Jongmin Park, Lisa Lyon and Glen T. Cameron
[WJMCR 4:1 December 2000]
Sections:
Abstract|Introduction|Analysis and Results|Conclusion
Abstract
Claims are made for the importance of corporate reputation as essential to the effective, integrated marketing of a company’s branded products. Based on the Elaboration Likelihood and Combined-Effects Models of persuasion theory, an experiment was conducted to examine the value of one tool in corporate reputation management�the corporate ad or corporate image ad. Using path analysis, findings indicate that the corporate reputation ad had a greater impact on purchase intention under low involvement conditions than under high involvement conditions. Subjects who are unmotivated seem to rely more heavily on reputation information when making a purchase decision. Notably, the reputation ad had no indirect impact in mediating the effects of attitude toward the brand and the brand ad on purchase intention. Reputation effects in this study impacted only in a direct path upon purchase decision. The Combined-Effects Model showed greater promise than the Elaboration Likelihood Model for further research on reputation and strategic communication campaign planning.
Introduction
Purpose of the Study. Historically, advertising in the United States has been oriented toward the product and the brand. With the advent of modern corporations that offer many products and services, new types of advertising, such as corporate advertising and institutional advertising, have appeared. Numerous articles have described the ways in which corporate advertising has fostered positive public attitudes toward businesses.1 Corporate advertising is used to establish, alter or maintain a corporation’s identity. In fact, much of corporate advertising serves as a support function to promote a corporation’s character and personality to consumers. In effect, corporate advertising is a key tool in developing and maintaining the corporation’s reputation.
Quite obviously, both corporate ads and brand ads are subject to many of the same dynamics of interest to information processing researchers. Using the Elaboration Likelihood Model and the combined-influence hypothesis as a framework, this study will examine central versus peripheral route processing by measuring consumers’ attitudes toward a brand, a brand ad and a corporate ad.
Several studies, such as MacKenzie, Lutz, and Belch2, Droge3, MacKenzie and Spreng4, and Brown and Stayman5 have found that consumers’ attitudes toward ads are formed using the peripheral route. Other studies, such as Burton and Lichenstein6, Muehling and Laczniak7, MacKenzie and Lutz8, Homer9, Miniard, Bhatla, and Rose10, and Mick11, however, have indicated that the central-route is used to form consumers’ attitudes toward the ad. Together, these studies indicate that, because subjects’ attitudes toward ads significantly affect brand attitude and purchase intention for highly involved subjects, both the central and peripheral route can be used simultaneously. Lord, Lee and Sauer12 proposed this combined-influence hypothesis, which is supported by their research.
Corporate reputation, built in part on corporate image advertising, is based on how the company conducts or is perceived as conducting its business.13 Corporate reputation is a set of attributes ascribed to a firm, inferred from the firm’s past actions, about which one infers that reliability is an attribute of this person or company. By doing so, one makes judgments about past information and uses these signals to form beliefs in predicting future actions.14 Likewise, corporate image is like a picture in a consumer’s mind about a corporation’s character and personality.15 This study defines corporate reputation advertising as promoting a corporation’s character and personality to consumers, and brand advertising as promoting a corporation’s brand to consumers.
Advertising for Corporate Reputation. In today’s corporate world, everything from toothpaste to internet browsers has become a commodity. Frequently, there is little or no distinction between prices, technologies, or product capabilities. A company’s reputation, therefore, can be the overriding basis for a consumer’s purchasing decision. Whether a company is ethical and honest, environmentally friendly, and responsive to customers’ needs can be a company’s most powerful asset or its costliest liability.16 Reputational effects are purported to impact everything from stock values of the company to employee morale to the effectiveness of brand advertising, as well as attitudes toward the brand itself.
Traditionally, lack of a widely accepted measure of reputation has caused difficulty in creating well-reasoned and defensible answers about corporate reputation and reputational dynamics.17 Only recently, however, has formal research outlined some of the strategic planning implications behind corporate reputation. For example, Hammond, Annis & Slocum found that corporate reputation is linked with a firm’s bottom-line financial performance. Investors may consider less socially responsible organizations riskier investments because of possible governmental intervention. Additionally, if a firm is viewed as socially rsponsible it may have a relatively low financial risk as a result of its strong relationship with the surrounding community. Weigelt and Camerer18 further the notion that bottom-line financial success is linked with corporate reputation by stating that reputation is an asset vital to an organization’s financial success. They also provide evidence suggesting that the reputation-financial performance effect is a two-way process: a firm’s financial performance affects its reputation and its reputation affects its performance.
