Compounding consumer interest: effects of advertising campaign publicity on the ability to...

It is not uncommon that an advertising campaign in and of itself becomes sufficiently remarkable to attract media coverage. Major metropolitan daily newspapers feature regular columns on advertising. The New York Times publishes about 700 advertising news stories per year, almost two items per

Advertising professionals recognize that news coverage about an ad can augment media campaign expenditures, potentially building expectations and heightening awareness of the ad (Harris 1998; "No PR, No Comment" 1995). Advertising as news adds value to an ad campaign because it is mentioned in a nonadvertising context rather than being thrown into the cluttered commercial pod (Rothenberg 1990). Yates (1995) noted that a common element among some successful brands is that their ad campaigns hit the headlines. Ries and Ries (1999, p. 42) describe the value of this preadvertising publicity thus: "publicity is the nail, advertising is the hammer." Harris (1998) labels it "value added public relations."

Although numerous examples about the effects of publicity on ad campaigns have appeared in various publications, empirical evidence is scant. Thorson and Moore (1996) argue that it is important and necessary to understand how cross-promotional tools (e.g., advertising, public relations, promotions, etc.) work together and relate to each other. The current study provides a timely, empirical examination of the synergetic effects of combining the promotional tools of publicity and advertising. The following research questions were addressed: (1) Does publicity of an advertising campaign improve consumers' memory of the subsequent advertisements? (2) What mechanism accounts for the impact of publicity?

INTEGRATED MARKETING COMMUNICATION (IMC) AS A CROSS-PROMOTIONAL TOOL

This exploratory investigation is particularly relevant in light of increasing interest in integrated marketing communication (IMC), which seeks to better understand and utilize the full array of tools in the promotional mix (Schultz, Tannenbaum, and Lauterborn 1992; Thorson and Moore 1996). Considerable discussion of IMC has been in progress, but academicians and practitioners appear to have developed their own concept of IMC (for details, see Duncan and Caywood 1996). However, a common ground regarding "What is IMC?" has surfaced in academic circles. IMC is "strategic coordination of multiple communication voices" and "pursuing synergy by integration" (Moriarty 1996; Schultz, Tannenbaum, and Lauterborn 1992; Thorson and Moore 1996).

By addressing the key elements of IMC, Thorson and Moore (1996) and Lutz (1996) proposed a general research agenda for IMC: How do cross-media and/or cross-promotional tools work together? There have been a small number of studies on cross-media in advertising campaigns. Edell and Keller (1989) examined media interactions in an advertising campaign employing television and radio to understand how advertising campaigns should be coordinated across media. It has been reported that print advertising can enhance the effectiveness of television ads when both ads are well coordinated (Confer 1992; Confer and McGlathery 1991).

As Thorson and Moore (1996) pointed out, however, there have been few studies of the combination of different promotional tools that keep the IMC framework in mind. Most studies that consider multiple promotional tools simultaneously have been found in the area of mathematical modeling research; these neither measured processing variables per se nor investigated integrated effects (e.g., Ailawadi, Farris, and Parry 1997; Balasubramanian and Kumar 1990).

It is reasonable to believe that the IMC perspective can provide advertising researchers with a more integrated way of thinking rather than breaking down fields or functions into increasingly narrow, specialized pockets. The present study is not an attempt to see the whole picture in one glance; rather, it is an attempt to look at two promotional tools at once--advertising and publicity--using traditional advertising effectiveness measures.

ADVERTISING AS NEWS: AVAILABILITY OF INFORMATION ABOUT AD CAMPAIGNS

Previous research provides baseline information on the extent of news coverage of advertising. News stories about advertising are fairly common on network newscasts. Keenan (1995) examined the full census of stories of three television network newscasts indexed under the heading "advertising" from 1970 through 1993. There were a total of 1,068 stories dealing with advertising, which represented an average of 44.5 per year, or nearly one every week. It was reported that more than 3,000 news items about various advertising campaigns were found in four major metropolitan daily newspapers over a three-year span. Major newspapers frequently report on the day-to-day business of advertising. The mere launch of a new campaign is often deemed newsworthy, and the topic of product advertising was frequently covered in the news media (Pasadeos, Phelps, and Lamme 2000).

Information generated by advertisers about planned campaigns and products is called "preannouncement" behavior. This is defined as "a formal, deliberate communication behavior before a firm actually undertakes a particular marketing action such as a price change, a new advertising campaign, or a product line change" (Eliashberg and Robertson 1988, p. 282). Such preannouncements often result in free publicity from the news media. Not every action by advertisers to get publicity results in media coverage, however; information provided must have some news value.

The amount of news coverage a preannouncement receives can be partly explained by the professional practices of journalism. News selection is a complex process that reflects various values, goals, and process factors. Shoemaker and Reese (1996) point out that newsworthiness involves the presence and prevalence of certain news values, specifically timeliness, proximity, prominence, conflict, oddity, and human interest. These values influence news selection by media gatekeepers. Traditional news values can in part explain why certain types of advertising campaign information or advertisements draw more media attention than others do. In the advertising context, Pasadeos, Phelps, and Lamme (2000) suggest two major categories of information that impact the newsworthiness of advertising: items that are either "important" or "interesting."

Reporting about political advertising, for example, satisfies the "important" criterion. Many journalists believe it is their Fourth Estate duty to provide a picture of political reality. Thus, analysis of political campaigns, including analysis of political ads, is considered crucial to the ideal of informed democratic choice (see, e.g., Tedesco, McKinnon, and Kaid 1996). Advertising thought to be deceptive also falls in this category. In reporting on deceptive ads, the media are fulfilling their watchdog role by reporting on misleading information in advertisements that might result in harm to consumers (Pasadeos, Phelps, and Lamme 2000).

On the other hand, Super Bowl advertising, in particular, has been the subject of a substantial number of news stories. This trend began with Apple Computer's startling commercial during the 1984 Super Bowl, which generated extensive publicity and introduced the concept of advertising as a news event (Johnson 1994). Recognition of "Bud Bowl" ads in Super Bowl broadcasts has benefited from substantial publicity and benefited the company by establishing a strong brand connection with the Super Bowl (Fitzgerald 2001).

The Super Bowl is arguably the most visible advertising event of the year. Many new commercials debut during the game, and the stratospheric network price tags for Super Bowl commercial time alone satisfy the "interesting" newsworthiness criterion. Indeed, cost of placing the ads is routinely the subject of news stories. Super Bowl commercials have even become a competition among advertisers--the equivalent of the Oscars for the movie business. USA Today, for example, evaluates all the commercials aired during the game through an annual national survey and announces which selections the public judged to be the best and worst. Jin and Zhao's (1999) empirical findings suggest that a substantial amount of news deals with Super Bowl ads before, and shortly after, the Super Bowl game. Therefore, timeliness is also a factor affecting which advertising campaigns become newsworthy.

