effectiveness include attitude toward the ad (Aad), attitude toward the brand (Abr), and purchase intent (PI) (MacKenzie and Lutz, 1989). The HSM is useful for understanding the persuasive effects of communication variables such as source credibility and argument strength (Chaiken, 1980; Chaiken and Maheswaran, 1994). In these cases, heuristics derive from experience, and the presence of relevant heuristic cues triggers heuristics (Todorov et al., 2002). Examples of persuasion heuristics include “length implies strength” and “consensus implies correctness.” Although prior research does not use animals as a specific heuristic cue, their appearance in advertisements is often an attention-getting device not relating directly to the product being advertised. Accordingly, with other things being equal for a given campaign, the study proposes the following hypotheses: H1a. An ad with an animal cue evokes greater heuristic processing than an ad without an animal cue. H1b. An ad without an animal cue evokes greater systematic processing than an ad with an animal cue. Much research hypothesizes that involvement leads to greater perceptions of attribute differences and product importance as measures of ad effectiveness (Zaichkowsky, 1985). Involvement may affect information processing and may serve as a good gauge of motivation to process (Celsi and Olson, 1988; Zaichkowsky, 1985). Consumers with higher involvement focus more attention on processing, put greater effort into processing, and better comprehend the information an ad presents (Celsi and Olson, 1988), as they are processing systematically. Consumers with lower involvement focus less attention on processing and put less effort into processing an ad, as they process the ad heuristically. Heuristic cues such as source expertise or attractiveness have a greater impact on persuasion under low involvement as opposed to high involvement conditions (Chaiken, 1980). Therefore, the study hypothesizes as follows. H2a. Individuals with lower levels of involvement use a more heuristic and less systematic processing strategy for an ad. H2b. Individuals with higher levels of involvement use a more systematic and less heuristic processing strategy for an ad. According to the HSM, contextual cues link with the heuristic route to attitude change, and prior research notes that Aad may be formed through a heuristic process in response to ad-execution cues rather than brand-message arguments (Lord et al., 1995). Brown et al. (1998) find that conditions favoring greater cognitive elaboration produce weaker Aad than conditions not favoring elaboration. Therefore, the study proposes the following hypothesis for the advertisements containing an animal cue. H3. Systematic processors have lower Aad scores than heuristic processors. Considerable research on advertising effectiveness finds a relationship between Aad and Abr and a relationship between Abr
and PI (e.g., MacKenzie and Lutz, 1989; Muehling and McCann, 1993). Brown and Stayman's (1992) meta-analysis shows significant correlations among Aad, Abr, and PI. In order to determine whether heuristic cues such as animals have the same effects on the dependent variables, the study hypothesizes as follows. H4. Aad relates positively to attitude toward the brand. H5. Abr relates positively to purchase intent. This research also proposes that an individual's attitude toward pets may affect the interpretation and evaluation of the ad. A great deal of psychological research links pets to a general sense of well-being in humans (American Pet Products Manufacturers Association, 2006). When appearing in ad messages, animals may tap into a human–animal bond but only for consumers with positive attitudes toward pets. H6. For the ad containing an animal cue, individuals having positive attitudes toward pets have more favorable attitudes toward the ad than those who have less favorable attitudes toward pets. In summary, animal companions in ads act as positive heuristic cues on Aad resulting in positive Abr and, subsequently, higher PI. The expectation is that systematic processing of the message in the ad results in these outcomes as well, although the effect of heuristic processing likely is stronger due to the presence of the animal in the ad. In addition, their effect depends on the level of product involvement, as well as pre-existing attitudes toward pets. 4. Method 4.1. Experimental design The study employs a 2 × 3 between-respondents fixed factor design, with ad format (with animal/without animal) and involvement (high/medium/low) as the experimental factors. Participants were recruited from business classes at a large midwestern university, were given class credit for their participation in the research, and participated as one large group in a classroom setting. A total of 87 female (33%) and 166 male (64%) undergraduates participated in the study (with 7 participants not indicating gender). Their ages range from 18 to 51 years, with an average of 20years. The participants were predominantly Caucasian (73%). The primary academic majors include marketing (17%), business, (16%), management (14%), and accounting (10%). Although students represent a convenience sample, they are appropriate targets for credit cards (the stimulus product). According to Nellie Mae, 76% of college students had at least one credit card in 2004, up from 67% in 1998 (Nellie Mae, 2005). In addition, the total number of cards held by students has increased from three cards in 2000 to four or more in 2004 (Nellie Mae, 2005). Participants randomly received one of two test-booklet conditions, passed out by student administrators (who were blind to the experimental conditions) with the understanding that they would be asked to
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
evaluate the color-advertising execution in the booklet as part of a test market. Within the booklet, the participants answered pre-exposure questions relating to brand evaluations and involvement regarding the product category of credit cards. Then they were exposed to one of two ads (see Appendix A). After exposure, the participants answered questions dealing with the dependent variables and with attitude toward pets. 4.2. Choice of stimuli (advertisements) Two original ads for MasterCard were created as the stimuli. One full-page ad included a dog and the other had no animal. The dog/no-dog conditions were manipulated by presenting respondents with only one ad in their test booklet. A dog was selected because they are the most sociable of the domestic animals kept as pets in Western culture; dogs are highly amenable to sociological analysis (Sanders, 1990). Further, more U.S. households own dogs than any other pets (American Pet Products Manufacturers Association, 2006). 4.3. Measures The independent variables include ad format (with or without dog), involvement, and attitude toward pets. The main dependent variables – Aad, Abr, and PI – were measured after exposure to the stimulus. Brand evaluations and product involvement were measured prior to exposure to the ad; and attitudes toward pets, product familiarity, and demographic information (including pet ownership) were measured after exposure. Six brands (Coca-Cola, Visa, McDonald's, Pepsi, MasterCard, and Burger King) were included in pre-exposure brand evaluations in order to minimize priming effects. Independent variables. Zaichkowsky's (1985) personal involvement inventory (PII) was utilized to assess participants' involvement with the stimulus-product category. The PII is a summative index composed of 20 seven-point bipolar adjective pairs (e.g., important/unimportant, appealing/unappealing, mundane/fascinating). In the current experiment, the scale measures involvement with the product category of credit cards in general. Research on processing modalities typically measures the number of cognitions an ad generates using a thought-listing technique (MacKenzie and Spreng, 1992). As applied to the HSM, systematic processors should produce more cognitions about the ad than heuristic processors. Under the heuristic mode, participants should generate fewer thoughts, and those thoughts are less likely to be message-related and more likely to be related to other cues. Advertising cognitions were obtained by asking respondents, immediately after viewing the ad, to write down all the thoughts, reactions, ideas, and feelings they had while looking at the ad (MacKenzie and Lutz, 1989; Miniard et al., 1991). Two trained judges coded the responses into four categories, with each “thought” as the unit of analysis. Examples of messagerelevant thoughts/systematic processing are: “With MasterCard you can earn points”; “For everything else there's MasterCard”; “You earn points the more you spend”; and “The whole price-
387
less thing is good.” Animal heuristic thoughts include: “I liked the dog”; “Cute dog”; “I want a dog”; and “I miss my puppy.” Examples of other heuristic thoughts are: “That room is a mess”; “Ugly furniture”; “I thought of furniture I had”; and “Colorful room.” Examples of non-related thoughts are: “I don't want to work tonight” and “I wonder if that girl below me is single.” Twenty-five percent of the sample was chosen to establish unitizing and coding reliability, and Guetzkow's (1950) U for the unit of analysis between coders was .96. The final overall inter-coder reliability using Scott's Pi (1955) to correct for chance agreement was .93. The Pet Attitude Scale (PAS) was utilized to measure participants' attitudes toward pets with higher scores indicating a more favorable attitude toward pets (Templer et al., 1981). PAS is an 18-item Likert-type scale that has been highly reliable in previous research. Dependent variables. Attitudes toward the ad and the brand were assessed using three, seven-position semantic differential items: good/bad, pleasant/unpleasant, and favorable/unfavorable (MacKenzie and Lutz, 1989). To measure purchase intent, respondents were asked three five-point Likert-type scale items (strongly agree/strongly disagree): “It is very likely that I would use [brand] if I had one”; “I will use [brand] the next time I need to use a credit card”; and “I will definitely try [brand]” (Putrevu and Lord, 1994). 5. Results 5.1. Manipulation check The dog/no-dog conditions were manipulated by presenting respondents with either the experimental ad, which includes a dog, or the control ad without a dog. In order to confirm that the respondents in the dog-present treatment notice the dog, the following questions were asked: “In the advertisement you just saw, was there an animal in the ad?” and “If yes, please write in the kind of animal that was in the ad.” Only three respondents claimed to have seen an animal in ads that did not have one, while 100% of the respondents correctly identified a dog in their ads. Booklets with incorrect answers were removed from the data set, leaving 260 respondents.