Morley19 suggests that in today’s global business atmosphere, a solid corporate reputation will increase customer loyalty to the products of companies with good reputations. A good reputation allows for easier product introduction because customers feel that a “guarantee” comes with it. Bromley20 even suggests that a good image or reputation can allow a company not only to implement its present plans, but also to pursue its goals in the future. This has been described as the”halo effect”�where a generally positive attitude toward the company lends the company to immunity to a certain extent. Goldberg and Hartwick21 found evidence of this in an experiment investigating the combined effects of a company’s reputation and advertisements on product evaluations. Subjects who formed a negative evaluation of the company based on a bad reputation found the advertisements less credible and rated the products less favorably than those who received positive reputation information about the company.
While managing a corporate reputation involves many factors, research suggests that advertising has been successful in promoting corporate reputations and company position statements to various audiences.22 Corporate advertising can help better position American products against the competition, to meet increased pressure from consumer groups and politicians, and to repair the corporate reputations of American companies that are criticized for their roles in creating adverse environmental conditions. Schumann, Hathcote, and West23 compiled a summary of the corporate image ad studies in Table 1. Almost all of the studies reported ratings of recognition, image and overall corporate reputation. They said “As business attempts to respond, corporate advertising is expected to promote images of environmentally responsible corporate citizens.”
The Brand Image and Corporate Image. Haedrich24 identified two levels of images: brand and corporate. He said that, in the advertising of brand image, “corporations address the extent to which each marketing instrument contributes to the brand image and the extent to which that image enhances desired economic behavior of markets.” Corporate images, however, consist of not only brand images, but also other factors such as the quality of management, corporate leadership and employee orientation.
Frombrun and Van Reil contend that in marketing research, “reputation,” often labeled brand image focuses on the nature of information processing, resulting in “pictures in the heads” of external subjects, attributing cognitive and affective meaning to cues received about an object they were directly or indirectly confronted with.25 Objects in advertising and marketing research are usually products (such as beer, toothpaste, internet browsers), while consumers are the principal subject of analyses. However, Page contends that to be useful, the lessons learned from this marketing view of reputation must be expanded to move beyond just the creation of brand equity to increase “share of mind.”26 Among the determinants of “share of mind” is reputation: the confidence one has that a product baring a particular brand will live up to the producer’s name.
Druckenmiller stated that a corporate reputation must be given continual attention because “it must carry a heavier load than corporate brands.” Good corporate reputations are the sum of many parts, including high quality products and services, years of recognition equity invested in a familiar logo, a well-known CEO, sponsorship of a high-profile event, and memorable advertising.27
Using factor-analysis, Winters28 categorized corporate image into three components: business conduct/marketing factors; social conduct/marketing factors; and contributions factors. He found that “marketing image is an important component of overall attitude toward a company. Corporate advertising messages should be aimed at improving the social conduct image of the company, (e.g., environmental image). Contribution messages would have the least impact on overall attitude toward the company.” Business conduct/marketing image or reputation is the most important component of corporate image in predicting overall favorability toward the corporation.
A study by Frombrun and Shanley,29 investigating the factors that influence corporate reputation found that audiences (publics) construct reputations on the basis of three main things. First, publics consider information about the firm’s relative structural positions within the organizational fields, specifically using market and accounting signals indicating performance. Second, institutional signals indicating conformity to social norms are considered. And third, strategy signals indicating strategic postures influence how publics construct reputation.
Corporations can harmonize their brand and corporate images by way of strategic advertising campaigns. Corporations can likewise bring their perceived images into harmony with those of all relevant populations. Haedrich insisted that “to harmonize these images, the planning of marketing and public relations should be integrated at the level of strategic management.”30 Corporate leadership can meet the overlapping demands of the market and society by integrating marketing, advertising, and public relations under the common roof of strategic corporate planning and making them equal instruments of management.
Theoretical Background
The Elaboration Likelihood Model. The Elaboration Likelihood Model was developed within many fields of study, such as psychological counseling, and consumers’ attitudes in advertising and has received solid empirical support. The ELM considers varying degrees of subject or client interest as critical variables in determining how attitudes are formed and what types of attitudes are formed.31Petty and Cacioppo32 insisted that:
Enduring attitude change, then, appears to depend on the likelihood that an issue or argument will be elaborated upon (thought about). It doesn’t make sense for a person to think carefully about every message received daily. Most of the messages that we receive, in fact, are on issues that are relatively trivial, and it is not worth our time and energy to scrutinize them carefully.
Reviews of attitude change study that measure persistence33 also support the view that the active cognitive involvement of a person in the persuasion situation is crucial for enduring attitude changes.