THE PROPOSED MODEL

Figure 1 represents a proposed model illustrating how preexposure to news about ads may affect consumers' information processing of subsequent ads. Effects of advertising news on ad memory were specified through a dual path model.

[FIGURE 1 OMITTED]

One path illustrates the motivational process--the way that news on ads stimulates active engagement with the ads, which will, in turn, improve memory of those ads. The other path suggests the repetition process--the way that memory derived from reading the news about the ads directly influences memory of the ads after being exposed to them. As the model suggests, the reader's perceived interest in Super Bowl advertising and ad attention are two mediators.

Motivational Process

News stories about Super Bowl advertising are a kind of entertainment news, similar to the way movie and drama critics provide entertainment news. Information provided by critics about entertainment products often stimulates consumers' curiosity (Eliashberg and Shugan 1997). Likewise, news information about advertising, especially in the case of Super Bowl advertising, can be viewed as an advertising event announcement that depicts who the participants are, what the creative ideas are, who the celebrities are, and how much money a 30-second commercial costs in the Super Bowl.

Edell and Keller (1989) conceptualized a model of media interaction to provide an information-processing perspective on a coordinated media campaign that employed both television and radio. They applied the concept of "curiosity" by presenting the audio track of a television ad on the radio followed by the full ad on television. They reasoned that "people who have heard the radio ad first may be more curious about what the accompanying video looks like than people who view the TV ad without having heard the radio ad. This curiosity may motivate greater processing of the video during TV ad exposure" (p. 151).

Similarly, news coverage on poll information during elections (so-called horse-race polls) triggers the public's curiosity about the election, thus increasing public interest in the election and enhancing knowledge about political issues (Zhao and Bleske 1998). In an experiment, Zhao and Bleske (1998) had subjects read a story about political issues. Subjects in the experimental group were given horse-race information that outlined which candidate was leading in the polls. The horse-race information was deleted for subjects in the control group. The main body of the story was identical for the two groups. They found that subjects given horse-race poll information showed a better understanding of the issues. They reasoned that "poll information may lead to increased exposure and attention to the media channels with issue content, which in turn lead to better understanding about candidates' issue position" (p. 30).

Advertising practitioners support the idea of curiosity effects, noting that news enhances ad interest, builds expectation, and heightens awareness for a new ad (Harris 1998; "No PR, No Comment" 1995; Ries and Ries 1999). This curiosity factor was applied to the present study, which leads to the following hypothesis:

   H1: Prior exposure to a news story about ads will have a
   positive effect on ad interest. Thus, people who read the news
   story will have a higher ad interest than those who do not read
   the news story.

Information-processing models (see, e.g., Greenwald and Leavitt 1984) assume that the attention allocated to an ad is a function of consumers' involvement. The conceptualization and measurement of involvement has long been a controversial topic, due to the use of the term in the varied research domains of social psychology, political science, consumer behavior, advertising, and mass communication. Whereas most involvement studies in the advertising and consumer behavior literature have focused on product involvement or advertising message involvement, more emphasis has been given to issue (e.g., political issues) or event (e.g, election) involvement in the mass communication and political communication literature. For example, some researchers have conducted studies of the effectiveness of communication using involvement operationalized as interest, that is, event or issue (e.g., Salmon et al. 1985).

Despite differences in the actual operationalization of involvement, there seems to be consensus that interest-involvement plays a key role in information processing by activating a heightened state of arousal with greater cognitive activity (Salmon 1986) and active information-search and acquisition strategies (see, e.g., Buchholz and Smith 1991; Burnkrant and Sawyer 1983; Greenwald and Leavitt 1984; Zaichkowsky 1986). Although a universally accepted definition of involvement has not yet surfaced, this study considers the involvement of Super Bowl advertising as the amount of event interest. The basic logic is identical in any context, however. When people have interest in ads, they are likely to pay attention to the ads and gain more information about them. Thus, the proposed model posits that perceived ad interest is a trigger for ad attention and, in turn, the ad attention increases recall of the subsequent ads. This leads to the following hypotheses:

   H2: Ad interest will have a positive effect on ad attention.
   Thus, people who are more interested in the ads will have higher
   ad attention.

   H3: Ad attention will have a positive effect on ad memory.
   Thus, people who pay more attention to the ads will have higher
   ad memory.

Repetition Process

Considering the stage of encoding of the news, reading a news story normally requires active participation by the audience because reading printed words is relatively demanding cognitive work. Print improves memory because reading involves self-initiated processing necessary to elaborate on the information provided in text (Frieske and Park 1999). Thus, some facts in the story (e.g., the topic, brand names, ad execution tactics, etc.) are thought to be available (Lynch and Srull 1982) when a reader comprehends and decodes the news. The stored information can then be retrieved and used as cues for the information being tested as suggested in the associative models of memory (e.g., Raaijmakers and Shiffrin 1981).

Arguably, the availability of cues depends on the degree of message repetition. In addition, the encoding variability that occurs when information is observed from two perspectives accounts for an enhanced memory effect (Melton 1970). During the interval between repetitions, the context of the target information changes and the information is associated with diverse contexts. This enhances memory owing to more retrieval paths used than in a single presentation (Young and Bellezza 1982). When a person first reads a news story about ads and then watches the ads later, encoding variability beyond the message repetition effect occurs because information is gleaned from two perspectives--the news story and the ad itself. This phenomenon has been used to explain the fact that the repetition of similar but nonidentical ads often enhances message recall more than repetition of identical ads (Unnava and Burnkrant 1991; Young and Bellezza 1982).

The encoding process of people reading news is one component that enhances their memory of subsequent ads relative to those who did not read the news; the other component is the retrieval process. During memory tests in which a person is asked to remember ads seen during a Super Bowl, one may start by recalling the game-viewing situation. Although some aspect of the recently viewed ads may be unique, distinguishing among ads seen during this event and other television-viewing situations may not be clear. At some point, the person may recall the news article mentioning the Super Bowl advertising and remember some brand ads. Pham and Johar (1997) called this process "memory-trace refreshment," which "entails an attempt to revive (bring to consciousness) the original learning episode with its perceptual and contextual details" (p. 250). Recalling the news stories, the person may correctly identify some brands from the Super Bowl ads. A person who did not read a news story beforehand has no such retrieval information, and thus remains confused about the jumble of ads seen during the game. Considering these cognitive processes, the following hypothesis was developed:

   H4: Prior exposure to a news story about ads will have a
   positive effect on subsequent ad memory. Thus, people who read
   the news story will have a higher subsequent ad memory than
   those who do not read the news story.

