C. Yoo, D MacInnis /Journal of Business Research 58(2005)1397-1406 content directly, it seems reasonable that aroused emotions talking to her father over the phone about how much she result from evaluation of the ads, as opposed to the misses him execution. Combined, our reasoning suggests that None of the pretested ads typified an informational ad H5. When the ad execution is informational, credibility execution. To create an informational ad execution. we oughts affect Aad(a) directly and(b) indirectly through replaced the audio of the emotional ad with a voice-over, their mediational effect on feelings which stated attribute information about the long-distance company(e.g, operation service, quality of sound, and etc. Another aspect of evaluative thoughts is that when the ad The visual aspects of the ad (with the girl talking to her execution is informational, it should affect beliefs. For father over the phone) were unchanged example, the more credible the ad execution is, the stronger efore manipulating the informational ad format con- consumers'beliefs about the brand are likely to be(petty dition, we reviewed informational type of the AT and T ads and Cacioppo, 1986; MacInnis and Jaworski, 1989). Since Typical type of the informational ad was to explain excellent the information provided by beliefs may provide the operation service, sound quality, 24 h service and etc. with a opportunity for consumers to form relatively strong brand talking scene. Therefore, we expected that we can create beliefs, beliefs may be an important predictor of brand typical type of informational ad by replacing audio of the attitudes(see Lutz, 1975). Thus we propose that Moreover, keeping the visual aspects of both ads H6. When the ad execution is informational, thoughts about constant across both ad execution conditions allowed us to the ad's credibility influence beliefs control for potentially confounding factors that might be H7. When the ad execution is informational, beliefs involved with the use of totally different ads. Ads in both influence brand attitudes conditions began with the same introductory song, followed by the narration(conversation of girl with her father; As noted above, informational ads can influence feelings announcer's statement of brand benef th ended with by affecting the credibility thoughts consumers have the logo and announcement of the company name. The towards ads. These feelings may, in turn, influence how visuals, announcer, music, logo, ad sequence and ad length much consumers like the ad. Thus, consumers are unlikely are thus identical in the two ads. to like ads that make them feel irritable, skeptical,or An additional pretest with forty consumers was conducted annoyed. We expect, then, that feelings influence ad to assess the success of the informational ad execution attitudes and that they affect brand attitudes through the manipulation. The mean emotionality and informationality mediational influence of Aad. Thus we propose that scores of this new ad were 2.98(emotionality score of original emotional ad: 5.07)and 4.23 (informationality score H8. When the ad execution is informational, feelings of original ad: 2.45). We found significant mean differences influence Ab(a)directly and(b) indirectly through between emotional and informational ad with respect to mediational influence of Aad emotionality and informationality scores (P<.001) Because we are interested in the processes by which brand attitudes are formed and because we wished to create 3. Method homogeneity across consumers in the nature of the exposure 3.1 Stimulus selection novel ad for a novel brand. for both the informational and emotional commercials, the original telephone companys To experimentally manipulate emotional and informa- logo and signature were removed and were replaced by a tional ad execution format and examine their effects on the novel company name and logo. The introductory song was brand attitude formation processes noted above, we pre- also replaced with new music, as the original song is likely tested the informational and emotional ad format of fourteen to have cued the company name. As described below, our commercials using forty subjects. All commercials were for analyses included only subjects who reported unfamiliarity a long-distance telephone company. Subjects were asked to with the commercial and had no prior association of it with view the ads and rate the extent to which they seemed to the original commercial characterize an emotional ys. informational ad execution format. Several 7-point Likert scaled items were used to 3.2. Subjects, design and procedur measure emotional ad format (this ad appeals to my emotion,, this ad creates a mood,)and informational ad Two hundred two students were randomly assigned to format ('this ad appeals to my rationality, this ad provides either the informational or emotional ad execution con- a lot of information). One ad typified the emotional ad dition. In each exposure setting, subjects watched the format (mean emotionality score of 5.07)and was regarded designated commercial twice with a 10-s lag between as relatively weak on informationality(mean information- exposures. Questionnaires were administered immediately ality score of 2.45). This commercial shows a daughter following the second exposure and measures of Ab and Aadcontent directly, it seems reasonable that aroused emotions result