The Elaboration Likelihood Model posits two basic routes of attitude change. The central route is taken when persuasion results from thinking about the issue or argument under consideration. The peripheral route is taken when persuasion comes from non-issue-relevant concern, or cues, such as impression management, the attractiveness of the message’s source, or one’s social role. This Elaboration Likelihood Model holds that consumers who are both motivated (personally concerned) and able to cognitively process the content of a persuasive message are likely to be persuaded by the central route. Consumers who are not motivated and able to think about the persuasive message are likely to be persuaded by the peripheral route.
Petty and Cacioppo34 have found that if the new attitudes of consumers results from the central route (issue-relevant cognitive activity), attitude change is likely to be relatively enduring. However, if the new attitude of consumers results from the peripheral route (various persuasion cues in the situation), the attitude is likely to exist only as long as the cues remain salient.
The Central Route versus The Peripheral Route. The central route involves a subject’s personal thoughts about a messages and his own evaluation of the arguments presented. Attitude change using the central route is more difficult to achieve than attitude change using the peripheral route. However, attitude change using the central route lasts longer and influences subjects’ and consumers’ behaviors more powerfully.
Petty, Cacioppo and Schumann35 noted that:
Clearly, the elaboration-likelihood model indicates that it is quite difficult to produce an enduring attitude change by exposing people to a persuasive communication. The recipient of the message must have both the motivation and the ability to process the information contained in the communication, and the information presented must elicit favorable cognitive responses that are rehearsed and stored in long-term memory.
On the peripheral route, subjects select, from the persuasion context, some cue associated favorably or unfavorably with the attitude topic. When the cue is no longer present or recalled, the previous attitudes are said to be elastic. Attitude changes using the peripheral route are easily produced in the laboratory but do not influence a subject’s behavior as strongly as the central route.
Low and High Involvement. This Elaboration Likelihood Model holds that when involvement is high, people are likely to be persuaded by the central route. When involvement is low, however, people are likely to be persuaded by the peripheral route. Petty, Cacioppo and Schumann36 noted the relationship between the degree of personal relevance and issue-relevant cognitive activity. They found that:
In studies where issue-relevant cognitive activity was likely to be intense (e.g., role-playing studies, experiments employing personally relevant issues, etc.), the attitude changes produced have been found to be relatively enduring. On the other hand, in studies where issue-relevant cognitive activity was likely to be weak (e.g., experiments employing issues of little personal relevance), the initial attitude changes produced have been relatively short-lived.
In one study, Petty, Cacioppo and Schumann studied the impact of the two routes using a print advertisement for a fictitious disposable razor. They found that the quality of the arguments (strong versus weak message claims about the product) had a strong impact on product evaluation under high involvement conditions and a weaker, but still significant, effect under low involvement conditions. In contrast, endorser manipulation (celebrities versus average citizens) influenced product attitude only for those subjects less involved during message processing.37
The Peripheral Cue Hypothesis and the Attitude Toward Ads. Using the Petty and Cacioppo’s Elaboration Likelihood Model, Lutz, MacKenzie, and Belch38examined the way in which the attitude toward an ad mediates brand attitude and purchase intention by testing four alternative hypotheses. Their results indicated the attitude toward an ad is a peripheral process on the part of their subjects.
MacKenzie and Spreng39 examined the role of central and peripheral processes in driving brand attitude and purchase intention under varying levels of motivation. They assumed the attitude toward an ad operated exclusively as a peripheral route. Specifically, they observed that “increasing motivation decreases the impact of a peripheral cue (the attitude toward the ad) on brand attitude.” Their hypotheses were supported. The results indicated that increasing the motivation to evaluate an advertised brand has no effect on brand cognitions, but it increases the impact of brand attitudes on purchase intentions by directly strengthening the attitude-intention relation and increasing brand attitudes.
Droge40 suggested that the attitude toward an ad operates as a peripheral route. Droge’s modeling of the impact of the attitude toward the ad provided results consistent with that explanation.
The Combined Influence Hypothesis of Attitude toward the Ad. Several studies; such as MacKenzie, Lutz, and Belch, Droge, MacKenzie and Spreng and Brown and Stayman; supported the assertion that consumers’ attitudes toward ads operate as a peripheral route. However, other studies; such as Burton and Lichenstein, MacKenzie and Lutz, Homer, Miniard, Bhatla, and Rose and Mick; have hinted at a substantive role of central-route processing in attitudes toward ads.
Muehling and Laczniak41 examined the moderating role of involvement in the relationship of brand attitude with brand beliefs and the attitude toward ads. They found that the attitude toward ads was shown to significantly affect brand attitude, and it also affected brand beliefs for highly involved subjects. Among subjects who had low involvement, the role of beliefs faded to non-significance, although the attitude toward the ad remained significant.
