the effects of cognitive and affective program involvement on cognitive and affective ad involvement

12
JOURNAL OF BUSINESS AND PSYCHOLOGY Volume 13, No. 1, Fall 1998 THE EFFECTS OF COGNITIVE AND AFFECTIVE PROGRAM INVOLVEMENT ON COGNITIVE AND AFFECTIVE AD INVOLVEMENT Kevin G. Celuch Mark Slama Illinois State University ABSTRACT: Advertisers attempt to create ads which gain the attention and in- volvement of their audience. These advertisements may be primarily cognitively or affectively involving. Similarly the programs in which the ads are embedded may be primarily cognitively or affectively involving. Does the type of program involvement affect involvement with ads? Building on previous research (Mc- Clung, Park and Sauer 1985; Park and McClung 1986) a study is presented which suggests that cognitive involvement in ads (particularlycognitive ads) suf- fers from an overload effect when the ads are placed in cognitively involving programs while affective involvement in ads is enhanced by a priming effect when the ads are placed in affectively involvingprograms. INTRODUCTION With, on average, over 90 percent of the advertising budget spent on media, clearly media planning is a financially as well as strategically significant advertising arena. Increasing media alternatives and audi- ence fragmentation serve to complicate media planning (cf. Arens and Bovee, 1994; Krugman, Reid, Dunn, and Barban, 1994; Schultz and Barnes 1995). Within this broader media landscape, a substantial amount of research supports the idea that the media environment, spe- cifically, programs can influence the effectiveness of ads (Axelrod 1963; Bryant and Comisky 1978; Clancy and Kewskin 1971; Crane 1964; Gold- berg and Gorn 1987; Kennedy 1971; Singh and Hitchon 1989). There- fore, the issue of program environment would seem to warrant careful Address correspondence to Kevin G. Celuch, Ph.D., Department of Marketing, College of Business, Illinois State University, Normal, Illinois 61761. 115 © 1998 Human Sciences Press, Inc.

Upload: independent

Post on 29-Nov-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

JOURNAL OF BUSINESS AND PSYCHOLOGYVolume 13, No. 1, Fall 1998

THE EFFECTS OF COGNITIVE ANDAFFECTIVE PROGRAM INVOLVEMENT ON

COGNITIVE AND AFFECTIVE ADINVOLVEMENT

Kevin G. CeluchMark Slama

Illinois State University

ABSTRACT: Advertisers attempt to create ads which gain the attention and in-volvement of their audience. These advertisements may be primarily cognitivelyor affectively involving. Similarly the programs in which the ads are embeddedmay be primarily cognitively or affectively involving. Does the type of programinvolvement affect involvement with ads? Building on previous research (Mc-Clung, Park and Sauer 1985; Park and McClung 1986) a study is presentedwhich suggests that cognitive involvement in ads (particularly cognitive ads) suf-fers from an overload effect when the ads are placed in cognitively involvingprograms while affective involvement in ads is enhanced by a priming effectwhen the ads are placed in affectively involving programs.

INTRODUCTION

With, on average, over 90 percent of the advertising budget spent onmedia, clearly media planning is a financially as well as strategicallysignificant advertising arena. Increasing media alternatives and audi-ence fragmentation serve to complicate media planning (cf. Arens andBovee, 1994; Krugman, Reid, Dunn, and Barban, 1994; Schultz andBarnes 1995). Within this broader media landscape, a substantialamount of research supports the idea that the media environment, spe-cifically, programs can influence the effectiveness of ads (Axelrod 1963;Bryant and Comisky 1978; Clancy and Kewskin 1971; Crane 1964; Gold-berg and Gorn 1987; Kennedy 1971; Singh and Hitchon 1989). There-fore, the issue of program environment would seem to warrant careful

Address correspondence to Kevin G. Celuch, Ph.D., Department of Marketing, Collegeof Business, Illinois State University, Normal, Illinois 61761.

115 © 1998 Human Sciences Press, Inc.

JOURNAL OF BUSINESS AND PSYCHOLOGY

consideration by parties responsible for the placement of advertise-ments.

Advertising practitioners have typically relied on program ratingsand audience demographics as a basis for the placement of an ad in aprogram which will maximize the ad's effectiveness. Soldow and Prin-cipe (1981) and Park and McClung (1986) have noted a limitation ofusing such information. Ratings and demographics provide informationabout who is watching what program but give no indication of how in-volved individuals are with a program.