from evaluation of the ads, as opposed to the execution. Combined, our reasoning suggests that: H5. When the ad execution is informational, credibility thoughts affect Aad (a) directly and (b) indirectly through their mediational effect on feelings. Another aspect of evaluative thoughts is that when the ad execution is informational, it should affect beliefs. For example, the more credible the ad execution is, the stronger consumers’ beliefs about the brand are likely to be (Petty and Cacioppo, 1986; MacInnis and Jaworski, 1989). Since the information provided by beliefs may provide the opportunity for consumers to form relatively strong brand beliefs, beliefs may be an important predictor of brand attitudes (see Lutz, 1975). Thus we propose that: H6. When the ad execution is informational, thoughts about the ad’s credibility influence beliefs. H7. When the ad execution is informational, beliefs influence brand attitudes. As noted above, informational ads can influence feelings by affecting the credibility thoughts consumers have towards ads. These feelings may, in turn, influence how much consumers like the ad. Thus, consumers are unlikely to like ads that make them feel irritable, skeptical, or annoyed. We expect, then, that feelings influence ad attitudes and that they affect brand attitudes through the mediational influence of Aad. Thus we propose that: H8. When the ad execution is informational, feelings influence Ab (a) directly and (b) indirectly through mediational influence of Aad. 3. Method 3.1. Stimulus selection To experimentally manipulate emotional and informa￾tional ad execution format and examine their effects on the brand attitude formation processes noted above, we pre￾tested the informational and emotional ad format of fourteen commercials using forty subjects. All commercials were for a long-distance telephone company. Subjects were asked to view the ads and rate the extent to which they seemed to characterize an emotional vs. informational ad execution format. Several 7-point Likert scaled items were used to measure emotional ad format (Fthis ad appeals to my emotion_, Fthis ad creates a mood_) and informational ad format (Fthis ad appeals to my rationality_, Fthis ad provides a lot of information_). One ad typified the emotional ad format (mean emotionality score of 5.07) and was regarded as relatively weak on informationality (mean information￾ality score of 2.45). This commercial shows a daughter talking to her father over the phone about how much she misses him. None of the pretested ads typified an informational ad execution. To create an informational ad execution, we replaced the audio of the emotional ad with a voice-over, which stated attribute information about the long-distance company (e.g., operation service, quality of sound, and etc.). The visual aspects of the ad (with the girl talking to her father over the phone) were unchanged. Before manipulating the informational ad format con￾dition, we reviewed informational type of the AT and T ads. Typical type of the informational ad was to explain excellent operation service, sound quality, 24 h service and etc. with a talking scene. Therefore, we expected that we can create typical type of informational ad by replacing audio of the emotional ad. Moreover, keeping the visual aspects of both ads constant across both ad execution conditions allowed us to control for potentially confounding factors that might be involved with the use of totally different ads. Ads in both conditions began with the same introductory song, followed by the narration (conversation of girl with her father; announcer’s statement of brand benefits). Both ended with the logo and announcement of the company name. The visuals, announcer, music, logo, ad sequence and ad length are thus identical in the two ads. An additional pretest with forty consumers was conducted to assess the success of the informational ad execution manipulation. The mean emotionality and informationality scores of this new ad were 2.98 (emotionality score of original emotional ad: 5.07) and 4.23 (informationality score of original ad: 2.45). We found significant mean differences between emotional and informational ad with respect to emotionality and informationality scores ( p <.001). Because we are interested in the processes by which brand attitudes are formed and because we wished to create homogeneity across consumers in the nature of the exposure and processing task, we wanted to expose consumers to a novel ad for a novel brand. For both the informational and emotional commercials, the original telephone company’s logo and signature were removed and were replaced by a novel company name and logo. The introductory song was also replaced with new music, as the original song is likely to have cued the company name. As described below, our analyses included only subjects who reported unfamiliarity with the commercial and had no prior association of it with the original commercial. 3.2. Subjects, design and procedures Two hundred two students were randomly assigned to either the informational or emotional ad execution con￾dition. In each exposure setting, subjects watched the designated commercial twice with a 10-s lag between exposures. Questionnaires were administered immediately following the second exposure and measures of Ab and Aad 1400 C. Yoo, D. MacInnis / Journal of Business Research 58 (2005) 1397 – 1406