It is possible that the length of information regarding the ads presented in the news story will produce differential recall effects among brand ads if multiple ads are embedded in the story. Greater detail in the news story about a particular ad would increase the depth of information processing when reading the story, which in turn could be a critical determinant of subsequent recall performance. In contrast, shorter stories in the news are less likely to create a strong cue in the reader's mind. Although people would likely retrieve the original learning episode with its perceptual and contextual details, they would have a higher chance of failing to recall brand names because of the relatively weak encoding process involved. Thus, the availability of cues and the strength of associations (see, e.g., Raaijmakers and Shiffrin 1981) created by reading the news might be dependent upon the amount of information presented regarding each brand ad in the news. This leads to the following hypothesis:

   H5: The story length of each brand ad in a news story about
   ads will produce a differential effect on subsequent ad memory.
   Thus, the longer the story about an ad, the greater the
   differences will be in the ability to recall the ad and its
   associated brand.

METHOD

Procedure

The specific goal of the experiment was to examine the relationship between prior news exposure about upcoming advertisements in the Super Bowl and memory of the subsequent ads. The treatment--exposure to news stories--was conducted under controlled experimental conditions; however, viewing the Super Bowl game and its attendant commercials were not controlled (natural setting).

Experiments were administered in five sessions. All sessions were conducted in regular university classes. To ensure that testing procedures in all the sessions were as similar as possible, the proctors followed detailed written directions. The treatment (exposure to news articles) was conducted the Friday (January 28, 2000) before the Sunday night Super Bowl game. The second visit to the classes was made on the Monday (January 31, 2000) immediately following the game. To increase the participation rate, lottery prizes were given in each visit.

During the Friday session, the proctor thanked students for participating and told them that the purpose of the study was to examine attitudes toward news stories. The real purpose of the study was not revealed. Participants were instructed when to begin, when to stop, and when to turn the page. Subjects were also instructed not to talk during the experiment. After all the subjects had received the booklet, the proctor asked them to write down the last four digits of their student identification numbers on the first page of the questionnaire as an entry for lottery prizes. The proctor then asked the participants to open the test booklet to the second page and read the news story. After two minutes, participants were told to stop and turn the page. Pretesting showed that students needed approximately two minutes to read the story. Subjects were randomly divided into two groups. Subjects in the experimental group were given a news story about selected Super Bowl ads, whereas subjects in the control group read a news article that did not mention any Super Bowl ads. In the first section of the questionnaire, attitudes toward the news story were measured. Super Bowl ad event interest and demographic information followed. Administration of the Friday test required approximately 20 minutes.

Participants were not told that an additional survey including memory tests on the Super Bowl ads would be conducted after the Super Bowl game. This was to ensure that attention to the advertisements during the Super Bowl game would take place in a natural way. On Monday after the game, the proctor revisited the class. The questionnaire booklet consisted of Questionnaires A and B. Subjects were asked to go to Questionnaire A if they watched any part of the Super Bowl game. Questionnaire B was prepared for the non-Super Bowl viewers.

Both questionnaires had the same number of pages. In the first section of Questionnaire A, respondents were asked to report whether they watched the game and, if so, what portion they watched. In the second section, respondents were asked to list brand names they could remember seeing during the Super Bowl broadcast. Ad recall and other measures followed the advertised brand recall test. Monday and Friday measures were matched together based on student ID numbers.

Subjects

Participants were recruited from three universities: two were in the southeastern United States and one was in the northeastern United States. As a readily accessible large population, students were considered appropriate participants for the purpose of this study, which examined the causal relationship between news exposure and ad memory. This study required students to complete questionnaires two times--before (treatment) and after the game (memory tests). On the Friday before the Super Bowl game, which is played on a Sunday, 623 students completed a questionnaire. For the memory tests on Monday, a day after the game, a total of 551 students responded. There was not any systematic difference between groups for the attrition rates (experimental group: 19% versus control group: 17%).

Target Stimuli

The focus items were two single-page layouts (8.5 x 11 inch), constructed to look like magazine articles. The stimulus news story for the experimental group was modified from actual news articles about Super Bowl ads that appeared in USA Today, the New York Times, and the Wall Street Journal. The story consisted of a headline and body, with no pictures imbedded. The news story for the experimental group included descriptions of advertisements for five different brands. Information about each brand's ad was subtitled over one to three paragraphs of text. In addition, without detailed descriptions (only brand names), the news story mentioned six other brand names. All brand names and advertisements were supposed to appear in the Super Bowl broadcast (see Appendix). The news story for the control group did not have any information about the Super Bowl ads.

Dependent Variables

Super Bowl ad interest (INTEREST) was measured by three items with seven-point scales: (1) "How much interest do you have in commercials that will appear during the Super Bowl broadcast?" (not very much/very much), (2) "How much do you want to see the Super Bowl advertisements?" (not very much/very much), and (3) "I'm very curious what the Super Bowl advertisements will be this year" (strongly disagree/ strongly agree).

Because the Super Bowl viewing situation was not controlled, ad attention could not be measured while subjects were watching television. Hence, the use of physiological responses or eye-movement tracking, which are available in tightly controlled lab experiments, was not feasible in this exploratory study. Instead, self-reported ad attention (ATTENTION) measured situational attention and was modified from television advertising avoidance studies (Speck and Elliot 1997). Three items with seven-point scales were: (1) "I did not want to leave the room during the commercial breaks because I did not want to miss the ads" (strongly disagree/ strongly agree), (2) "I did not want to change the channel during the commercial break because I wanted to watch the ads" (strongly agree/strongly disagree), and (3) "How much attention did you pay to the ads during the Super Bowl broadcast?" (not very much/very much).

Ad Memory (MEMORY) had two indicators: brand recall and ad recall. For the advertised brand recall measure (B-RECALL), unaided free recall tests were performed by directing subjects to a blank sheet and asking them to list as many brand names as they could recall from the Super Bowl (Pavelchak, Antil, and Munch 1988). The number of brands correctly recalled was counted. For the ad recall (AD-RECALL), subjects were instructed to write down everything they could remember about the ads themselves (e.g., description of the ads, characters, story line, or theme). AD-RECALL scores were calculated by counting the number of ad components correctly recalled (Edell and Keller 1989).