Recently, program-related advertising issues have been approachedfrom the theoretical standpoint of involvement (Soldow and Principe1981; Krugman 1983; Park and Young 1983; 1986; Park and McClung1986). Involvement with the content of communications (i.e., issues,products, etc.) has long been thought of as an important mediator ofindividuals' responses to persuasive communications.

Both cognitive and affective involvement have been found to influ-ence viewers' reactions to advertisements (Park and Young 1983; 1986).Cognitive involvement may be defined as a concern with the functionalinformation content of a communication. Affective involvement may bedefined as a concern with the emotional and value expressive content ofa communication.

A person may have cognitive or affective involvement with any typeof communication including television programs and ads. Given that ad-vertisers wish to maximize the effectiveness of their ads, they need toknow whether to choose cognitively or affectively involving programs inwhich to place their ads. Such a choice is likely to depend on the type ofinvolvement created by the ad. Image appeals tend to create affective adinvolvement while functional appeals tend to create cognitive ad in-volvement. Is it better to place a cognitively involving ad in a cognitivelyinvolving program or an affectively involving program? What type ofprogram would be best for an affectively involving ad?

As of yet, no empirical research has been presented which examineshow the levels of cognitive and affective involvement in programs influ-ence the levels of cognitive and affective involvement in ads. The pur-pose of this research is to fill this void.

Previous Research

Conceptualizing by McClung, Park and Sauer (1985) has suggestedhow program involvement and ad involvement may be linked and em-pirical work by Park and McClung (1985) has supported the existence ofinteractions between program involvement and ad involvement. Mc-Clung, Park and Sauer (1985) have proposed that the effectiveness ofcognitively involving ads will be enhanced by moderate levels of cogni-

116

KEVIN G. CELUCH AND MARK SLAMA

tive involvement in programs because of congruity between the process-ing style required for the program and the ad. However, high levels ofcognitive program involvement could reduce ad effectiveness becausethe detailed analytical processing style associated with high cognitiveinvolvement may not be able to process both the information from theprogram and the ad; therefore, involvement in the ad as well as effec-tiveness of the ad could be reduced.

An affectively involving ad, however, will be subject to analogicalprocessing in which information is grouped into larger "chunks" andthere should be little if any detriment in the analogical processing of anaffective ad due to high affective program involvement. Therefore, undermost conditions there should simply be a priming effect in which in-volvement in and effectiveness of affectively involving ads are enhancedby higher levels of affective involvement in the program.

In an empirical test with overall ad involvement as the dependentvariable (not cognitive and affective ad involvement as studied here),Park and McClung (1986) found some support for these ideas. Consis-tent with the notion of an overload effect for cognitive involvement, theyfound small decreases in ad involvement as cognitive program involve-ment was increased from low to moderate to high. Affective programinvolvement, however, seemed to produce a priming effect as ad involve-ment increased substantially from low to moderate affective programinvolvement. Only a small and apparently insignificant decrease in adinvolvement was associated with the increase in affective program in-volvement from moderate to high.

While the Park and McClung (1986) study provides valuable insighton the interactions between program and ad involvement certain limita-tions should be noted. Park and McClung (1986) used an affective and acognitive program to produce affective and cognitive program involve-ment, but they manipulated the level of program involvement by givingthe subjects in the low, moderate and high conditions different instruc-tions for processing the two programs. These instructions directly cre-ated the processing style and type of involvement rather than allowingit to arise from a more natural viewing of the program.

In addition, Park and McClung measured only the overall level ofad involvement rather than cognitive and affective types of ad involve-ment. It would be useful to consider both affective and cognitive types ofinvolvement with ads because McClung, Park and Sauer (1985) makedifferent predictions regarding the impact of program involvement forcognitively and affectively involving ads, specifically, that cognitively in-volving ads should suffer more from overload than affectively involvingads. The current research is concerned with the influence of programinvolvement on ad involvement and extends the Park and McClung(1986) study in two important ways. First, program involvement is al-

117

JOURNAL OF BUSINESS AND PSYCHOLOGY

lowed to occur naturally instead of being induced through processinginstructions. Second, cognitive and affective ads are employed and cog-nitive and affective ad involvement are measured rather than just over-all ad involvement. An attempt is made to be as realistic as possibleusing typical choices for cognitive and affective programs (60 Minutesand Cheers, respectively). It is expected based on the theory of McClung,Park and Sauer (1985, which suggests cognitive overload effects occurfor high cognitive program involvement) that when advertisers place ei-ther cognitive or affective ads within a typical cognitively involving pro-gram there will be a decrease in cognitive ad involvement due to anoverload effect as compared to placing the same ad in an affective pro-gram.