They pointed out that, because subjects’ attitudes toward the ad was shown to significantly affect brand attitude and purchase intention for highly involved subjects, attitudes toward ads can consist of central and peripheral components.
In contrast with the peripheral-cue hypothesis, Lord, Lee and Sauer42 proposed the combined-influence hypothesis. According to this hypothesis, consumer attitudes toward ads are a function of consumer response to both message arguments and peripheral cues. Their experiment followed a 2 (argument strength in the ad) x 2 (peripheral cue, the liked music versus the disliked music) x 2 (number of exposures) x 2 (involvement) treatment design. The results supported the combined-influence hypothesis. Consumer attitudes toward ads were found to be directly influenced by the consumers’ responses to and evaluation of message arguments in addition to peripheral cues in the ad or program context. Consumer attitudes toward ads also exert a consistently significant direct impact on purchase intentions and an indirect influence that is mediated by brand attitudes.
Lord, Lee and Sauer called for future research to examine alternative peripheral cues including humor, vocal pattern, testimonials and mood or image in the context of the combined-effects model. The current study partially fulfills this call.
Research Hypotheses and Path Analysis Design
In this study, consumers’ attitudes toward two kinds of ads, corporate reputation and brand ads, were studied. Subjects using the peripheral route associate the advocated position not with the issue itself but with secondary sources, such as an expert, an attractive source or a powerful source. Petty and Cacioppo43 described the peripheral route:
Although all of these secondary techniques have been successful in producing attitude changes, the elaboration-likelihood model suggests that all of these changes will not be very permanent. Furthermore, these techniques are not likely to be very successful in changing people’s attitudes when they have a lot of prior information about the issue or if the issue is very involving to them. When people have a lot of prior information about an issue and the issue has personal relevance, they will be motivated to process the issue-relevant information presented, and peripheral aspects of the persuasion situation will be less important.
A corporate reputation ad serves as a peripheral cue by providing source information while the brand ad is central. Based upon the Elaboration Likelihood Model, four hypotheses are generated. H1. Attitude toward the brand will have a greater impact on purchase intention under high involvement conditions than under low involvement conditions.
H2. Attitude toward the corporate reputation ad will have a greater impact on attitude toward the brand ad under low involvement conditions than under high involvement conditions.
H3. Attitude toward the corporate reputation ad will have a greater impact on attitude toward the brand under low involvement conditions than under high involvement conditions.
H4. Attitude toward the corporate reputation ad will have a greater impact on purchase intention under low involvement conditions than under high involvement conditions.
In addition, the Combined Effects Model proposes that attitude toward the brand ad serves a central one. Based upon the CEM, two hypotheses are generated.
H5. Attitude toward the brand ad will be shown to exert a direct influence on purchase intention independent of the level of the involvement.
H6. Attitude toward the brand ad will be shown to exert a direct influence on attitude toward the brand independent of the level of the involvement.
As Figure 1 indicates, the exogenous variable in this study is the attitude toward the corporate reputation ad. The endogenous variables are attitude toward the brand ad, attitude toward the brand, and the purchase intention. Both kinds of variables are measured by the response to a questionnaire.
Methodology
Subjects. Using mall intercepts in a mid-western city, a pilot study was done in April 1998. Fifty respondents (half were men, half were women) were asked to rate the questionnaire. Respondents found several problems in the questionnaires and offered suggestions. Some changes were made based on their advice.
A total of 295 male and female students enrolled in an undergraduate course participated in the study to earn extra credit. Two different ad booklets were prepared for the study. The first contained the ads in the high involvement conditions and the second contained the ads in the low involvement conditions.
Advertisement Samples and Measures. Subjects were assigned randomly to the treatment conditions. Two ad booklets contained the same ad samples (corporate reputation ad and brand ad) and questionnaires and two different cover pages that introduced the cereal ad differed in the high and low involvement conditions. Subjects saw the corporate reputation ad first and were instructed to indicate their attitudes. Then subjects also saw the brand ad and indicated their attitudes. On the last page of the questionnaire, subjects’ attitudes toward the brand, as well as purchase intentions, were measured. Upon completion of the questionnaire, subjects were thoroughly debriefed and thanked for their participation.
After collecting the data, frequencies were analyzed to determine the subjects’ demographic characteristics and media and advertisement consumption. Next, t-tests were used to compare the levels of involvement. Then, multiple regression analysis was conducted to find path coefficients and path routes among the variables. Finally, two-way analysis of variance was performed to examine the research hypotheses of this study.