Table 1 Mean scores and standard deviations for independent and dependent variables
Independent variables Involvement Attitude toward the pet Dependent variables Attitude toward the ad Attitude toward the brand Purchase intent Total systematic thoughts Total heuristic thoughts Total thoughts N = 260.
Scale mean
Std. dev.
a
95.58 94.16
23.80 21.21
.954 .934
13.88 13.74 9.01 1.77 2.67 4.72
4.75 4.51 2.79 1.46 1.94 2.26
.920 .936 .807 n.a. n.a. n.a.
388
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
Table 2 Pearson correlations among dependent variables
Attitude toward the ad Attitude toward the brand Purchase intent
Attitude toward the ad
Attitude toward the brand
Purchase intent
1.00 .535 ⁎⁎ .250 ⁎⁎
1.00 .510 ⁎⁎
1.00
N = 260. ⁎⁎ p b .01, one-tailed.
5.2. Initial analyses Before addressing the specific hypotheses, the scales used to measure the independent and dependent variables were checked for internal consistency and unidimensionality. Table 1 summarizes the mean, standard deviation, and Cronbach's coefficient alpha for each independent and dependent composite variable. The respondents were split to create three subsamples with high, medium, and low involvement. Respondents with involvement totals ranging from 20 to 90 were classified as low involvement; involvement totals from 91 through 108 were classed as medium involvement; and involvement totals from 109 to 140 were designated as high involvement. Unless otherwise stated, all further analyses with involvement as a variable use this three-way split. 5.3. Hypothesis testing Hypothesis 1a proposes that the presence of an animal cue in an ad leads to greater heuristic processing than occurs if no animal is present. A one-way between-respondents ANOVA on the number of heuristic thoughts indicates that respondents viewing the ad with versus without the dog generate more heuristic thoughts: Mdog = 3.11 versus Mnodog = 2.23, F(1, 258) = 14.16, p b .01. These findings support H1a. In contrast, H1b posits that ads without an animal cue evoke greater systematic processing than ads with the animal present. A oneway between-respondents ANOVA on the number of systematic thoughts indicates that respondents viewing the ad without versus with the dog generate more systematic thoughts: Mnodog = 2.00 versus Mdog = 1.56, F(1, 258) = 6.10, p b .01. These results support H1b. Hypotheses 2a and 2b suggest that involvement with the product influences the number of heuristic versus systematic thoughts, with individuals having low involvement utilizing a more heuristic strategy and individuals with high involvement utilizing a more systematic strategy. A one-way between-respondents ANOVA indicates that there is a statistically significant difference in the total number of heuristic thoughts for different levels of involvement (F(2, 236) = 4.37, p = .01). However, Scheffe's post-hoc analysis reveals that the results are not in the predicted direction. Those with medium levels of involvement have significantly fewer heuristic thoughts (M = 2.31) than those with the highest level of involvement (M = 3.18), and those with the lowest level of involvement
(M = 2.57) are not significantly different from those with either medium or high levels of involvement. These results, in fact, suggest a curvilinear relationship and do not support H2a. According to results of a separate ANOVA, although those with low involvement have fewer systematic thoughts than those with higher levels of involvement as predicted (Mlow = 1.65, Mmed = 2.05, Mhigh = 1.73), there is no statistically significant difference in the total number of systematic thoughts for different levels of involvement (F(2, 236) = 1.65, p = .194). Again, there is a slight curvilinear relationship and no support for H2b. Two-way ANOVAs for the number of heuristic and systematic thoughts, testing for both ad type and involvement, confirm the results for H1a, H1b, H2a and H2b and suggest that there are no interaction effects. For the ad with a dog, Hypothesis 3 proposes that processing strategy will impact Aad in that greater systematic processing will result in lower Aad, while greater heuristic processing will result in higher Aad. To test this hypothesis, the study separates respondents into categories (systematic, concurrent, or heuristic processors) depending on the relative balance of systematic or heuristic thoughts on the thought-listing measure. This results in three uneven groups with four respondents being classified as systematic processors, 89 as concurrent processors, and 30 as heuristic processors. A one-way between-respondents ANOVA indicates no significant differences among the three groups: Msystematic = 12.50, Mconcurrent = 14.26, Mheuristic = 13.97, F(2, 120) = .275, NS. These results do not support H3. The same analysis for the ad without a dog produces different results. In this case, systematic processors have the highest score for Aad (n = 22, M = 15.82), with concurrent processors (n = 78, M = 13.73) and heuristic processors (n = 23, M = 12.35) at lower levels (F(2,120) = 3.24, p b .05). Scheffe's post-hoc analysis shows that the difference in Aad between systematic and heuristic processors is statistically significant. In support of H4 and H5, Table 2 shows significant correlations between Aad and Abr and between Abr and PI. As Table 1 indicates, scores on the attitude-toward-pets scale are fairly high, with the mean well above the mid-point of the scale (94 versus 72). One hundred fifty-two students (58.5%) currently own a pet, and only fourteen of the participants have never owned one. Participants cite dogs most frequently as the type of pet owned. Hypothesis 6 proposes that attitude toward pets (Apet) relates to respondents' Aad for those who saw ads with a dog. The Pearson's correlation between Apet and Aad is .37 ( p b .01), supporting H6. In order to investigate the effects of attitude toward pets more thoroughly, an ANOVA was conducted for Aad with processing Table 3 Effects of processing style on attitude toward ad Source
Degrees of freedom
Mean square
F
Significance
Model a Attitude toward pet Processing style Error
4 1 3 111
5749.18 333.36 79.03 20.91
274.92 15.941 3.779
.000 .000 .013
a
Adjusted R Squared = .905.
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
type as the independent variable and Apet as a covariate. Here, we find a main effect for processing type and also a significant effect for Apet (Table 3). Concurrent processors have the highest scores for Aad (Mconcurrent = 14.16, compared with Msystematic = 12.5 and Mheuristic = 13.89).
389
search and theory. Respondents have generally high scores for involvement with credit cards. Some individuals in the low involvement group have scores above the mid-point of the scale; so there may not be enough variation among the three groups to test this variable appropriately. Thus, involvement does not influence processing strategy or Aad.