To the contrary, no decrease in affective ad involvement is expectedwhen either cognitive or affective ads are placed in an affectively involv-ing program as compared to a cognitively involving program. In fact, itseems that under relatively natural conditions (no processing instruc-tions such as used in the Park and McClung experiment) the averagelevel of affective involvement for viewers of typical affectively involvingtelevision programs will be moderate to high and will produce a primingrather than an overload effect. This is consistent with the very largeincrease in ad involvement found by Park and McClung (1986) goingfrom the low to moderate affective program involvement conditions. Thehigh affective program involvement condition in their experiment, whichproduced a small decrease in ad involvement relative to the moderateaffective program involvement condition, is probably higher (due to thenature of the instructions they used) than would be expected of a typicaltelevision audience.

Given the previous discussion the following hypotheses will betested by this research. H1 is presumed to be due to the priming effectfor affective involvement. H2 is due to the cognitive overload effect.

H1: Affective ad involvement will be higher for both cognitive andaffective ads when placed in an affective program.

H2: Cognitive involvement will be lower for both cognitive and af-fective ads when placed in a cognitive program.

METHOD

Procedure and Subjects

The experiment employed a 2 (program involvement type: cognitiveor affective) X 2 (ad involvement type: cognitive or affective) factorial

118

KEVIN G. CELUCH AND MARK SLAMA

design in which subjects were randomly assigned to one of four condi-tions. Individuals were provided with a cover story that they were par-ticipating in research related to the development of new measurementinstruments. Subjects were informed that the study would involvewatching a program segment followed by a commercial. Next, they wereprovided instructions for completing the measures. Subjects were thenexposed to one of the stimulus program/ad conditions and, following ex-posure, asked to complete a measurement packet.

Two hundred and thirty four students enrolled in business classesat a large midwestern university participated in this study. Subjectswere tested in small groups, a setting which is not uncommon for collegestudents, and earned extra credit points for participation.

Stimulus Materials

Programs and ads were selected after reviewing many televisionprograms and ads and attempting to find two of each which would scorequite differently in terms of cognitive and affective involvement. Stim-ulus materials for the study consisted of two 10-15 minute programsegments and two 1-2 minute commercials. Neither of the ads had beenrun in the geographic market of the study. Further, although the adsrelated to existing brands, attitudinal contamination was deemed to below given the type of individuals comprising our sample and the natureof the products (i.e., pagers and cognac). All existing commercials wereerased from the programs and one stimulus ad embedded at the end of aprogram in accordance with design conditions.

A segment of 60 Minutes featuring a report on unethical loss pre-vention tactics utilized by a retail chain was selected for the cognitivelyinvolving program. Part of an episode of the program Cheers focusing onthe interaction/relationship between the primary male and female char-acters was selected for the affectively involving program.

Two 30 second commercials running consecutively for a brand ofpagers were used for the cognitively involving ad. The ad executionsemployed words flashed on the screen accompanied by upbeat music.The ads contained references to the use of the advertised pagers by keymanagement at two prominent U.S. companies. An 800 telephone num-ber was provided at the end of each ad. A two minute commercial forcognac was used for the affectively involving ad. The ad execution in-cluded six vignettes of evocative male/female images (i.e., romantic/erotic themes) accompanied by dramatic music. The brand name wasflashed at the beginning of the ad while a "product shot" concluded thecommercial.

119

JOURNAL OF BUSINESS AND PSYCHOLOGY

Measures

Cognitive and affective program and ad involvement were assessedvia scales adapted from Stanford (1984) and Schlinger (1979) (please seeTable 1 for specific items). Measures were formed by summing relateditem scores and dividing by the number of items in each scale.