In order to provide a structure that integrated traditional theories and Learn-Feel-Do hierarchy models with consumer involvement and brain specialization theories, Vaughn44 introduced a new Foote, Cone, and Belding approach to advertising strategy depicted in Table 2. Based on this model most food and packaged staple goods belong to the third group: Low Involvement/Thinking (Habit Formation).
Product decisions in this group involve minimal thought and a tendency to form buying habits based on convenience. Also, consumers of food/household items have several “acceptable” brands, and brand loyalty is a function of habit. Therefore, this group has a Do-Learn-Feel pattern. This study focused specifically on print advertisements for cereal, an appropriate product to use to study consumer attitudes toward a brand and corporate reputation to examine how these factors affect purchase intention. Also, because consumers have preconceptions about special brands and corporate reputations of products, such as Kellogg’s, the fictitious names (Crown for a company name) and Good Morning Mates as a brand name were used.
Measures developed by Lord, Lee, and Sauer were used in this study. Attitude toward the brand ad was measured by three seven-point semantic differential items anchored with good/bad, pleasant/unpleasant and favorable/unfavorable. General attitude toward the brand ad was measured by the same systems in this study. The intercorrelation among these measures was very high (average r = 0.8092). The same scales were used to assess the attitude toward the corporate reputation ad, which also yielded a high intercorrelation (average r= 0.8209). To identify the attitude toward the brand, favorability was used in a seven-point semantic differential measure.
To adjust for the possibility of response differences between intentions and estimates45, purchase likelihood was measured with both an intention item (a ten-point semantic differential of very likely/very unlikely) and an estimation question (“probability that you will purchase this cereal,” ranging from 1=0% to 10 =100%). The correlation between the two measures was 0.76.
Petty, Cacioppo, and Schuman’s46 involvement measure was adapted to bolster the involvement manipulation for this study. Involvement was embedded in two places in the ad booklet. First, before subjects were asked their attitudes toward ads and the brand, two different groups of subjects saw two different cover pages respectively, and these cover pages that introduced the cereal ad differed in the high and low involvement conditions. High involvement subjects were told that this survey and research were sponsored by the Crown Cereal Company and the Advertising Research Center of the sponsoring university. They were told the advertisement and product would soon be test-marketed in medium-sized cities throughout the Midwest, including the study site. Low involvement subjects were told that the advertisement and product were being test-marketed only on the East Coast. Thus, high involvement subjects were led to believe that the product would be available in their area in the near future, and low involvement subjects were led to believe that this cereal would not be available for purchase in their area in the foreseeable future.
Second, on the last part of the questionnaire, involvement also was assessed through the use of a seven-point semantic differential item: “while viewing the ads, I was: very involved/very uninvolved.” One hundred and fifty subjects saw the cover page in high involvement conditions, and 145 subjects saw the cover page in low involvement conditions. However, twenty of the 150 subjects who saw the cover page in high involvement conditions indicated low involvement leading to exclusion of 20 questionnaires that were not under the expected involvement condition. Fourteen of the 145 subjects who saw the cover page in the low involvement condition indicated high involvement and thus were also excluded. Finally, 261 questionnaires were used for this study.
Analysis and Results
Demographics and Manipulation Checks for Involvement. A sample of 261 valid subjects was used. One hundred thirty-three (51%) of the subjects were male and 127 subjects (48.7%) were female. One hundred and fourteen subjects (43.7%) said that they read magazines two or three times a week, 133 subjects (51%) read a newspaper 1 hour or less every day or two or three times a week and 109 subjects (41.8%) watched TV more than 1 hour every day. One hundred and fifty-six subjects (59.8%) responded that they sometimes enjoy seeing advertising. Ninety-four subjects (36%) answered that they always enjoyed seeing advertising on TV, magazines and newspapers.
To measure the difference in subjects’ attitudes according to the level of involvement, questions regarding their attitudes toward the corporate reputation ad, the brand ad, the brand, and purchase intention were included. Attitude scores on the corporate reputation ad and the brand ad represent the average rating of the three seven-point semantic differential scales that are anchored at bad and good, unpleasant and pleasant, and unfavorable and favorable.
Attitudes differed as a function of involvement. Subjects under high involvement (mean = 5.5) showed more positive attitudes toward the corporate reputation ad than subjects under low involvement (mean = 4.8; t = 1.37 d.f = 259 p = .0001 < .05 ). Subjects with low involvement (mean = 3.5929) showed more positive attitudes toward the brand ad than subjects under high involvement (mean = 3.1564; t = -2.42 d.f = 259 p = .016 < .05 ). Subjects with low involvement (mean = 3.7939) showed more positive attitudes toward the brand than subjects under high involvement (mean = 3.3154; t = -3.11 d.f = 230 p = .002 < .05 ). Subjects with low involvement (mean = 3.8702) showed more positive attitudes toward the purchase intention than subjects under high involvement (mean = 2.8154; t = -3.99 d.f = 259 p = .0001 < .05 ).