6. Discussion 6.1. Managerial implications, future research, and limitations This study illustrates the impact that executional elements in an ad have on audience processing style and subsequent attitudes toward the ad and brand. Although the findings are not as straightforward as hypothesized, they suggest that animals are popular in advertising for a reason. The two stimulus ads in the experiment have nearly identical Aad scores (Mdog = 14.08; Mnodog = 13.75; F(1, 248) = .294, NS), but the processes through which those scores are attained are quite different. In the no-dog condition, respondents generate about the same number of systematic and heuristic thoughts (2.0 versus 2.2, respectively). The majority of respondents are considered to be concurrent processors, but they are more likely to be classified as systematic processors than those in the dog condition. Systematic processors have more favorable Aad than heuristic processors for this ad. Attitude toward the ad is strongly related to Abr, which is strongly linked to purchase intent. Respondents in the dog-ad group generate significantly more heuristic than systematic thoughts (3.1 versus 1.56, respectively, t = 7.1, p b .01); thus they are more likely to be classified as heuristic processors than those in the no-dog group. The majority are concurrent processors, and there are very few systematic processors. It is interesting that this group has, on average, more total thoughts relevant to the ad than those in the no-dog group. They are not using mental shortcuts, as predicted by theory, but process the executional elements at some length. For those who saw the dog ad, processing style has no simple, direct effect on Aad, but it does have a positive influence when Apet is used as a covariate. For Aad, concurrent processors have the highest score, and systematic processors have the lowest. In other words, if readers only evaluate the message arguments, they may not find the ad particularly appealing; but if they also process the heuristic cues, then ad liking increases. The results of this research support the premise of the heuristic–systematic model that both systematic and heuristic processing can occur when people receive a persuasive message. This study hypothesizes that the presence of the dog in the ad suppresses systematic processing and increases heuristic processing, ultimately leading to increased attitude toward the ad. The results indicate an increased attitude toward the ad when respondents heuristically process the animal cues in the advertising, whether they do that exclusively or concurrently with message arguments about product attributes. Either way, Aad is strongly related to Abr, which is strongly linked to PI. Involvement with the product category does not play a role in processing for either ad, which is contrary to previous re-
When considering the objectives of the advertisement, advertisers may want to consider what type of processing they are attempting to invoke (systematic versus heuristic). In the current study, the presence of the dog increases heuristic processing, concurrent processing, and ultimately attitude toward the ad. If advertisers seek to invoke heuristic processing, the use of an animal companion in an ad may help to achieve this strategy. However, if advertisers wish to invoke systematic processing in which audience members focus on the quality of message arguments, then the use of an animal in the advertising message may derail this process. Advertisers for whom systematic processing and changes in brand beliefs are important goals need to be mindful in considering the appropriateness of utilizing animals in their advertising campaigns. When animals win ratings for Super Bowl commercials, they are often the stars of the ad. Those commercials typically have a relatively simple message. Likewise, in the current research, the dog was a focal point for the ad and the message was not very complex. However, the outcomes may not be the same if the animal companion is not the focus of the ad or if the message is a complex one. Future research should explore additional creative executions that utilize animal companions to determine the effects of variations in companion prominence and message length/ complexity. Another area of fruitful exploration concerns the congruence of the animal with the product in the advertisement. Prior research supports a strong positive effect of thematic congruence on memory for advertising and attitude toward the ad (Moorman et al., 2002). Future research should compare congruent (e.g., a dog in a dog food ad) and non-congruent (e.g., a dog in a life insurance ad) advertisements containing animals in order to determine if congruency is an important variable. The congruence issue is irrelevant, however, if animals are included in the ad simply as attention-getting devices (the “cute factor”). Attention as an outcome measure was not tested in this study. The experiment presented a single ad to the respondents in an artificial setting in order to test their reactions. If participants receive multiple ads or if they are flipping through magazines and responding to ads in a natural manner, measuring attention (recall) would make sense. This criterion is important for future studies that look at whether type of animal (popular companions or more exotic pets) makes a difference. Yet another avenue for future research relates to our measurement of involvement. Involvement is often a very difficult variable to quantify, as different researchers have noted several
390
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
types of broad involvement or involvement as a multidimensional construct with several underlying dimensions (Broderick and Mueller, 1999; Laurent and Kapferer, 1985) — including normative, enduring, situational, and hedonic involvement. In our study, involvement was assessed through the use of Zaichkowsky's (1985) involvement with the product category scale. Future research should consider other measures of involvement, as different types of involvement (e.g., hedonic versus normative) might correspond with more/less systematic or heuristic processing and might ultimately lead to differences in the processing of the animal cue in the advertising message. The current research is subject to the usual limitations of experimental research. First, the use of college students as participants might be a concern that limits the generalizability of results. However, undergraduates are especially appropriate for this research because they comprise a homogeneous target market for marketers. Future research should attempt to investigate perceptions of animals in advertising messages among non-student consumer groups. Another limitation of this research relates to the choice of the stimuli in the experiment. Although the ads were experimentally designed, the brand (MasterCard) and its overall message strategy (“Priceless”) are familiar to students. When using real brands, the researcher runs the risk that participants have preexisting knowledge, attitudes, and beliefs about the brand that will affect their responses. Obviously, the results of this study are specific to the advertising stimuli used and cannot be generalized to all credit-card advertisements. Future research
Appendix A. Advertising stimuli
should investigate whether the results hold across other product categories. Another important limitation is the fact that the ads in the treatment and control manipulations were similar in format and message, but not in visual cues. This is appropriate in that it mirrors common real-world advertising campaigns, but is problematic in that the differences between the two conditions (not using advertisements identical in every way except for the animal cue) may have confounded the results and should be recognized as potential rival explanations for the pattern of data. While we attempted to control for extraneous variables that might confound our experiment, we erred on the side of “practical application” utilized by practitioners in an effort to achieve a “unified field theory” (Carlson et al., 2005) that walks the line between theoretical study and practical application. 7. Conclusion Using animals is a common practice for advertising campaigns today, despite the “hassles” and costs associated with working with animals. The accepted belief is that animals in advertisements inspire good feelings about the advertisement and the brand. The current research supports the anecdotal evidence that the use of animals in advertising has positive effects, especially when combined with a favorable attitude toward pets, but it also suggests that non-animal ad executions can be equally effective.
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
References The American Pet Products Manufacturers Association. Pet industry spending figures; 2006 [March 23], http://www.appma.org. Aylesworth A, Chapman K, Dobscha S. Animal companions and marketing: dogs are more than just a cell in the BCG matrix! Adv Consum Res 1999;26: 385–91. Belk RW. Possessions and the extended self. J Consum Res 1988;15:139–68 [September]. Broderick AJ, Mueller RD. A theoretical and empirical exegesis of the consumer involvement construct. J Mark Theory Prac 1999;7(4):97–108. Brown SP, Homer PM, Inman JJ. A meta-analysis of relationships between adevoked feeling and advertising responses. J Mark Res 1998;35(1):114–26. Brown SP, Stayman DM. Antecedents and consequences of attitude toward the ad: a meta-analysis. J Consum Res 1992;19(1):34–51. Carlson L, Grove SJ, Stafford MR. Perspectives on advertising research. J Advert 2005;34(2):117–48. Celsi R, Olson JC. The role of involvement in attention and comprehension processes. J Consum Res 1988;15:210–24 [September]. Chaiken S. Heuristic versus systematic information processing and the use of source versus message cues in persuasion. J Pers Soc Psychol 1980;39 (5):752–66. Chaiken S, Maheswaran D. Heuristic processing can bias systematic processing: effects of source credibility, argument ambiguity, and task importance on attitude judgment. J Pers Soc Psychol 1994;66:460–73 [March]. Croke V. It's a bull market for animals in ads. Austin American-Statesman; 1992. p. D9. [December 5]. Dale S. Animal smart. USA Weekend 2005; July 10. Feldhamer G, Whittaker J, Monty A, Weickert C. Charismatic mammalian megafauna: public empathy and marketing strategy. J Pop Cult 2002;36 (1):160–8. Griffin RJ, Neuwirth K, Giese J, Dunwoody S. Linking the heuristic–systematic model and depth of processing. Communic Res 2002;29(6):705–32. Guetzkow H. Unitizing and categorizing problems in coding qualitative data. J Clin Psychol 1950;6:47–58. Hirschman EC. Consumers and their animal companions. J Consum Res 1994;20 (March):616–32. Hoggan K. Four legs good, two legs bad. Marketing 1989; March 2: 1. Holbrook MB, Stephens DL, Day E, Holbrook SM, Strazar G. A collective stereographic photo essay on key aspects of animal companionship: the truth about dogs and cats. Acad Market Sci Rev 2001:1–17. Horovitz B. Smile! You're the stars of the Super ad Bowl: America loves to laugh at ads with comical critters. USA Today 2003a; January 24. Horovitz B. Anheuser-Busch's ‘Replay’ is top Super Bowl ad. USA Today 2003b; January 27. Horovitz B. Viewers' favorite ads crude, rude and furry. USA Today 2004; February 2.