Cognitive responses generated while watching the ad were also elic-ited through a procedure reported in MacKenzie and Lutz (1989). Inaccordance with the coding scheme presented in Laczniak et al. (1989),responses were coded as message, brand, product, ad-related, or other.Subject thoughts were categorized by trained graduate student coderswho had prior experience classifying cognitive responses yet were blindto the hypotheses of this study.

Cronbach's coefficient alpha was used to assess the internal consis-tency of the multiple item measures used in the study (Peter, 1979).Three of the four scales evidenced estimates greater than .8 with theexception being affective program involvement (.63). Overall, the inter-nal consistency estimates for measures used in this study are compara-ble with estimates reported in related research (Celuch and Slama, 1995).

120

Table 1Program and Ad Involvement Measures

Measures*

Cognitive Program Involvement1. I learned information from the program.2. The program was informative.3. The program contained information that is relevant to current events/topics.

Affective Program Involvement1. Aspects of the program were attractive to me.2. I felt as though I was right there in the program.3. The program was personal and intimate.4. Ifelt that the characters/people in the program were acting out what I feel at times

Cognitive Ad Involvement1. I learned information from the commercial.2. The commercial was informative.3. The commercial contained product-relevant information.

Affective Ad Involvement1. Aspects of the commercial were attractive to me.2. I felt as though I was right there in the commercial.3. The commercial was personal and intimate.4. I felt that the characters/people in the commercial were acting out what I feel at

times.5. I could personally relate to aspects of the commercial.6. The commercial portrayed the way people feel at times.

*A1llitems utilized seven-point scales with endpoints "strongly disagree"/"stronglyagree."

KEVIN G. CELUCH AND MARK SLAMA

Pretesting

Program pretesting utilizing the cognitive and affective program in-volvement measures described above confirmed that the 60 Minute seg-ment was more cognitively involving (M = 6.6) compared to the Cheerssegment (M = 5.3; t(50) = 4.04,p<.001). Conversely, Cheers was foundto be more affectively involving (M = 4.3) relative to 60 Minutes (M =2.9; t(50) = 6.00, p<.001).

Ad pretesting utilizing the cognitive and affective ad involvementmeasures confirmed that the pager ad was more cognitively involving(M = 4.1) compared to the cognac ad (M = 3.3; t(54) = 2.09, jo<.05). Incontrast, the cognac ad was found to be more affectively involving (M =4.7) relative to the pager ad (M = 3.3; t(54) = 4.51,p<.001). Ad pretest-ing notwithstanding, it should be noted that these ads are for differentproducts and this is a potential limitation of the study.

Manipulation Checks

As expected based on pretesting, 60 Minutes was found to be moder-ate-high on cognitive program involvement (M = 5.31 on 1-7 scales) andsignificantly more cognitively involving than Cheers, while Cheers wasfound to be moderate-high on affective program involvement (M = 4.92on 1-7 scales) and significantly more affectively involving than 60 Min-utes. Further, the pager ad was found to be significantly higher on cog-nitive ad involvement than the cognac ad, while the cognac ad was foundto higher on affective ad involvement than the pager ad (see Table 2).

Cognitive responses also served as an ad manipulation check. Twoindependent coders who were unaware of the expected findings agreed

Table 2Program/Ad Manipulation Checks: Means and T-Test Results

Stimulus Measure

Program Involvement

CognitiveAffective

Ad Involvement

CognitiveAffectiveCognitive Responses

Message/Brand-RelatedAd-Related

Program/Ad

60 Minutes(n=118)

5.313.96

Pager(n = 116)

3.902.29

1.102.94

Cheers(n = 116)

4.414.92

Cognac(n=118)

2.674.28

0.204.35

T-Value

6.39*6.68*

6.35*12.70*

7.80*5.72*

*p<.01.

121

JOURNAL OF BUSINESS AND PSYCHOLOGY

75% of the time regarding response categorization. In cases of disagree-ment, one of the researchers served as arbiter. Subjects generated sig-nificantly more message and brand-related thoughts for the pager adthan for the cognac ad and more ad-related cognitive responses for thecognac ad than for the pager ad. Such responding is consistent with ahigher degree of cognitive involvement with the pager ad and a higherdegree of affective involvement with the cognac ad. In sum, checks con-firmed the intended manipulation of program and ad involvement type(see Table 2).