Interestingly, as can be seen from Q1 of Figure 2, subjects with high involvement showed more positive attitudes than did subjects under low involvement only on the corporate reputation ad. However, these results have no significance in this study because the subjects’ attitudes toward ads are affected by multiple variables within the context of ad, such as the endorser, the argument and mood. For the variable of interest, different levels of involvement showed different attitudes to each question about the corporate reputation ad, the brand ad and the brand.
To verify the research hypotheses of this study, therefore, the relationships between variables will be examined in the next section.
Path analysis. Path analysis was employed to determine the causal effects of the research variables in this study. It is used to articulate the causal direction and causal time ordering of dependent variables in relation to each other and to the independent variables. Path analysis is useful in making explicit the rationale of conventional regression calculations. Although path analysis is not a method for discovering causes, it is an analytical technique that can be applied to a causal model formulated by researchers on the basis of knowledge and theoretical considerations.
Generally, exogenous and endogenous variables are combined in a path analysis to identify significant paths corresponding to the hypothesized relations. In Figure 3(see Table 3), the attitude toward the corporate reputation ad is shown as an exogenous variable. The relationships among the exogenous variables remain unspecified and are characterized by standardized partial regression coefficients. The other factors; attitude toward the brand ad, attitude toward the brand and purchase intention are considered endogenous variables because their variation is determined by other variables in the model. Therefore, exogenous variables are treated as independent factors and the endogenous variables are regarded as dependent variables.
Each arrow in the model bears a path coefficient that represents the direct effect of the antecedent variable on the dependent factor. In Figure 3, the path coefficient (p1) is <.05, indicating that the attitude toward the corporate reputation ad and attitude toward the brand ad are mainly due to indirect effects. The attitude toward the corporate reputation ad has a weak direct effect (-.08) on the attitude toward the brand. The direct effect of the attitude toward the corporate reputation ad on purchase intention is .115, whereas the total of indirect effects is .023 (.115�.092). In other words, the attitude toward the corporate reputation ad has practically a more direct effect than an indirect one on purchase intention.
Therefore, the observations regarding these variables lead to the conclusion that the present model can be trimmed. The consistency of the path model can be increased by excluding the variables which have weak direct effects.47 The more parsimonious model is presented in Figure 4 (see Table 4). The attitude toward the corporate reputation ad in this model, however, has no statistically significant effect on purchase intention (R Square=.0084, Signif F=.13191). The attitude toward the corporate reputation ad has to be excluded in this model.
Figure 4 shows causal relationships between the research variables and specifies causal ordering. In this model, the attitude toward the brand ad is an exogenous variable and the attitude toward the brand and purchase intention are considered endogenous variables. In this case, the attitude toward the brand ad affects the attitude toward the brand (path coefficient=.633) and purchase intention (path coefficient=.3423). This exogenous variable has a positive relationship with the variables. Also, the attitude toward the brand affects purchase intention (path coefficient =.4848). That is, there is a positive causal relationship between two endogenous variables. This path model draws the actual path coefficients calculated from the study data. The zero-order correlations of variables from which the path coefficients were calculated are shown in Table 5.
The path model under high involvement. The path coefficients under high involvement do not show much difference from the path coefficients (Figure 5; see Table 6) without controlling involvement. An exogenous variable, the attitude toward the corporate reputation ad, has weak direct effects on the other variables, including the attitude toward the brand ad (path coefficient =.019), the attitude toward the brand (path coefficient=-.118) and purchase intention (path coefficient=.07). Therefore, the attitude toward the corporate reputation ad is excluded in the more parsimonious model.
The more parsimonious path model is presented in Figure 6 (see Table 7), which shows casual relationships among the research variables and specifies causal ordering. In this model, the attitude toward the brand ad is an exogenous variable and the attitude toward the brand and purchase intention are endogenous variables. In these relationships, the attitude toward the brand ad highly affects the attitude toward the brand (path coefficient=.686) and purchase intention (path coefficient=.274) is also affected by it. This exogenous variable has a positive relationship to the other variables. Also, the attitude toward the brand affects purchase intention (path coefficient=.6018). That is, there is a positive causal relationship between the two endogenous variables. The zero-order correlations of variables from which the path coefficients were calculated can be seen in Table 8.