ANALYSES AND RESULTS

Screening Data

Responses to specific questions were included if the respondent answered at least half of the items on the survey; none of the responses were used in cases in which fewer than half of the items on the questionnaire were answered. Because the goal of the study involved news effects, it was reasonable to analyze a data set in which all respondents were exposed to the experimental stimulus only. Thus, subjects who reported they had been exposed to other news stories about the Super Bowl ads before the memory test were not included in the data analysis. All subjects in the final sample watched at least a part of the game and completed both survey questionnaires. It is important to note that it was assumed that the subjects in the control group had not heard, read, or watched any news coverage about the ads during the experiment, whereas the subjects in the experimental group were exposed to only the experimental stimulus news story. This left a final sample of 169 subjects for analysis. The final sample was well balanced in terms of gender (48% male, 52% female). A majority of the sample was between the ages of 18 and 22 (96.9%), which was expected given that they were college students. Subjects' academic majors varied, also as expected. No significant differences were noted in the dependent variables among the three different schools (Wilks's [lambda] = .91; F = .94; p > .48).

False Alarm Tests

For the recall measures, subjects' falsely recalled items should be examined to avoid the misinterpretation of the data beforehand. Hence, two independent coders analyzed all Super Bowl ads using videotapes recorded during the game, and the number of correctly recalled items was counted. The correlation coefficients for intercoder reliability on the two recall measures (B-RECALL and AD-RECALL) exceeded .9, which was deemed acceptable. To see if any of the recalls may have been a false alarm, the coders searched for incorrectly recalled items. A total of 17 brand names (correct: 656; incorrect: 17) were falsely recalled, amounting to about 2% of total items recalled. False alarms were rarely found for AD-RECALL, with only one false recall. These results suggest that false alarms should not significantly bias the memory variables.

Hypotheses Testing

Structural equation modeling (SEM) was used to test the proposed model and the relevant hypotheses. Before testing the structural model, a measurement model with three constructs (INTEREST, ATTENTION, and MEMORY) was tested. The results suggested the model was fairly adequate ([chi square] = 25.9, df = 17, p > .07; GFI [goodness-of-fit index] = .96; AGFI [adjusted goodness-of-fit index] = .92; CFI [comparative fit index] = .99). Item reliability coefficient alpha was in the acceptable range for INTEREST ([alpha] = .94), ATTENTION ([alpha] = .94), and MEMORY ([alpha] = .69). All factor loadings for constructs were significant. Thus, the data supported convergent validity. For discriminant validity, a two-standard error interval estimate of each coefficient was calculated to examine whether one (1) was within the interval. Data indicated that interval estimates for any coefficients did not include 1, providing supporting evidence for discriminant validity.

Table 1 presents the results of the structural model. The proposed model showed good model fit. Chi-square and other model fit indices were within acceptable ranges ([chi square] = 34.5, df = 23,p > .05; GFI = .95; AGFI = .91; CFI = .99; RMSEA [root mean square error of approximation] = .05). All structural path coefficients were significant. The paths, NEWS [right arrow] INTEREST (H1) and NEWS [right arrow] MEMORY (H4) were significant at .05 [alpha] level, indicating that the news exposure group had significantly higher ad interest and ad memory than did the control group. Hypothesis 2 predicted that INTEREST would have a positive influence on ATTENTION. As indicated in Table 1, the effect of INTEREST on ATTENTION was significant and in the hypothesized direction (p < .001). Thus, subjects who were more interested in the ads were more likely to be attentive to the ads. Hypothesis 3 proposed that ATTENTION would have a positive effect on MEMORY. The results demonstrated a strong positive relationship between the two constructs (p < .001), indicating that subjects who paid more attention to the ads had higher memory scores. Additional analysis regarding the path NEWS [right arrow] ATTENTION was not significant (p > .2), suggesting that news reading did not directly affect ATTENTION; that is, news reading only appeared to affect ATTENTION via INTEREST. The structural model summary is presented in Figure 2.

[FIGURE 2 OMITTED]

To test whether there were differences in the effects on each brand in the news story (H5), all recalled items of B-RECALL and AD-RECALL were recoded based on brand names. Eleven brand categories for each memory measure were created. Each brand name was coded 1 if it was recalled and 0 if it was not. Then each brand was evaluated using a chi-square test. Results (see Table 2) showed that the largest group difference for both B-RECALL ([chi square] = 33.4; p < .001) and AD-RECALL ([chi square] = 11.4; p < .01) was observed in Tropicana, which with 27 lines had the largest coverage in the story. EDS, which had the second-largest coverage with 12 lines in the news, had the second-largest effect for both recall measures (B-RECALL: [chi square] = 10.75;p < .01, and AD-RECALL: [chi square] = 6.49;p < .05). Two other brands (7 UP and Motorola), which had similar length in each story, but shorter coverage than the previous two brands, showed similar effects on B-RECALL, but effects were much less than those of the longer stories (7 UP: [chi square] = 6.49; p < .05, and Motorola: [chi square] = 5.52;p < .05).

There was one exception to the general pattern of larger effects with longer coverage. No group difference for either memory measure was found for "Bud" (Budweiser beer) (B-RECALL: [chi square] = .32; p > .1, and AD-RECALL: [chi square] = 1.54; p > . 1). This exception likely occurred because Bud has been well known through the popular "Bud Bowl" from past Super Bowl advertising. Thus, the Bud ad in the news might not have provided any new information to the subjects. In addition, the Bud ads appeared eight times during the game--by far the most frequently shown ad compared with other brand ads (see the ad frequency in Table 2). The higher frequency of the ad could have attenuated the news effects.

Effects were weaker for the six brands that mentioned only brand names in the news story, although significant group differences were found for two brands (Autotrade and Pet) in B-RECALL. None of the brands showed any significant group differences in AD-RECALL. In sum, two distinct patterns emerged from the analysis: (1) stronger news effects were found in the brands that were described in more detail, and (2) news effects on B-RECALL were more significant than AD-RECALL. Yet these findings were inconclusive due to some exceptions and potential limitations, which will be addressed in the Discussion and Conclusions section below.

Additional Analysis

Treatment effects were further analyzed with the inclusion of those who did not watch the Super Bowl. This provided an opportunity to better understand the effects of the news story and depth of coverage by comparing which brands nonwatchers would expect to appear in Super Bowl ads. Subjects who read the experimental news story or the control news story but did not watch the game were asked to list the brand names that they thought appeared during the Super Bowl broadcast.