RESULTS

Analysis of Variance

In order to test the hypotheses, two factor analyses of variance wereperformed with programs and ads as the factors and cognitive and affec-tive ad involvement as dependent variables. H1 predicts a program ef-fect for affective ad involvement such that it should be higher in theaffective program than in the cognitive program (due to priming effects).An examination of Table 3 and Figure 1 clearly supports this hypothesis(p = .005).

H2 predicts a program effect for cognitive ad involvement such thatit should be lower in the cognitive program. Table 3 and Figure 1 alsostrongly support this hypothesis (p < .001). However, given that there isan interaction effect for cognitive ad involvement it is necessary to ex-amine cell means to determine whether cognitive ad involvement de-creases for both the affective and cognitive ads when placed in the cogni-tive program. The Scheffe procedure was used to test for differencesbetween the cell means and it was found that only the cognitive ad inthe affective program was significantly different from the other cellmeans. This implies that the reduction in cognitive ad involvement oc-

122

Table 3Analysis of Variance Results for Cognitive and Affective Ad Involvement

Dependent Variable

Cognitive Ad InvolvementProgramAdProgram X Ad

F

19.9044.38

8.42

df

1,2241,2241,224

P

.000

.000

.004Affective Ad Involvement

ProgramAdProgram X Ad

8.18160.48

.08

1,2241,2241,224

.005

.000

.784

KEVIN G. CELUCH AND MARK SLAMA 123

Figure 1Cognitive and Affective Involvement Means for Each Experimental Condition

curred primarily for the cognitive ad. This is consistent with the predic-tions of McClung, Park and Sauer (1985) regarding overload effects forcognitively oriented ads embedded in cognitively involving programs.

SUMMARY AND DISCUSSION

This study was designed to explore the influence of program in-volvement on ad involvement. Program type (cognitively or affectivelyinvolving) was found to influence levels of cognitive and affective ad in-

JOURNAL OF BUSINESS AND PSYCHOLOGY

volvement. It was found that placing cognitive ads in cognitively involv-ing programs decreased the level of cognitive involvement with theseads. Affective ad involvement, however, increased for both the affec-tively and the cognitively oriented ads when the ads were placed in theaffective program.

Limitations of this research relate to the specific programs and adsused and the nature of the sample. The programs and ads employed inthis research for the sake of realism (they produce cognitive and affec-tive involvement more naturally without the unrealistic processing in-structions employed in previous research) are multidimensional stimuliwhich may differ in many ways other than those focused on in thisstudy. For example, the ads were of different length and for differentproducts, therefore differences between conditions could be due to fac-tors other than cognitive and affective involvement. Furthermore, theinclusion of additional products in the advertisements tested in futureresearch is needed to extend the generalizability of the findings beyondthe products used here. An interesting question relates to the gener-alizability of the findings to print compared to broadcast media. The useof alternative samples in related research is also warranted.

These limitations notwithstanding, the present study points to thedifferential effects programs can have on ad involvement. For the cogni-tive ad condition both types of ad involvement were enhanced by placingthe ads in the affective as opposed to the cognitive program. For theaffective ad, affective involvement was enhanced by placing the ad inthe affective program. While it must be acknowledged that ad involve-ment is only one indicator of ad effectiveness, the literature has indis-putably linked the construct to a host of related information processingoutcomes. With so much time and effort (both strategic and production-related) devoted to the creation of involving ads, it should be recognizedthat failure to consider contexts effects, particularly program effects,may mitigate their effectiveness. Specifically, the placement of cognitiveads in cognitive programs may result in diminished ad effectiveness.

If the results of the current study are replicated and extended it ispossible that advertisers would want to place a premium on affectivelyinvolving programming such as drama and situation comedy as opposedto more cognitive programming such as news shows and documentaries.This would occur to the extent that the target audience was equallyrepresented in both types of programming. As noted by Park and Mc-Clung (1986), the media planning goal related to the placement of adswithin programs to maximize ad effectiveness cannot be adequately ad-dressed through the use of program ratings and demographics withoutconsidering viewer involvement with the program in which the ad is toappear.