391
Horovitz B. Anheuser-Busch still king of ads as marketers play it safe. USA Today 2005; February 7. Horovitz B. ‘Magic fridge’ of Bud Light ices an advertising win. USA Today 2006; February 5. Kim J, Lim J, Bhargava M. The role of affect in attitude formation: a classical conditioning approach. J Acad Mark Sci 1998;26(2):143–52. Laurent G, Kapferer J. Measuring consumer involvement profiles. J Mark Res 1985;22(1):41–53. Lerner JE, Kalof L. The animal text: message and meaning in television advertisements. Sociol Q 1999;40(4):565–86. Lord K, Lee M, Sauer PL. The combined influence hypothesis: central and peripheral antecedents of attitude toward the ad. J Advert 1995;24(1):73–85. MacKenzie SB, Lutz RJ. An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. J Mark 1989;53:48–65 [April]. MacKenzie SB, Spreng RA. How does motivation moderate the impact of central and peripheral processing of brand attitudes and intentions? J Consum Res 1992;18(4):519–29. Miniard PW, Bhatla S, Lord KR, Dickson PR, Unnava HR. Picture-based persuasion processes and the moderating role of involvement. J Consum Res 1991;18(1):92–107. Moorman M, Neijens PC, Smith EG. The effects of magazine-induced psychological responses and thematic congruence on memory and attitude toward the ad in a real-life setting. J Advert 2002;31(4):27–40. Muehling DD, McCann M. Attitude toward the ad: a review. J Curr Issues Res Advert 1993;15(2):25–58. Nellie Mae. Undergraduate students and credit cards in 2004; 2005 [May], www.nelliemae.com. Phillips BJ. Advertising and the cultural meaning of animals. Adv Consum Res 1996;23:354–60. Putrevu S, Lord KR. Comparative and noncomparative advertising: attitudinal effects under cognitive and affective involvement conditions. J Advert 1994;23(2):77–90. Sanders CR. The animal ‘other’: self-definition, social identity and companion animals. Adv Consum Res 1990;17:662–8. Scott W. Reliability of content analysis: the case of nominal scale coding. Public Opin Q 1955;19:321–5. Spears NE, Mowen JC, Chakraborty G. Symbolic role of animals in print advertising: content analysis and conceptual development. J Bus Res 1996;37:87–95. Templer DI, Slater CA, Dickey S, Baldwin R, Veleber DM. The construction of a pet attitude scale. Psychol Rec 1981;31:343–8. Todorov A, Chaiken S, Henderson MD. The heuristic–systematic model of social information processing. In: Dillard JP, Pfau M, editors. The persuasion handbook. Thousand Oaks, CA: Sage Publications; 2002. p. 195–211. Zaichkowsky JL. Measuring the involvement construct. J Consum Res 1985;12:341–52 [December].
Journal of Business Research 61 (2008) 384 – 391
Animals in advertising: Love dogs? Love the ad!☆ Karen M. Lancendorfer a,⁎, JoAnn L. Atkin b,1 , Bonnie B. Reece c,2 a
c
Haworth College of Business, 3127 Schneider Hall, Western Michigan University, Kalamazoo, MI 49008-3812, United States b Haworth College of Business, 3243 Schneider Hall, Western Michigan University, Kalamazoo, MI 49008, United States Department of Advertising, Public Relations, & Retailing, Michigan State University, 315 Communication Arts & Sciences, East Lansing, MI 48824, United States Received 1 March 2006; received in revised form 1 July 2006; accepted 1 August 2006
Abstract Advertisers frequently use animals in ads, but little academic research focuses on consumer reactions to their use. This study uses the heuristic–systematic model (HSM) to examine consumer response to animal companions in advertisements. Specifically, HSM serves as the theoretical foundation for testing the effects of animal heuristic cues on the formation of attitude toward the ad, attitude toward the brand, and purchase intention. In the current study, the presence of the dog increases heuristic processing, concurrent processing, and ultimately attitude toward the ad. The article proposes managerial implications and avenues for future research. © 2007 Elsevier Inc. All rights reserved. Keywords: Heuristic–systematic model; Animals in advertising; Advertising effects
1. Introduction Animals often appear in advertisements, either in real or cartoon form. Past winners in USA Today's Ad Meter research include Pepsi commercials featuring flying geese and dancing bears, as well as Budweiser ads with ferrets, frogs, and lizards (Kim, Lim, and Bhargava, 1998). In the last four years, the top Ad Meter commercials feature animals, with almost one in five Super Bowl commercials in the past decade featuring an animal: Clydesdale horses, sheep, a zebra, a donkey, baboons, monkeys, a cat, and several varieties of dogs (Horovitz, 2003a,b, 2004, 2005, 2006). The use of animals in advertising implies that, by associating a brand with an attractive cue/stimulus, advertisers can favorably influence consumers' attitudes, even if the stimulus does not relate to the product and provides no product information (Kim, Lim, and Bhargava, 1998). Many advertising ☆ The authors gratefully acknowledge the comments of Nora J. Rifon, Sandi Smith, and Teresa Mastin, as well as the comments and helpful suggestions provided by the Journal of Business Research Editors, Arch G. Woodside, and Morris B. Holbrook, and the anonymous Journal of Business Research reviewers. ⁎ Corresponding author. Tel.: +1 269 387 5996. E-mail addresses: [email protected] (K.M. Lancendorfer), [email protected] (J.L. Atkin), [email protected] (B.B. Reece). 1 Tel.: +1 269 387 6238. 2 Tel.: +1 517 432 3586.