The path model under low involvement. As an exogenous variable, the attitude toward the corporate reputation ad under low involvement has weak direct effects on the other two variables: the attitude toward the brand ad (path coefficient =.071) and the attitude toward the brand (path coefficient =.013). However, the direct effect of attitude toward the corporate reputation ad on purchase intention is .203, and the total of indirect effects is .044 (Figure 7; see Table 9). In other words, the attitude toward the corporate reputation ad has a more direct effect than an indirect one on purchase intention. Therefore, the path model under low involvement will be trimmed. The more parsimonious model is presented in Figure 8 (see Table 10).
In this model, the exogenous variable, which is the attitude toward the corporate reputation ad, has no direct effect on the attitude toward the brand ad and the attitude toward the brand. But it has a direct effect on purchase intention. Therefore, this exogenous variable changes to a residual variable. “Residual variables are introduced to indicate the effect of variables not included in the model. It is assumed that a residual variable is neither correlated with other residuals nor with other variables in the system.”48 This variable has a statistically significant direct effect on purchase intention (R Square=.0612, Signif F=.0044) but has to be excluded in this model because this variable has no direct effects on the other variables in the model.
The most parsimonious path model is presented in Figure 8, which shows causal relationships among the research variables. In this model, the attitude toward the brand ad is an exogenous variable and the attitude toward the brand and purchase intention are endogenous variables. In these relationships, the attitude toward the brand ad affects the attitude toward the brand (path coefficient=.539) and purchase intention (path coefficient=.401). This exogenous variable has a positive relationship with the variables. Also, the attitude toward the brand affects purchase intention (path coefficient=.331). That is, there is a positive causal relationship between the two endogenous variables. The zero-order correlations of variables from which the path coefficients were calculated can be seen in Table 11.
Test of Hypotheses. In the preceding section, the path models were examined in order to find the relationships between the research variables. Using the path models, tests of the research hypotheses are examined in this section. The attitude toward the corporate reputation ad and the attitude toward the brand ad represent the average rating of the product on the three seven-point semantic differential scales anchored at good-bad, pleasant-unpleasant, and favorable-unfavorable.
Hypothesis 1 states: Attitude toward the brand will have a greater impact on purchase intention under high involvement conditions than under low involvement conditions. The path coefficients between the attitude toward the brand and purchase intention are .602 under high involvement and .331 under low involvement (See Figure 6 and Figure 8). Therefore, hypothesis 1 is not supported. That is, the attitude toward the brand has a significant impact on purchase intention regardless of the level of the involvement.
Hypothesis 2 states: Attitude toward the corporate reputation ad will have a greater impact on attitude toward the brand ad under low involvement conditions than under high involvement conditions. The path coefficients between the attitude toward the corporate reputation ad and the attitude toward the brand ad are .019 under high involvement and .071 under low involvement (See Figure 5 and Figure 7). Therefore, hypothesis 2 is not supported. That is, the attitude toward the corporate reputation ad has no significant impact on the attitude toward the brand ad regardless of the level of involvement.
Hypothesis 3 states: Attitude toward the corporate reputation ad will have a greater impact on attitude toward the brand under low involvement conditions than under high involvement conditions. The path coefficients between the attitude toward the corporate reputation ad and the attitude toward the brand are -.118 under high involvement and .013 under low involvement (See Figure 5 and Figure 7). Therefore, hypothesis 3 is not supported. That is, the attitude toward the corporate reputation ad has no significant impact on the attitude toward the brand regardless the level of the involvement.
Hypothesis 4 states: Attitude toward the corporate reputation ad will have a greater impact on purchase intention under low involvement conditions than under high involvement conditions. The path coefficients between the attitude toward the corporate reputation ad and purchase intention are .07 under high involvement and .247 under low involvement (See Figure 5 and Figure 7). Therefore, Hypothesis 4 is supported. That is, the attitude toward the corporate reputation ad has a greater impact on purchase intention under low involvement conditions compared to high involvement conditions.
Hypothesis 5 states: The attitude toward the brand ad will be shown to exert a direct influence on purchase intention independent of the level of the involvement The path coefficients between the attitude toward the brand and purchase intention are .274 under high involvement and .401 under low involvement (See Figure 6and Figure 8). Therefore, hypothesis 5 is supported. That is, the attitude toward the brand ad has a significant impact on purchase intention regardless of the level of the involvement.
Hypothesis 6 states: The attitude toward the brand ad will be shown to exert a direct influence on attitude toward the brand independent of the level of the involvement. The path coefficients between the attitude toward the brand ad and the attitude toward the brand are .686 under high involvement and .539 under low involvement (See Figure 6 and Figure 8). Therefore, hypothesis 6 is supported. That is, the attitude toward the brand ad has a significant impact on the attitude toward the brand regardless the level of the involvement.