From a research perspective, it is important to consider research

124

KEVIN G. CELUCH AND MARK SLAMA

which would replicate and extend the current findings. If programs caninteract with ads to produce cognitive overload, then at what levels of adinvolvement and program involvement do overload effects begin to oc-cur? Is the influence of affective program involvement always positive?Might overload effects also occur for affective ad involvement at suffi-ciently high levels of affective program involvement? Future research isneeded to help determine the relationship between ad involvement andprogram involvement at different levels of each. Such research couldhelp advertisers to better predict when either positive or negative pro-gram effects are likely.

REFERENCES

Arena, W. F., & Bovee, C. L. (1994). Contemporary advertising (5th ed.). Burr Ridge, IL:Irwin.

Axelrod, J. N. (1963). Induced moods and attitudes toward products, Journal of Advertis-ing Research, 3(2), 19-24.

Bryant J., & Comisky, P. W. (1978). The effect of positioning a message within differen-tially cognitively involving portions of a television segment on recall of the message,Human Communication Research, 5, 63-75.

Celuch, K, G., & Slama, M. (1995). Cognitive and affective components of Aad in a lowmotivation processing set, Psychology & Marketing, 12(2), 45-55.

Clancy, K. J., & Kewskin, D. M. (1971). TV commercial recall correlates, Journal of Adver-tising Research, 2, 18-20.

Crane, L. E, (1964). How product appeal and program affect attitudes toward commercials,Journal of Advertising Research, 4(1), 15-18.

Goldberg, M., & Gorn, G. (1987). Happy and sad TV programs: How they affect reactionsto commercials, Journal of Consumer Research, 14, 387-403.

Kennedy, J. R. (1971). How program environment affects TV commercials, Journal of Ad-vertising Research, 11, 33-38.

Krugman, D., Reid, L., Dunn, S., & Barban, A. (1994). Advertising: Its role in modernmarketing (8th ed.). Fort Worth, TX: The Dryden Press.

Krugman, H. E. (1983). Television program interest and commercial interruption, Journalof Advertising Research, 23(1), 21-23.

Laczniak, R., Muehling, D., & Grossbart, S. (1989). Manipulating message involvement inadvertising research, Journal of Advertising, 18 (2), 28-38.

MacKenzie, S., & Lutz, S. (1989). An empirical examination of the structural antecedentsof attitude toward the ad in an advertising pretesting context, Journal of MarketingResearch, 23(5), 130-143.

McClung, G. W., Park, C. W., & Sauer, W. J. (1985). Viewer processing of commercialmessages: Context and involvement. In Advances in Consumer Research, E. Hirsch-man & M. Holbrook, eds., Provo, UT: Association for Consumer Research, 12, 351-355.

Park, C. W., & McClung, G. W. (1986). The effect of TV program involvement on involve-ment with commercials. In Advances in Consumer Research, R. Lutz, ed., Provo, UT:Association for Consumer Research, 12, 544-547.

Park, C. W., & Mittal, B. (1985). A theory of involvement in consumer behavior: Problemsand issues. In Research in Consumer Behavior, J. N. Sheth, ed., Greenwich, CT: JAIPress.

Park, C. W., & Young, S. M. (1983). Types and levels of involvement and brand attitudeformation. In Advances in Consumer Research, R. P. Bagozzi and A. M. Tybout, eds.,Ann Arbor, MI: Association for Consumer Research, 10, 320-324.

125

JOURNAL OF BUSINESS AND PSYCHOLOGY

Park, C. W., & Young, S. M. (1986). Consumer response to television commercials: Theimpact of involvement and background music on brand attitude formation, Journal ofMarketing Research, 23(2), 11-24.

Peter, J. P. (1979). Reliability: A review of psychometric basics and recent marketing prac-tices. Journal of Marketing Research, 16, 6-17.

Schlinger, M. (1979). A profile of responses to commercials, Journal of Advertising Re-search, 19(2), 37-46.

Schultz, D., & Barnes, B. (1995). Strategic advertising campaigns (4th ed.). Lincolnwood,IL: NTC Business Books.

Singh, S. N., & Hitchon, J. C. (1989). The intensifying effects of exciting television pro-grams on the reception of subsequent commercials, Psychology & Marketing, 6(1), 1-31.

Soldow, G. R, & Principe, V. (1981). Response to commercials as a function of programcontext, Journal of Advertising Research, 21(2), 59-65.

Stanford, S. (1984). Predicting favorite TV program gratifications from general orienta-tions, Communication Research, 11(4), 519-536.

126