0148-2963/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2006.08.011
executives appear to believe that animals get attention, and they do not need research to confirm this proposal (Croke, 1992). In Super Bowl commercials, the animals generally have a leading role in the storyline; these commercials often involve humor. This article investigates whether or not animals (specifically, animal companions) are successful in print ads. The article explores the effect that animal companions have on consumers' information processing styles and on their attitudes toward the ad and the brand. The study also evaluates the impact that attitude toward pets has on ad processing and ad effects. 2. Background 2.1. The human–animal bond Pets are found in over 69 million U.S. homes; love, companionship, company, and affection are primary benefits people derive from sharing their lives with their pets, and 92% of pet owners view their pets as family members (American Pet Products Manufacturers Association, 2006). With almost twothirds of U.S. households having at least one pet, pets represent a large and growing market (Dale, 2005). Spending by U.S. consumers on their 164 million dogs and cats has doubled in the past decade to $36 billion in 2005 and an estimated $38.4 billion in 2006 (American Pet Products Manufacturers
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
Association, 2006). With these trends, marketers recognize the importance of animal companions in the lives and experiences of consumers (Aylesworth, Chapman, and Dobscha, 1999). Animal companions often serve as extensions of a consumer's self (Belk, 1988), and animal companions affect the social identity of the owner, as well as the owner's self perceptions (Sanders, 1990). Hirschman's (1994) seminal article suggests that animals may also act as friends and family members. Research shows that animal companions enhance consumers' well-being through the medical and psychological benefits of pet-relating consumption experiences (American Pet Products Manufacturers Association, 2006). Holbrook et al. (2001) conclude that pets provide opportunities for deeply involving experiences; for example, to appreciate nature and wildlife, for inspiration, to be playful, to be altruistic and nurturing. Using a photo-essay approach, Holbrook et al. (2001) illustrate the warm and enduring companionship pets offer and the ways in which humans love pets, treat pets as family members, and deeply mourn pets when their lives end. Advertisers often attempt to tap this human–animal bond when they use animals in persuasive messages. 2.2. Animals and advertising Marketing promotions regularly use animals as visual symbols, ranging from the AFLAC duck to the Energizer Bunny to the Taco Bell Chihuahua (Feldhamer et al., 2002). Often, consumers' awareness of the animal link to the brand is higher than that for advertisements starring human celebrities (Hoggan, 1989). Despite their widespread use in advertising, however, little academic research is available on this topic. When it is available, such research focuses only on studies using animals as stimuli to explore hypotheses regarding classical conditioning or the effects of pictures on persuasion (Kim et al., 1998; Miniard et al., 1991). Lerner and Kalof (1999) examine the use of animals as cultural symbols in television commercials. They study six primary themes in the portrayal of animals: as loved ones, as symbols, as tools, as allegories, as nuisances, and as part of nature. Phillips (1996) hypothesizes that animal characters proliferate in advertising because animal characters transfer meanings to the brands. If most consumers associate positive cultural meanings with specific animal characters, advertisers can exploit those meanings (Phillips, 1996). Spears et al. (1996, p. 90) suggest that, when advertising uses animals, “consumers are influenced by both the symbolic meanings that have been culturally assigned to that animal as well as the physical attractiveness and likeability of how the animal is portrayed.” Their analysis of print advertisements featuring animals reveals that advertisers associate particular animals with particular product categories. 2.3. Heuristic–systematic model of persuasive communication (HSM) The heuristic–systematic model (HSM) is a dual-process model that is useful for explaining how individuals process
385
information in persuasive messages and helps clarify the influence of animal companions in advertisements on consumer attitudes toward the ad and the brand. The HSM posits that people can process information systematically or heuristically and that these are qualitatively different modes of processing (Todorov et al., 2002). According to the HSM, individuals are cognitive misers and expend effort on a cognitive task only when they have sufficient motivation and resources to do so (Chaiken and Maheswaran, 1994). Highly motivated individuals engage in systematic processing by making an effort to analyze and understand relevant information. Less motivated individuals base their decisions on simple decision rules (heuristics). These heuristic processors use mental shortcuts in lieu of engaging in issue-relevant thinking to form a judgment. Many people process messages this way, judging their validity and making decisions through the use of superficial cues such as message length, a spokesperson, or the presentation of statistical data (Griffin et al., 2002). Such heuristics often derive from experience and have empirical validity. HSM allows both systematic and heuristic processing to occur. Depending on the level of an individual's involvement with the message content, heuristic and systematic processing can proceed concurrently, or one mode may dominate. Additivity is the concurrent occurrence of heuristic and systematic processing. Additivity shows through the generation of both heuristic cue-related and attribute-related thoughts (Chaiken and Maheswaran, 1994). Empirical research notes the role of both systematic and heuristic processing as a mediator of advertising's effect on brand attitudes (see MacKenzie and Spreng, 1992 for a review). Because the HSM presents two different but intertwined routes to persuasion, HSM is important for understanding the factors that determine when brand attitudes might be influenced by consumers' diligent processing of advertising information and when they might be influenced primarily by cues that trigger heuristic processing (MacKenzie and Spreng, 1992). The study here investigates the use of animal companions as advertising cues that should induce a heuristic processing strategy. Although little research is available for guiding a heuristic view of persuasion, individuals often agree or disagree with a message on the basis of their reactions to cues such as communicator credibility and mood (Chaiken, 1980). While marketers are beginning to study the animals-andadvertising link, this area remains fertile ground for research attention. The majority of published research is qualitative in nature. Quantitative research can also add to understanding in this area. Aylesworth et al. (1999, p. 388) suggest that exploring the “effect of using animals in advertising on constructs such as attitude toward the ad, attitude toward the brand, and memorability are areas in which both researchers and practitioners should have a keen interest.” 3. Theoretical concepts and hypotheses The current study extends research on the use of animals and animal companions in ads by applying the HSM model of persuasion in order to illustrate the effects of heuristic
386
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
processing on typical measures of advertising effectiveness. The principal outcome variables of advertising effectiveness include attitude toward the ad (Aad), attitude toward the brand (Abr), and purchase intent (PI) (MacKenzie and Lutz, 1989). The HSM is useful for understanding the persuasive effects of communication variables such as source credibility and argument strength (Chaiken, 1980; Chaiken and Maheswaran, 1994). In these cases, heuristics derive from experience, and the presence of relevant heuristic cues triggers heuristics (Todorov et al., 2002). Examples of persuasion heuristics include “length implies strength” and “consensus implies correctness.” Although prior research does not use animals as a specific heuristic cue, their appearance in advertisements is often an attention-getting device not relating directly to the product being advertised. Accordingly, with other things being equal for a given campaign, the study proposes the following hypotheses: H1a. An ad with an animal cue evokes greater heuristic processing than an ad without an animal cue. H1b. An ad without an animal cue evokes greater systematic processing than an ad with an animal cue. Much research hypothesizes that involvement leads to greater perceptions of attribute differences and product importance as measures of ad effectiveness (Zaichkowsky, 1985). Involvement may affect information processing and may serve as a good gauge of motivation to process (Celsi and Olson, 1988; Zaichkowsky, 1985). Consumers with higher involvement focus more attention on processing, put greater effort into processing, and better comprehend the information an ad presents (Celsi and Olson, 1988), as they are processing systematically. Consumers with lower involvement focus less attention on processing and put less effort into processing an ad, as they process the ad heuristically. Heuristic cues such as source expertise or attractiveness have a greater impact on persuasion under low involvement as opposed to high involvement conditions (Chaiken, 1980). Therefore, the study hypothesizes as follows. H2a. Individuals with lower levels of involvement use a more heuristic and less systematic processing strategy for an ad. H2b. Individuals with higher levels of involvement use a more systematic and less heuristic processing strategy for an ad. According to the HSM, contextual cues link with the heuristic route to attitude change, and prior research notes that Aad may be formed through a heuristic process in response to ad-execution cues rather than brand-message arguments (Lord et al., 1995). Brown et al. (1998) find that conditions favoring greater cognitive elaboration produce weaker Aad than conditions not favoring elaboration. Therefore, the study proposes the following hypothesis for the advertisements containing an animal cue. H3. Systematic processors have lower Aad scores than heuristic processors. Considerable research on advertising effectiveness finds a relationship between Aad and Abr and a relationship between Abr