Conclusion
Implications of the study. The purpose of this study was to verify hypotheses based on the Elaboration Likelihood Model and the Combined-Effects Model by taking several central and peripheral routes such as the attitude toward the brand, the attitude toward the brand ad, and the attitude toward the corporate reputation ad, all of which can influence consumers’ purchase intentions.
Currently, much of corporate advertising plays an important role in the promotion of a company’s products and services. Many researchers disagree about whether consumers’ attitude toward ads has a consistently significant direct impact on purchase intention. These are reasons why the above three variables were chosen as the independent variables for this study.
The application of the Elaboration Likelihood Model in this study was not found perfectly appropriate in measuring the effects of advertisements on purchase intention. Hypotheses 1, 2 and 3 which were based on the Elaboration Likelihood Model were not supported. That is, the attitude toward the brand had a significant impact on purchase intention regardless of the level of involvement, and the attitude toward the corporate reputation ad had no significant impact on the attitude toward the brand ad and the brand, again regardless of the level of involvement.
However, hypothesis 4 was supported. That is, the attitude toward the corporate reputation ad had a greater impact on purchase intention under low involvement conditions than under high involvement conditions. In addition, under low involvement, consumer’s attitude toward the corporate reputation ad had a statistically significant direct effect on purchase intention.
This evidence indicates that strategic use of corporate advertising as a means of managing reputation could not only strengthen consumers’ image of the organizations, but may also have a direct impact on corporate bottom-line performance under certain conditions. Subjects who are perhaps unmotivated and/or unable to cognitively process the content of a persuasive message seem to rely more heavily on reputation information when making a purchasing decision. Results of this study show the direct, main effect a reputational factor can have on a behavioral intention measure, bearing out platitudes about bottom-line importance of reputation management. This direct effect of reputation on purchase intention supports a study by Lyon and Cameron49 that found subjects were generally more likely to invest in and hold more favorable attitudes about companies with a good reputation over companies with a bad reputation. The results of this study, however, help better define the circumstances and conditions under which subjects will use reputation information in a consumer, decision-making scenario. This information can help corporations attempting to harmonize their brand and corporate images and then bring their perceived images into harmony with those of all relevant populations, by more accurately defining those populations.50
According to the results, two hypotheses based on the combined effects model were supported. In other words, the attitude toward the brand ad has a significant impact on the attitude toward the brand and purchase intention regardless of the level of involvement.
Finally, the results of this study indicate that the attitudes toward the brand ad are shown to exert a consistently significant direct impact on purchase intention in addition to its indirect influence as mediated by the attitude toward the brand, regardless of the level of involvement. The attitude toward the corporate reputation ad, however, was shown to exert no consistently significant direct impact on the attitude toward the brand, regardless of its indirect influence mediated by the attitude toward the brand ad, or the level of involvement. Although the attitude toward the corporate reputation ad had no indirect impact in mediating the effects of attitude toward the brand and the brand ad on purchase intention, it had a direct impact on purchase intention under low involvement conditions rather than under high involvement conditions.
Limitations and Suggestions. Additional research is needed to replicate and expand upon these findings. In this study, the corporate reputation ad, functioning as a peripheral cue, had significant effects such as a direct impact on purchase intention under low involvement conditions that should be tested under other conditions to build upon the results of this study because subjects were affected by various factors, including stories, pictures and testimonials, within the context of each ad. Also, additional studies on relations between consumers’ attitudes toward corporations based on reputation and purchase intention are necessary.
Finally, the combined-influence hypothesis toward the brand ad is supported in this study. Conceptual and empirical research is encouraged to examine the implications of the Combined Effects Model and its relationship to reputation research and integrated communication campaign efforts.
About the Authors:
Jongmin Park is an assistant professor at Pusan National University; Lisa Lyon is a doctoral student in the Henry W. Grady College of Journalism and Mass Communication at the University of Georgia; and Glen T. Cameron is a professor at the School of Journalism, University of Missouri. This paper was presented at the Annual Conference of Association for Education in Journalism and Mass Communication, New Orleans, LA, August 1999. Correspondence concerning this manuscript should be addressed to Jongmin Park, Assistant Professor, Department of Communication, College of Social Sciences, Pusan National University, San 30, Changjun-Dong, Kumjeoung-gu, Pusan, Korea, 609-735 office phone: 011-82-51-510-2149, home phone: 011-82-51-515-8665, office fax: 011-82-51-512-0945, Internet: jongmin2000@hotmail.com