Table 3 shows the results of ANOVA (analysis of variance) post hoc rests. Some distinct patterns were observed. For the five brands described in more detail in the news story, those who read news and saw the ads outperformed all others in B-RECALL. Those who read about the ads in the news but did not watch the Super Bowl recalled ads equally effectively as those who did not read the experimental news story but watched the Super Bowl. No treatment effects were observed for brands that appeared in the news but lacked detailed descriptions, or for other brand ads that were not mentioned in the news. It appears, then, that watching the Super Bowl telecast and reading the news story improved ad recall over either factor alone. These results also show that it is likely that news memory was available, but only for the brands with detailed descriptions.

DISCUSSION AND CONCLUSIONS

Summary of Findings and Implications

Marketing practitioners commonly assume that ad effectiveness could be improved substantially by preannouncement of an advertising campaign (Harris 1998; "No PR, No Comment" 1995; Yates 1995). This assumption was empirically tested by examining the effects of Super Bowl advertising campaign information in a news story on consumers' memory of the subsequent ads. Effects were examined with a dual processing model: reading news stories about ads affects the subsequent ad memory through (1) a motivational process and (2) a repetition process.

The motivational process posits that news about incoming advertising campaigns increases subsequent ad memory through active involvement in the advertising event (Edell and Keller 1989; Zhao and Bleske 1998). The data supported this process. The news exposure group had significantly greater interest in the upcoming Super Bowl ads than did the control group. Results of the present study also show positive causal relationships in ad memory through the motivational process: interest in the ads [right arrow] attention to the ads [right arrow] subsequent memory of the ads.

The key element of the repetition process is that memory derived from reading a news story about ads directly influences memory after exposure to the ads. In this study, the recall of brand ad information of subjects who read the news story but did not watch the Super Bowl was similar to those who watched the game but were not exposed to the news story. Those who both read the news story and watched the ads had the best brand ad recall. These findings support the benefits of the message repetition process. The better ad memory of subjects who read the news and watched the ads could result from stored information in long-term memory from reading the story, which was reinforced through repeat exposure to the ads during the Super Bowl. Thus, it is likely that both message repetition and motivational processes were involved in recall of Super Bowl ads.

Reading the news story did not boost recall of brands or ads that were not mentioned in the story but that aired during the Super Bowl. Thus, it is likely that the news story created greater interest, leading to greater recall, for ads covered in depth. However, ads not mentioned in the news appeared to be of less interest to subjects, and thus were not recalled as well. When some salient information in the news story is strongly encoded, part-list cueing effects could emerge (Alba and Chattopadhyay 1986). Hence, the retrieval process could strengthen recall of the salient brand ads mentioned in the news but weaken recall of other brand ads. The part-list cueing effects may explain why the brand ads that were described in more detail were recalled more frequently and why the news story did not increase memory of other brand ads that were not shown in the news.

Beyond the aggregated memory measures, which were based on the total number of correctly recalled items as in the structural equation modeling, additional analyses for individual brands in the news story were conducted. Although the group differences for ad memory were somewhat positively related to the amount of coverage of each brand in the news story, the finding is inconclusive due to some exceptions and potential limitations.

This study suggests an effective strategy in marketing communication. Using preannouncement or publicity to kick off an advertising campaign can enhance both the efficiency and impact of the ad campaign. Preannouncement could be an important first step in developing a total marketing package, especially when considering the high price tag of television commercials, particularly Super Bowl advertising. Prospective buyers of Super Bowl ads should consider the opportunity for substantial ad campaign publicity from Super Bowl ads, which typical ads can scarcely obtain, as well as other criteria in selecting media and vehicles (e.g., audience size and CPM [cost per thousand]). Also, advertising practitioners should consider the value of publicity, even in typical ad campaigns, and utilize public relations methods to preannounce campaigns. However, it is important that practitioners also provide some news value--be it interest or importance-to maximize campaign publicity.

The most frequently recalled brand and ad by the subjects was Budweiser ads. It is interesting to note that there was no group difference for Bud ads regarding memory measures. This exception likely occurred because Bud was already well known through the popular "Bud Bowl" Super Bowl advertising campaign and the significant publicity regarding it (Newell and Henderson 1998). This prior knowledge of the long-established connection between the brand and the Super Bowl probably did not provide other unique information to the subjects in the experimental group. This kind of established connection highlights the importance of long-term marketing and advertising campaigns.

Limitations and Future Research

The research design for the present study has limitations that should be addressed. The natural setting of Super Bowl viewing was designed to measure the memory of commercials within the subjects' normal television-viewing habits. This approach provides the advantage of ecological validity over controlled experiments by using real news stories, real commercials, a real audience, and a real event. There were some trade-offs, however. Super Bowl advertising is a highly visible event and a special case within the advertising environment. Some people are willing to watch the ads even though they are not interested in the game. In most news stories about advertising campaigns, people may not know when the ads from the news stories would be seen. On the other hand, readers or viewers of news about Super Bowl ads know exactly when they will be seen. This may result in anticipation of or heightened interest in the ads, which may improve recall. Yet it is unclear how preadvertising news affects information processing of the subsequent ad when the ad was shown in an attentive versus an inattentive situation. In this sense, the present study did not examine the more typical television commercial exposure, that is, the so-called low-involvement learning process (Krugman 1965). Thus, the findings are limited and should not be generalized to all advertising. Different research designs should be utilized for more generalizable assessments of ad recall. In addition, the proposed model in the present study assumed that exposure to news on ads took place firsthand. The model, therefore, did not consider why and how people read news stories about advertising.

Students are not frequent news readers; they often like television commercials better, especially when compared with baby boomers or other older generations. Results of the experiment may have differed had the experimental group included different populations with a broader age spectrum. Likewise, research is needed to study how news about ads from different media--broadcast, Web, and print--may affect ad and brand recall. In addition, the present study examined only one sequence of message exposure: news first and then ads. Further study is needed to discretely identify the similarities and differences across these dimensions and the consequences of the opposite sequence: ads first and then news. The encoding specificity principle (Tulving and Thomson 1973) suggests that retention increases to the degree that the environment during testing is similar to the learning environment. The news reading manipulation and the final memory measures after the Super Bowl were conducted in the same place and at the same time of the day. Thus, the findings could be inflated by location and time factors.

Furthermore, we assumed no prior knowledge of Super Bowl ads in the control group. However, as our results with Budweiser showed, in some cases, there may be prior knowledge of ads through the long-term commitment of particular products to a particular advertising campaign and event. This long-term strategy may increase ad recall regardless of news exposure. Thus, our assumptions of no prior ad knowledge may not have been valid for all products and may have influenced recall to a degree.