and PI (e.g., MacKenzie and Lutz, 1989; Muehling and McCann, 1993). Brown and Stayman's (1992) meta-analysis shows significant correlations among Aad, Abr, and PI. In order to determine whether heuristic cues such as animals have the same effects on the dependent variables, the study hypothesizes as follows. H4. Aad relates positively to attitude toward the brand. H5. Abr relates positively to purchase intent. This research also proposes that an individual's attitude toward pets may affect the interpretation and evaluation of the ad. A great deal of psychological research links pets to a general sense of well-being in humans (American Pet Products Manufacturers Association, 2006). When appearing in ad messages, animals may tap into a human–animal bond but only for consumers with positive attitudes toward pets. H6. For the ad containing an animal cue, individuals having positive attitudes toward pets have more favorable attitudes toward the ad than those who have less favorable attitudes toward pets. In summary, animal companions in ads act as positive heuristic cues on Aad resulting in positive Abr and, subsequently, higher PI. The expectation is that systematic processing of the message in the ad results in these outcomes as well, although the effect of heuristic processing likely is stronger due to the presence of the animal in the ad. In addition, their effect depends on the level of product involvement, as well as pre-existing attitudes toward pets. 4. Method 4.1. Experimental design The study employs a 2 × 3 between-respondents fixed factor design, with ad format (with animal/without animal) and involvement (high/medium/low) as the experimental factors. Participants were recruited from business classes at a large midwestern university, were given class credit for their participation in the research, and participated as one large group in a classroom setting. A total of 87 female (33%) and 166 male (64%) undergraduates participated in the study (with 7 participants not indicating gender). Their ages range from 18 to 51 years, with an average of 20years. The participants were predominantly Caucasian (73%). The primary academic majors include marketing (17%), business, (16%), management (14%), and accounting (10%). Although students represent a convenience sample, they are appropriate targets for credit cards (the stimulus product). According to Nellie Mae, 76% of college students had at least one credit card in 2004, up from 67% in 1998 (Nellie Mae, 2005). In addition, the total number of cards held by students has increased from three cards in 2000 to four or more in 2004 (Nellie Mae, 2005). Participants randomly received one of two test-booklet conditions, passed out by student administrators (who were blind to the experimental conditions) with the understanding that they would be asked to
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
evaluate the color-advertising execution in the booklet as part of a test market. Within the booklet, the participants answered pre-exposure questions relating to brand evaluations and involvement regarding the product category of credit cards. Then they were exposed to one of two ads (see Appendix A). After exposure, the participants answered questions dealing with the dependent variables and with attitude toward pets. 4.2. Choice of stimuli (advertisements) Two original ads for MasterCard were created as the stimuli. One full-page ad included a dog and the other had no animal. The dog/no-dog conditions were manipulated by presenting respondents with only one ad in their test booklet. A dog was selected because they are the most sociable of the domestic animals kept as pets in Western culture; dogs are highly amenable to sociological analysis (Sanders, 1990). Further, more U.S. households own dogs than any other pets (American Pet Products Manufacturers Association, 2006). 4.3. Measures The independent variables include ad format (with or without dog), involvement, and attitude toward pets. The main dependent variables – Aad, Abr, and PI – were measured after exposure to the stimulus. Brand evaluations and product involvement were measured prior to exposure to the ad; and attitudes toward pets, product familiarity, and demographic information (including pet ownership) were measured after exposure. Six brands (Coca-Cola, Visa, McDonald's, Pepsi, MasterCard, and Burger King) were included in pre-exposure brand evaluations in order to minimize priming effects. Independent variables. Zaichkowsky's (1985) personal involvement inventory (PII) was utilized to assess participants' involvement with the stimulus-product category. The PII is a summative index composed of 20 seven-point bipolar adjective pairs (e.g., important/unimportant, appealing/unappealing, mundane/fascinating). In the current experiment, the scale measures involvement with the product category of credit cards in general. Research on processing modalities typically measures the number of cognitions an ad generates using a thought-listing technique (MacKenzie and Spreng, 1992). As applied to the HSM, systematic processors should produce more cognitions about the ad than heuristic processors. Under the heuristic mode, participants should generate fewer thoughts, and those thoughts are less likely to be message-related and more likely to be related to other cues. Advertising cognitions were obtained by asking respondents, immediately after viewing the ad, to write down all the thoughts, reactions, ideas, and feelings they had while looking at the ad (MacKenzie and Lutz, 1989; Miniard et al., 1991). Two trained judges coded the responses into four categories, with each “thought” as the unit of analysis. Examples of messagerelevant thoughts/systematic processing are: “With MasterCard you can earn points”; “For everything else there's MasterCard”; “You earn points the more you spend”; and “The whole price-
387
less thing is good.” Animal heuristic thoughts include: “I liked the dog”; “Cute dog”; “I want a dog”; and “I miss my puppy.” Examples of other heuristic thoughts are: “That room is a mess”; “Ugly furniture”; “I thought of furniture I had”; and “Colorful room.” Examples of non-related thoughts are: “I don't want to work tonight” and “I wonder if that girl below me is single.” Twenty-five percent of the sample was chosen to establish unitizing and coding reliability, and Guetzkow's (1950) U for the unit of analysis between coders was .96. The final overall inter-coder reliability using Scott's Pi (1955) to correct for chance agreement was .93. The Pet Attitude Scale (PAS) was utilized to measure participants' attitudes toward pets with higher scores indicating a more favorable attitude toward pets (Templer et al., 1981). PAS is an 18-item Likert-type scale that has been highly reliable in previous research. Dependent variables. Attitudes toward the ad and the brand were assessed using three, seven-position semantic differential items: good/bad, pleasant/unpleasant, and favorable/unfavorable (MacKenzie and Lutz, 1989). To measure purchase intent, respondents were asked three five-point Likert-type scale items (strongly agree/strongly disagree): “It is very likely that I would use [brand] if I had one”; “I will use [brand] the next time I need to use a credit card”; and “I will definitely try [brand]” (Putrevu and Lord, 1994). 5. Results 5.1. Manipulation check The dog/no-dog conditions were manipulated by presenting respondents with either the experimental ad, which includes a dog, or the control ad without a dog. In order to confirm that the respondents in the dog-present treatment notice the dog, the following questions were asked: “In the advertisement you just saw, was there an animal in the ad?” and “If yes, please write in the kind of animal that was in the ad.” Only three respondents claimed to have seen an animal in ads that did not have one, while 100% of the respondents correctly identified a dog in their ads. Booklets with incorrect answers were removed from the data set, leaving 260 respondents.
Table 1 Mean scores and standard deviations for independent and dependent variables
Independent variables Involvement Attitude toward the pet Dependent variables Attitude toward the ad Attitude toward the brand Purchase intent Total systematic thoughts Total heuristic thoughts Total thoughts N = 260.
Scale mean
Std. dev.
a
95.58 94.16
23.80 21.21
.954 .934
13.88 13.74 9.01 1.77 2.67 4.72
4.75 4.51 2.79 1.46 1.94 2.26
.920 .936 .807 n.a. n.a. n.a.
388
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
Table 2 Pearson correlations among dependent variables
Attitude toward the ad Attitude toward the brand Purchase intent
Attitude toward the ad
Attitude toward the brand
Purchase intent
1.00 .535 ⁎⁎ .250 ⁎⁎
1.00 .510 ⁎⁎
1.00
N = 260. ⁎⁎ p b .01, one-tailed.