Message length for each brand ad in the news story was hypothesized to explore any differential news effects on the subsequent ad memory. The data showed a general pattern-greater recall of ads covered in greater detail in the news story. Several limitations should be addressed regarding this finding, however. First of all, without randomizing the order of presentation, the study could not exclude effects from presentation order. In this study, the longer brand ad stories were presented earlier in the news story. Thus, it was expected that primacy effects may be present in cases where readers could not or did not read all the way to the end of the news story. Second, the longer news coverage was primarily for well-known brands. In that sense, the study did not consider preexisting brand familiarity, which could have impacted the findings. Third, one could argue that the length of brand ad information might not be as important as the relevance of the content. In other words, this study did nor clarify what types of information embedded in the news could be better or worse for memory. Therefore, differential effects could be expected even if the amount of information given was equal. Fourth, the quality of the creative message in the ads was not measured and was therefore not included in the data analysis. A particular brand's advertisement may have been a better "creative product" than others. It is possible that news impact (group difference) would be decreased in cases where the focal ad was creatively superior, which in turn made it more memorable, regardless of its appearance or the length of discussion about it in a news story.

The study did not detect ad frequency effects that could have some impact on the findings. The notion that message repetition increases the memory of the message could be a plausible argument for lack of news impact when ad frequency increases, especially when the news exposure is given only once, as in this study. Accordingly, this study was unable to address whether the improved ad memory for prior exposure to the news is additive. For example, if the news has an additive effect, news + ad condition (one ad exposure after reading news), then it is possible that repeated exposure to the ad might have a similar effect (no news and a repetition of the same ad). Consequently, use of story length as an independent variable to gauge memory differences between the experimental group and control group might be misleading due to these limitations. Further research that considers the above issues could shed light on some of the inconsistent and inconclusive findings in this study.

Conclusions

Memory plays a major role in consumer choice (Bettman 1979; Lynch and Srull 1982). Presentation order, ad repetition, creative executions, and other factors can, to a degree, explain why some ads are recalled better from the crowded pool of ads. Results from the present study suggest that pre-publicity about ad campaigns is another factor that affects recall. It is apparent that a synergy exists between two promotional tools--marketing publicity and advertising--and advertisers could take advantage of this by utilizing an integrated marketing communications approach rather than focusing on one tool only.

APPENDIX

News Story for the Experimental Group

MOVE OVER, 'dot-coms' Other advertisers gear up for the Super Bowl

Madison Avenue has been calling Super Bowl XXXIV, to be broadcast by ABC on Jan. 30, the "dot-com bowl" or the "e-bowl." The reason: More than a dozen Internet companies are scheduled to make their Super Bowl advertising debuts during the game, compared with three in 1999 and one apiece in 1998. The eagerness of dot-corns to parade themselves before what is usually the largest viewing audience of the year helped drive ad rates for Super Bowl XXXIV to a record average of more than $2 million for each 30-second spot during the game. Other advertisers will join the lengthy list of initial advertisers dominated by such dot-coms as Autotrader.com, Etrade.com, HotJob.com, Monster.com, OurBeginning.com, and Pets.com.

* Tropicana: Unlike most of those dot-coms, "we have good awareness of our brand," Mr. Berdusco said, adding: "Our purpose is to seed a new campaign with an important message. And what better way to expose that message to a lot of people who would benefit from it than kicking it off during the Super Bowl?"

The campaign portrays the Tropicana Pure Premium brand of orange juice as a quick, nutritional fuel for busy people. The campaign, carrying the theme "The best 9.3 seconds you can possibly spend on yourself," will replace a four-year-old effort that characterized the brand as "perfect." The commercial to run during the Super Bowl will present in high speed a sequence of events that are part of a person's morning routine. The footage is interrupted when the person drinks a glass of Tropicana in 9.3 seconds, which is shown in real time. That accomplished, the footage reverts to the artificially faster pace as the spot ends.

"The number '9.3 seconds' had a good ring to it," Mr. Fischer said, "and we did some empirical testing by having some people at the agency guzzle a glass and time them." "We've taken it to some focus groups," he added, "and we've gotten great response." There's also a fiendishly clever and subtle dig in the "9.3 seconds" pitch, aimed at Tropicana's principal competitor, the Minute Maid brand of orange juice sold by the Coca-Cola Company. The implication is that 9.3 seconds is faster-and therefore better-than a minute, a word that occurs frequently in the current Minute Maid campaign, which carries the theme "Squeeze the day."

* Electronic Data System: EDS is bringing something entirely new to the Super Bowl advertising party: Cats. The stodgy technology company's 60-second parody of a Western epic features a cadre of cowboys on horseback rounding up thousands of cats, not cattle, running through an expansive mountain range. And as one kitten wrangler, his face raw and scratched from a close encounter with a cat, puts it plainly: "Don't let anybody tell you it's easy." David Lubars, president and creative director says, they tried to dream up a commercial that is both entertaining and delivers the message that EDS creates order out of chaos. In technology, he says, it is also "an expression to talk about how impossibly difficult something is."

* Seven-Up: Give 7-Up credit for persistence. After ABC taste monitors shot down its proposed spot "T-Shirt," it won approval for another called "Show Us Your Can." The spot uses the edgy "Make 7-Up Yours" theme that made its debut this fall, says spokeswoman Philppa Dworkin. The second-quarter spot stars Mad TV's Orlando Jones as a befuddled marketing exec. 7-Up is also running five pre-game spots with Jones.

* Anheuser-Busch: The beer baron is the game's largest advertiser, with ten 30-second slots. Expect to see at least one ad starring a Busch family member and one safe-driving spot. A-B says it's still weighing cantankerous lizards, talking turtles and Clydesdale horses for the rest. A-B, the exclusive beer advertiser for the 12th straight year, has secured the coveted first spot after kickoff. A-B commercials will promote Budweiser, Bud Light or Michelob Light.

* Motorola: The $30 billion technology company scooped up rights in April as the NFL's "official wireless communications sponsor." The Super Bowl rookie now is running two in-game spots for its array of wireless Web-browsing products. The theme: "Web Without Wires." Super Bowl ads are part of "integrating and leveraging our sponsorship," explains Dave Weisz, director of global sports and event marketing.

TABLE 1
Parameter Estimates and Model Fit Indices

                                        Unstandardized
                                         coefficients
Path                                    (standardized)    t value

NEWS [right arrow] INTEREST (H1)           .52 (.17)      2.20 *
INTEREST [right arrow] ATTENTION (H2)      .63 (.55)      7.19 ***
ATTENTION [right arrow] MEMORY (H3)        .37 (.48)      5.01 ***
NEWS [right arrow] MEMORY (H4)             .48 (.18)      2.10 *
NEWS [right arrow] ATTENTION               .32 (.09)      1.31
INTEREST [right arrow] Y1
  (how much interest)                     1.00 (.94)
INTEREST [right arrow] Y2
  (how much want to see)                   .99 (.92)      21.12 ***
INTEREST [right arrow] Y3
  (how much curiosity)                     .98 (.92)      21.05 ***
ATTENTION [right arrow] Y4
  (leave the room)                        1.00 (.93)
ATTENTION [right arrow] Y5
  (change the channel)                     .92 (.89)      16.05 ***
ATTENTION [right arrow] Y6
  (how much attention)                     .77 (.77)      12.24 ***
MEMORY [right arrow] Y7 (B-RECALL)        1.00 (.78)
MEMORY [right arrow] Y8 (AD-RECALL)        .99 (.68)      5.07 ***

Notes: Model fit indices: [chi square] = 34.5, p > .05;
GFI = .95; CFI = .99; AGFI = .91; RMSEA = .05.
Squared multiple correlations ([R.sup.2]): INTEREST = .03;
ATTENTION = .32; MEMORY = .30.

* p < .05.

** p < .01.

*** p <.001.

TABLE 2
Group Differences for Each Brand in the News Story

               Brand names (ad frequency    Percentage of people
Memory         during the game)                 who recalled

                                           Control   Experimental
                                            group       group
Brand recall   Brands with descriptions
                 Tropicana (1)              3.2%        37.8%
                 EDS (1)                    4.2         20.3
                 7 UP (1)                  14.7         31.1
                 Bud (8)                   78.9         82.4
                 Motorola (2)               5.3         16.2
               Brand names only
                 Autotrade (1)              0%          6.8%
                 Etrade (3)                56.8        55.4
                 Hotjobs (1)                9.5         6.8
                 Monster (2)               15.8        13.5
                 Ourbeginning (1)           0           2.7
                 Pet (1)                   11.6        24.3
Ad recall      Brands with descriptions
                 Tropicana (1)              3.2%       18.9%
                 EDS (1)                   16.8        33.8
                 7 UP (1)                  11.6        23.0
                 Bud (8)                   56.8        66.2
                 Motorola (2)               3.2         5.4
               Brand names only
                 Autotrade (1)               0%          0%
                 Etrade (3)                 38.9        36.5
                 Hotjobs (1)                 1.1         1.4
                 Monster (2)                 3.2         4.1
                 Ourbeginning (1)            1.1         1.4
                 Pet (1)                    10.5         9.5

               Brand names (ad frequency
Memory         during the game)            [chi sqaure]

Brand recall   Brands with descriptions
                 Tropicana (1)               33.4 ***
                 EDS (1)                     10.8 **
                 7 UP (1)                     6.5 ***
                 Bud (8)                      3
                 Motorola (2)                 5.5 ***
               Brand names only
                 Autotrade (1)                6.6 *
                 Etrade (3)                   0
                 Hotjobs (1)                    .4
                 Monster (2)                    .2
                 Ourbeginning (1)              2.6
                 Pet (1)                       4.8 *
Ad recall      Brands with descriptions
                 Tropicana (1)                11.4 **
                 EDS (1)                       6.5 *
                 7 UP (1)                      3.9 *
                 Bud (8)                       1.5
                 Motorola (2)                   .5
               Brand names only
                 Autotrade (1)                 n.a.
                 Etrade (3)                     .1
                 Hotjobs (1)                   0
                 Monster (2)                   0
                 Ourbeginning (1)              0
                 Pet (1)                       0

Note: n.a. = not available.

* p < .05.

** p < .01.

*** p<.001.

TABLE 3
Group Means and Significance Tests for Brand Names Recalled

                                Group 1        Group 2    Group 3
                             No news/No ads   News only   Ads only
                                (n = 52)      (n = 41)    (n = 93)

Five descriptive brand ads
  in the news (mean)            .68           1.39 (a)    1.07 (a)
Six nondescriptive brands
  in the news (mean)            .12 (a)        .28 (a)     .93 (b)
Other brand ads (mean)          .76 (a)        .38 (a)    1.44 (b)

                             Group 4
                             News/Ads
                             (n = 74)

Five descriptive brand ads
  in the news (mean)         1.87
Six nondescriptive brands
  in the news (mean)         1.09 (b)
Other brand ads (mean)       1.29 (b)

Note: Brand names recalled in Group 1 and Group 2 are based
on subjects' guesses about what they thought appeared during
the Super Bowl broadcast (they did not watch the game).
Group 3 is the control group (no news story) and Group 4 is
the experimental group (exposure to news story). Means with
same letter within a row are not significantly different.

REFERENCES

Ailawadi, Kusum, Paul Farris, and Mark Parry (1997), "Explaining Variations in the Advertising and Promotional Costs/ Sales Ratio: A Rejoinder," Journal of Marketing, 61 (January), 93-96.

Alba, Joseph W., and Amitava Chattopadhyay (1986), "Salience Effects in Brand Recall," Journal of Marketing Research, 23 (November), 363-369.

Balasubramanian, Kiva K., and V. Kumar (1990), "Analyzing Variations in Advertising and Promotional Expenditure: Key Correlates in Consumer, Industrial, and Service Markets," Journal of Marketing, 54 (April), 57-68.

Bettman, James R. (1979), "Memory Factors in Consumer Choice: A Review," Journal of Marketing, 43 (Spring), 37-53.

Buchholz, Laura M., and Robert E. Smith (1991), "The Role of Consumer Involvement in Determining Cognitive Response to Broadcast Advertising," Journal of Advertising, 20 (1), 4-17.

Burnkrant, Robert E., and Alan G. Sawyer (1983), "Effects of Involvement and Message Content on Information Processing Intensity," in Information Processing Research in Advertising, Richard Jackson Harris, ed., Hillsdale, NJ: Lawrence Erlbaum, 43-64.

Confer, Marian G. (1992), "The Media Multiplier: Nine Studies Conducted in Seven Countries," Journal of Advertising Research, 32 (1), RC4-RC10.

--, and Donald McGlathery (1991), "The Research Study: The Advertising Impact of Magazines in Conjunction with Television," Journal of Advertising Research, 31 (1), RC2-RC5.

Duncan, Tom, and Clarke Caywood (1996), "The Concept, Process, and Evolution of Integrated Marketing Communication," in Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds., Mahwah, NJ: Lawrence Erlbaum, 13-34.

Edell, Julie A., and Kevin Lane Keller (1989), "The Information Processing of Coordinated Media Campaigns," Journal of Marketing Research, 26 (May), 149-163.