5.2. Initial analyses Before addressing the specific hypotheses, the scales used to measure the independent and dependent variables were checked for internal consistency and unidimensionality. Table 1 summarizes the mean, standard deviation, and Cronbach's coefficient alpha for each independent and dependent composite variable. The respondents were split to create three subsamples with high, medium, and low involvement. Respondents with involvement totals ranging from 20 to 90 were classified as low involvement; involvement totals from 91 through 108 were classed as medium involvement; and involvement totals from 109 to 140 were designated as high involvement. Unless otherwise stated, all further analyses with involvement as a variable use this three-way split. 5.3. Hypothesis testing Hypothesis 1a proposes that the presence of an animal cue in an ad leads to greater heuristic processing than occurs if no animal is present. A one-way between-respondents ANOVA on the number of heuristic thoughts indicates that respondents viewing the ad with versus without the dog generate more heuristic thoughts: Mdog = 3.11 versus Mnodog = 2.23, F(1, 258) = 14.16, p b .01. These findings support H1a. In contrast, H1b posits that ads without an animal cue evoke greater systematic processing than ads with the animal present. A oneway between-respondents ANOVA on the number of systematic thoughts indicates that respondents viewing the ad without versus with the dog generate more systematic thoughts: Mnodog = 2.00 versus Mdog = 1.56, F(1, 258) = 6.10, p b .01. These results support H1b. Hypotheses 2a and 2b suggest that involvement with the product influences the number of heuristic versus systematic thoughts, with individuals having low involvement utilizing a more heuristic strategy and individuals with high involvement utilizing a more systematic strategy. A one-way between-respondents ANOVA indicates that there is a statistically significant difference in the total number of heuristic thoughts for different levels of involvement (F(2, 236) = 4.37, p = .01). However, Scheffe's post-hoc analysis reveals that the results are not in the predicted direction. Those with medium levels of involvement have significantly fewer heuristic thoughts (M = 2.31) than those with the highest level of involvement (M = 3.18), and those with the lowest level of involvement
(M = 2.57) are not significantly different from those with either medium or high levels of involvement. These results, in fact, suggest a curvilinear relationship and do not support H2a. According to results of a separate ANOVA, although those with low involvement have fewer systematic thoughts than those with higher levels of involvement as predicted (Mlow = 1.65, Mmed = 2.05, Mhigh = 1.73), there is no statistically significant difference in the total number of systematic thoughts for different levels of involvement (F(2, 236) = 1.65, p = .194). Again, there is a slight curvilinear relationship and no support for H2b. Two-way ANOVAs for the number of heuristic and systematic thoughts, testing for both ad type and involvement, confirm the results for H1a, H1b, H2a and H2b and suggest that there are no interaction effects. For the ad with a dog, Hypothesis 3 proposes that processing strategy will impact Aad in that greater systematic processing will result in lower Aad, while greater heuristic processing will result in higher Aad. To test this hypothesis, the study separates respondents into categories (systematic, concurrent, or heuristic processors) depending on the relative balance of systematic or heuristic thoughts on the thought-listing measure. This results in three uneven groups with four respondents being classified as systematic processors, 89 as concurrent processors, and 30 as heuristic processors. A one-way between-respondents ANOVA indicates no significant differences among the three groups: Msystematic = 12.50, Mconcurrent = 14.26, Mheuristic = 13.97, F(2, 120) = .275, NS. These results do not support H3. The same analysis for the ad without a dog produces different results. In this case, systematic processors have the highest score for Aad (n = 22, M = 15.82), with concurrent processors (n = 78, M = 13.73) and heuristic processors (n = 23, M = 12.35) at lower levels (F(2,120) = 3.24, p b .05). Scheffe's post-hoc analysis shows that the difference in Aad between systematic and heuristic processors is statistically significant. In support of H4 and H5, Table 2 shows significant correlations between Aad and Abr and between Abr and PI. As Table 1 indicates, scores on the attitude-toward-pets scale are fairly high, with the mean well above the mid-point of the scale (94 versus 72). One hundred fifty-two students (58.5%) currently own a pet, and only fourteen of the participants have never owned one. Participants cite dogs most frequently as the type of pet owned. Hypothesis 6 proposes that attitude toward pets (Apet) relates to respondents' Aad for those who saw ads with a dog. The Pearson's correlation between Apet and Aad is .37 ( p b .01), supporting H6. In order to investigate the effects of attitude toward pets more thoroughly, an ANOVA was conducted for Aad with processing Table 3 Effects of processing style on attitude toward ad Source
Degrees of freedom
Mean square
F
Significance
Model a Attitude toward pet Processing style Error
4 1 3 111
5749.18 333.36 79.03 20.91
274.92 15.941 3.779
.000 .000 .013
a
Adjusted R Squared = .905.
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
type as the independent variable and Apet as a covariate. Here, we find a main effect for processing type and also a significant effect for Apet (Table 3). Concurrent processors have the highest scores for Aad (Mconcurrent = 14.16, compared with Msystematic = 12.5 and Mheuristic = 13.89).
389
search and theory. Respondents have generally high scores for involvement with credit cards. Some individuals in the low involvement group have scores above the mid-point of the scale; so there may not be enough variation among the three groups to test this variable appropriately. Thus, involvement does not influence processing strategy or Aad.
6. Discussion 6.1. Managerial implications, future research, and limitations This study illustrates the impact that executional elements in an ad have on audience processing style and subsequent attitudes toward the ad and brand. Although the findings are not as straightforward as hypothesized, they suggest that animals are popular in advertising for a reason. The two stimulus ads in the experiment have nearly identical Aad scores (Mdog = 14.08; Mnodog = 13.75; F(1, 248) = .294, NS), but the processes through which those scores are attained are quite different. In the no-dog condition, respondents generate about the same number of systematic and heuristic thoughts (2.0 versus 2.2, respectively). The majority of respondents are considered to be concurrent processors, but they are more likely to be classified as systematic processors than those in the dog condition. Systematic processors have more favorable Aad than heuristic processors for this ad. Attitude toward the ad is strongly related to Abr, which is strongly linked to purchase intent. Respondents in the dog-ad group generate significantly more heuristic than systematic thoughts (3.1 versus 1.56, respectively, t = 7.1, p b .01); thus they are more likely to be classified as heuristic processors than those in the no-dog group. The majority are concurrent processors, and there are very few systematic processors. It is interesting that this group has, on average, more total thoughts relevant to the ad than those in the no-dog group. They are not using mental shortcuts, as predicted by theory, but process the executional elements at some length. For those who saw the dog ad, processing style has no simple, direct effect on Aad, but it does have a positive influence when Apet is used as a covariate. For Aad, concurrent processors have the highest score, and systematic processors have the lowest. In other words, if readers only evaluate the message arguments, they may not find the ad particularly appealing; but if they also process the heuristic cues, then ad liking increases. The results of this research support the premise of the heuristic–systematic model that both systematic and heuristic processing can occur when people receive a persuasive message. This study hypothesizes that the presence of the dog in the ad suppresses systematic processing and increases heuristic processing, ultimately leading to increased attitude toward the ad. The results indicate an increased attitude toward the ad when respondents heuristically process the animal cues in the advertising, whether they do that exclusively or concurrently with message arguments about product attributes. Either way, Aad is strongly related to Abr, which is strongly linked to PI. Involvement with the product category does not play a role in processing for either ad, which is contrary to previous re-
When considering the objectives of the advertisement, advertisers may want to consider what type of processing they are attempting to invoke (systematic versus heuristic). In the current study, the presence of the dog increases heuristic processing, concurrent processing, and ultimately attitude toward the ad. If advertisers seek to invoke heuristic processing, the use of an animal companion in an ad may help to achieve this strategy. However, if advertisers wish to invoke systematic processing in which audience members focus on the quality of message arguments, then the use of an animal in the advertising message may derail this process. Advertisers for whom systematic processing and changes in brand beliefs are important goals need to be mindful in considering the appropriateness of utilizing animals in their advertising campaigns. When animals win ratings for Super Bowl commercials, they are often the stars of the ad. Those commercials typically have a relatively simple message. Likewise, in the current research, the dog was a focal point for the ad and the message was not very complex. However, the outcomes may not be the same if the animal companion is not the focus of the ad or if the message is a complex one. Future research should explore additional creative executions that utilize animal companions to determine the effects of variations in companion prominence and message length/ complexity. Another area of fruitful exploration concerns the congruence of the animal with the product in the advertisement. Prior research supports a strong positive effect of thematic congruence on memory for advertising and attitude toward the ad (Moorman et al., 2002). Future research should compare congruent (e.g., a dog in a dog food ad) and non-congruent (e.g., a dog in a life insurance ad) advertisements containing animals in order to determine if congruency is an important variable. The congruence issue is irrelevant, however, if animals are included in the ad simply as attention-getting devices (the “cute factor”). Attention as an outcome measure was not tested in this study. The experiment presented a single ad to the respondents in an artificial setting in order to test their reactions. If participants receive multiple ads or if they are flipping through magazines and responding to ads in a natural manner, measuring attention (recall) would make sense. This criterion is important for future studies that look at whether type of animal (popular companions or more exotic pets) makes a difference. Yet another avenue for future research relates to our measurement of involvement. Involvement is often a very difficult variable to quantify, as different researchers have noted several
390
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
types of broad involvement or involvement as a multidimensional construct with several underlying dimensions (Broderick and Mueller, 1999; Laurent and Kapferer, 1985) — including normative, enduring, situational, and hedonic involvement. In our study, involvement was assessed through the use of Zaichkowsky's (1985) involvement with the product category scale. Future research should consider other measures of involvement, as different types of involvement (e.g., hedonic versus normative) might correspond with more/less systematic or heuristic processing and might ultimately lead to differences in the processing of the animal cue in the advertising message. The current research is subject to the usual limitations of experimental research. First, the use of college students as participants might be a concern that limits the generalizability of results. However, undergraduates are especially appropriate for this research because they comprise a homogeneous target market for marketers. Future research should attempt to investigate perceptions of animals in advertising messages among non-student consumer groups. Another limitation of this research relates to the choice of the stimuli in the experiment. Although the ads were experimentally designed, the brand (MasterCard) and its overall message strategy (“Priceless”) are familiar to students. When using real brands, the researcher runs the risk that participants have preexisting knowledge, attitudes, and beliefs about the brand that will affect their responses. Obviously, the results of this study are specific to the advertising stimuli used and cannot be generalized to all credit-card advertisements. Future research
Appendix A. Advertising stimuli
should investigate whether the results hold across other product categories. Another important limitation is the fact that the ads in the treatment and control manipulations were similar in format and message, but not in visual cues. This is appropriate in that it mirrors common real-world advertising campaigns, but is problematic in that the differences between the two conditions (not using advertisements identical in every way except for the animal cue) may have confounded the results and should be recognized as potential rival explanations for the pattern of data. While we attempted to control for extraneous variables that might confound our experiment, we erred on the side of “practical application” utilized by practitioners in an effort to achieve a “unified field theory” (Carlson et al., 2005) that walks the line between theoretical study and practical application. 7. Conclusion Using animals is a common practice for advertising campaigns today, despite the “hassles” and costs associated with working with animals. The accepted belief is that animals in advertisements inspire good feelings about the advertisement and the brand. The current research supports the anecdotal evidence that the use of animals in advertising has positive effects, especially when combined with a favorable attitude toward pets, but it also suggests that non-animal ad executions can be equally effective.