Eliashberg, Jehoshua, and Steven M. Shugan (1997), "Film Critics: Influencers or Predictors?" Journal of Marketing, 61 (April), 68-78.

--, and Thomas S. Robertson (1988), "New Product Preannouncing Behavior: A Market Signaling Study," Journal of Marketing Research, 25 (August), 282-292.

Fitzgerald, Kate (2001), "Bud's Super Bowl," Advertising Age, 72 (January 15), 20.

Frieske, David A., and Denise C. Park (1999), "Memory for News in Young and Old Adults," Psychology and Aging, 14 (1), 90-98.

Greenwald, Anthony G., and Clark Leavitt (1984), "Audience Involvement in Advertising: Four levels," Journal of Consumer Research, 11 (June), 581-592.

Harris, Thomas L. (1998), Value-Added Public Relations: The Secret Weapon of Integrated Marketing, Lincolnwood, IL: NTC Business Books.

Jin, Hyun Seung, and Xinshu Zhao (1999), "Effects of Media Coverage on Advertised Brand Recall and Recognition," in Proceedings of 1999 Conference of the American Academy of Advertising, Marilyn S. Roberts, ed., Gainesville, FL: University of Florida, 54-61.

Johnson, Bradley (1994), "Ten Years After '1984': The Commercial and the Product That Changed Advertising," Advertising Age, 65 (January 10), 1.

Keenan, Kevin L. (1995), "Television News Coverage of Advertising: An Exploratory Census of Content," in Proceedings of the 1995 Conference of the American Academy of Advertising, Charles S. Madden, ed., Waco, TX: Baylor University, 174-180.

Krugman, Herbert E. (1965), "The Impact of Television Advertising: Learning Without Involvement," Public Opinion Quarterly, 29 (Fall), 349-356.

Lutz, Richard J. (1996), "Some General Observations About Research on Integrated Marketing Communications," in Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds., Mahwah, NJ: Lawrence Erlbaum, 355-367.

Lynch, John S., Jr., and Thomas K. Srull (1982), "Memory and Attentional Factors in Consumer Choice: Concepts and Research Methods," Journal of Consumer Research, 9 (June), 18-37.

Melton, Arthur W. (1970), "The Situation with Respect to the Spacing of Repetition and Memory," Journal of Verbal Learning and Verbal Behavior, 9 (October), 596-606.

Moriarty, Sandra E. (1996), "The Circle of Synergy: Theoretical Perspectives and an Evolving IMC Research Agenda," in Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds., Mahwah, NJ: Lawrence Erlbaum, 333-353.

Newell, Stephen J., and Kenneth V. Henderson (1998), "Super Bowl Advertising: Field Testing the Importance of Advertisement Frequency, Length and Placement on Recall," Journal of Marketing Communications, 4, 237-248.

"No PR, No Comment" (1995), Marketing (January 26), 37.

Pasadeos, Yorgo, Joe Phelps, and Opdycke Lamme (2000), "Newspaper Coverage of Advertising: The Unfolding Story," in Proceedings of 2000 Conference of the American Academy of Advertising, Mary Alice Shaver, ed., East Lansing: Michigan State University, 79-89.

Pavelchak, Mark A., John H. Antil, and James M. Munch (1988), "The Super Bowl: An Investigation into the Relationship Among Program Context, Emotional Experience, and Ad Recall," Journal of Consumer Research. 15 (December), 360-367.

Pham, Michel Tuan, and Gita Venkataramani Johar (1997), "Contingent Processes of Source Identification," Journal of Consumer Research, 24 (December), 249-265.

Raaijmakers, Jeroen G., and Richard M. Shiffrin (1981), "Search of Associative Memory," Psychological Review, 99, 92-134.

Ries, Al, and Laura Ries (1999), "First Do Some Great Publicity: Building a Brand Takes More Than a Big Burst of Advertising Dollars," Advertising Age (February 8), 42.

Rothenberg, Randall (1990), "Commercials Become News and the Airtime Is Free," New York Times (January 8), D1.

Salmon, Charles T. (1986), "Perspectives on Involvement in Consumer and Communication Research," in Progress in Communication Sciences, B. Dervin and M. J. Voight, eds., Norwood, NJ: Ablex, 243-268.

--, Leonard N. Reid, James Pokrywcznski, and Robert W. Willett (1985), "The Effectiveness of Advocacy Advertising Relative to News Coverage," Communication Research, 12 (October), 546-567.

Schultz, Don E., Stanley I. Tannenbaum, and Robert F. Lauterborn (1992), Integrated Marketing Communications, Lincolnwood, IL: NTC.

Shoemaker, Pamela J., and Stephen D. Reese (1996), Mediating the Message, White Plains, NY: Longman.

Speck, Surgi, and Michael T. Elliott (1997), "Predictors of Advertising Avoidance in Print and Broadcast Media," Journal of Advertising, 26 (3) (Fall), 61-76.

Tedesco, John C., Lori Melton McKinnon, and Lynda Lee Kaid (1996), "Advertising Watchdogs: A Content Analysis of Print and Broadcast Ad Watches," Harvard International Journal of Press and Politics, 1 (4), 76-93.

Thorson, Esther, and Jeri Moore (1996), "Introduction," in Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds., Mahwah, NJ: Lawrence Erlbaum, 1-10.

Tulving, Endel, and Donald M. Thomson (1973), "Encoding Specificity and Retrieval Processes in Episodic Memory," Psychological Review, 80, 352-373.

Unnava, H. Rao, and Robert E. Burnkrant (1991), "Effects of Repeating Varied Ad Execution on Brand Name Memory," Journal of Marketing Research, 28 (November), 406-416.

Yates, Karen (1995), "Making News: Press Coverage of Advertising Campaign," Campaign (May 26), 36.

Young, Daniel R., and Francis S. Bellezza (1982), "Encoding Variability, Memory Organization and the Repetition Effect," Journal of Experimental Psychology: Learning, Memory and Cognition, 8 (November), 545-559.

Zaichkowsky, Judith Lynne (1986), "Conceptualizing Involvement," Journal of Advertising, 15 (2), 4-14.

Zhao, Xinshu, and Glen L. Bleske (1998), "Polling Effects: Horse-Race Polls and Audience Issue Learning," Harvard International Journal of Press and Politics. 3 (4), 13-34.

Hyun Seung Jin (Ph.D., University of North Carolina at Chapel Hill) is assistant professor, A.Q. Miller School of Journalism and Mass Communications, Kansas State University.

The author thanks Jane Brown, Xinshu Zhao, James Bettman, Richard Lutz, Ronald Faber, the former editor, and three anonymous reviewers for their helpful comments on a previous version of the manuscript.