K.M. Lancendorfer et al. / Journal of Business Research 61 (2008) 384–391
References The American Pet Products Manufacturers Association. Pet industry spending figures; 2006 [March 23], http://www.appma.org. Aylesworth A, Chapman K, Dobscha S. Animal companions and marketing: dogs are more than just a cell in the BCG matrix! Adv Consum Res 1999;26: 385–91. Belk RW. Possessions and the extended self. J Consum Res 1988;15:139–68 [September]. Broderick AJ, Mueller RD. A theoretical and empirical exegesis of the consumer involvement construct. J Mark Theory Prac 1999;7(4):97–108. Brown SP, Homer PM, Inman JJ. A meta-analysis of relationships between adevoked feeling and advertising responses. J Mark Res 1998;35(1):114–26. Brown SP, Stayman DM. Antecedents and consequences of attitude toward the ad: a meta-analysis. J Consum Res 1992;19(1):34–51. Carlson L, Grove SJ, Stafford MR. Perspectives on advertising research. J Advert 2005;34(2):117–48. Celsi R, Olson JC. The role of involvement in attention and comprehension processes. J Consum Res 1988;15:210–24 [September]. Chaiken S. Heuristic versus systematic information processing and the use of source versus message cues in persuasion. J Pers Soc Psychol 1980;39 (5):752–66. Chaiken S, Maheswaran D. Heuristic processing can bias systematic processing: effects of source credibility, argument ambiguity, and task importance on attitude judgment. J Pers Soc Psychol 1994;66:460–73 [March]. Croke V. It's a bull market for animals in ads. Austin American-Statesman; 1992. p. D9. [December 5]. Dale S. Animal smart. USA Weekend 2005; July 10. Feldhamer G, Whittaker J, Monty A, Weickert C. Charismatic mammalian megafauna: public empathy and marketing strategy. J Pop Cult 2002;36 (1):160–8. Griffin RJ, Neuwirth K, Giese J, Dunwoody S. Linking the heuristic–systematic model and depth of processing. Communic Res 2002;29(6):705–32. Guetzkow H. Unitizing and categorizing problems in coding qualitative data. J Clin Psychol 1950;6:47–58. Hirschman EC. Consumers and their animal companions. J Consum Res 1994;20 (March):616–32. Hoggan K. Four legs good, two legs bad. Marketing 1989; March 2: 1. Holbrook MB, Stephens DL, Day E, Holbrook SM, Strazar G. A collective stereographic photo essay on key aspects of animal companionship: the truth about dogs and cats. Acad Market Sci Rev 2001:1–17. Horovitz B. Smile! You're the stars of the Super ad Bowl: America loves to laugh at ads with comical critters. USA Today 2003a; January 24. Horovitz B. Anheuser-Busch's ‘Replay’ is top Super Bowl ad. USA Today 2003b; January 27. Horovitz B. Viewers' favorite ads crude, rude and furry. USA Today 2004; February 2.
391
Horovitz B. Anheuser-Busch still king of ads as marketers play it safe. USA Today 2005; February 7. Horovitz B. ‘Magic fridge’ of Bud Light ices an advertising win. USA Today 2006; February 5. Kim J, Lim J, Bhargava M. The role of affect in attitude formation: a classical conditioning approach. J Acad Mark Sci 1998;26(2):143–52. Laurent G, Kapferer J. Measuring consumer involvement profiles. J Mark Res 1985;22(1):41–53. Lerner JE, Kalof L. The animal text: message and meaning in television advertisements. Sociol Q 1999;40(4):565–86. Lord K, Lee M, Sauer PL. The combined influence hypothesis: central and peripheral antecedents of attitude toward the ad. J Advert 1995;24(1):73–85. MacKenzie SB, Lutz RJ. An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. J Mark 1989;53:48–65 [April]. MacKenzie SB, Spreng RA. How does motivation moderate the impact of central and peripheral processing of brand attitudes and intentions? J Consum Res 1992;18(4):519–29. Miniard PW, Bhatla S, Lord KR, Dickson PR, Unnava HR. Picture-based persuasion processes and the moderating role of involvement. J Consum Res 1991;18(1):92–107. Moorman M, Neijens PC, Smith EG. The effects of magazine-induced psychological responses and thematic congruence on memory and attitude toward the ad in a real-life setting. J Advert 2002;31(4):27–40. Muehling DD, McCann M. Attitude toward the ad: a review. J Curr Issues Res Advert 1993;15(2):25–58. Nellie Mae. Undergraduate students and credit cards in 2004; 2005 [May], www.nelliemae.com. Phillips BJ. Advertising and the cultural meaning of animals. Adv Consum Res 1996;23:354–60. Putrevu S, Lord KR. Comparative and noncomparative advertising: attitudinal effects under cognitive and affective involvement conditions. J Advert 1994;23(2):77–90. Sanders CR. The animal ‘other’: self-definition, social identity and companion animals. Adv Consum Res 1990;17:662–8. Scott W. Reliability of content analysis: the case of nominal scale coding. Public Opin Q 1955;19:321–5. Spears NE, Mowen JC, Chakraborty G. Symbolic role of animals in print advertising: content analysis and conceptual development. J Bus Res 1996;37:87–95. Templer DI, Slater CA, Dickey S, Baldwin R, Veleber DM. The construction of a pet attitude scale. Psychol Rec 1981;31:343–8. Todorov A, Chaiken S, Henderson MD. The heuristic–systematic model of social information processing. In: Dillard JP, Pfau M, editors. The persuasion handbook. Thousand Oaks, CA: Sage Publications; 2002. p. 195–211. Zaichkowsky JL. Measuring the involvement construct. J Consum Res 1985;12:341–52 [December].