Corrective advertising information processing by teenagers

Journal of Consumer Srudies and Home Economics (1988) 12,49-58.

Corrective advertising information processing by teenagers

B R E N D A S T E R N Q U I S T A N D K A R E N C L A R K Human Environment and Design, Michigan State University, Michigan

Department of

In this study teenagers’ understanding of corrective advertisements was measured at three levels: beliefs, attitudes and behavioural intentions. The results support the use of beliefs rather than attitudes or behavioural intentions to detect deception.

Introduction

In the U.S., corrective advertising is one method utilized by the Federal Trade Commission (FTC) to eliminate delay profits and persistent deception. Most simply stated, corrective advertising is ‘advertising designed to inform consumers that previous advertising by the company was deceptive’.’ The exact wording and length of exposure of this advertising is generally stipulated in a consent decree. Corrective advertising was first formally proposed by SOUP (Students Opposing Unfair Practices), a group of George Washington law students who sought to intervene in a FTC action against the Campbell Soup Company.* This case concerned the soup company’s practice of putting marbles in soup shown in television commercials, giving the soup a deceptively rich appearance. Although the Com- mission declined to use corrective advertising in that case, it reserved the right to use the remedy in future cases. SOUP again sought the corrective advertising remedy in a second case, Firestone Tire and Rubber Company, with the same r e ~ u l t s . ~ The FTC has subsequently utilized corrective advertising in a number of cases as part of the consent order.

Although the purpose of corrective advertising is to change or correct pur- portedly erroneous beliefs, there is some question as to the actual comprehension versus miscomprehension the corrective messages carry. Mazis and Adkinson4 noted that 39% of their subjects misinterpreted the corrective message used in the study. Little information has been obtained about how teenagers process and in- terpret corrective advertising information.

Teenagers’ information processing The teenage years are a time of strong social development. In this transition period, individuals are asserting greater independence from the family unit but

Correspondence: Brenda Sternquist, Associate Professor, 114 Human Ecology, Michigan State University, East Lansing, MI 48824, U.S.A.

49

Corrective advertising

greater dependence on peer approval. Studies on teenagers’ consumer behaviour have mainly been based on two models: the cognitive development model and the social learning model. The cognitive development model centers around the function of age as a proxy variable for cognitive development. The effects of adolescent maturation on consumer behaviour have been examined by several researchers.”’ The basis of the social learning model is the formation of consumer behaviour based on an interaction of various sources of influence such as parents, peers and mass media.

Moschis’ investigated both affective and cognitive orientations. Age was used as an index of cognitive development, and mass media, parents, peers and schools were investigated as socialization agents. The findings of that study indicated that brand preferences are acquired as a result of teenagers’ interaction with mass media rather than through increasing consumer experience as the child matures. A strong predictor of teenage attitudes toward advertising was peer communication about purchasing products; this suggests that the development of favourable attitudes about advertising is based on an interpersonal process. The study also found that teenagers become more sceptical of advertising claims and develop a greater cognitive defence toward them as they grow older.

Beliefs, attitudes and behavioural intentions

Most of the studies dealing with deceptive and corrective advertising have used the theoretical framework developed by Fishbeim8 He views the formation of attitudes as part of the learning process in which beliefs are associated with various objects. As an object is associated with various positively or negatively perceived attributes, characteristics, or qualities, the individual acquires a positive or negative attitude toward the object. Thus, at any point in time, an attitude toward an object is a function of beliefs about the object, and the evaluative aspects of those beliefs.

Although evidence seems to support the predictive quality of the Fishbein attitude model, marketing research is really concerned with the prediction, not of attitudes, but of consumer behaviour. Fishbein’ theorized that behavioural intentions are influenced not only by attitudes, but by social or normative considerations as well. Fishbein has represented this behavioural intention model as:

B = BI = Aact + [(NB),,I(MC),,],

where: B = overt behaviour; B1 = behavioural intention; Aact = attitude toward an act; NB = normative belief; MC = motivation to comply with the normative belief; wo and w1 = empirically determined weights.

This model added two new components to the original Aact Fishbein equation, a normative belief (NB) component and a motivation-to-comply (MC) component, to predict behavioural intentions. A relative weighting system has been added to components because it is theorized that intention to engage in certain

50

B. Sternquist and K. Clark

types of behaviour is more strongly influenced by social o r normative considerations than intention to engage in other types of behaviour. For example, social or normative considerations may have a greater part in influencing an individual to buy a particular brand of status-related items such as designer jeans, than rela- tively status-free items such as milk.

In this study, teenagers’ understanding of corrective advertisements was measured at three levels: beliefs, attitudes and behavioural intentions. The normative belief approach used by Gardner” was used to compare subjects’ beliefs about the products. The original Fishbein model (Aact) was used to measure attitudes and the extended Fishbein model (BI) was used to measure behavioural intentions.

Methods

The products

Two product groups selected for use in this study were from among those pre- viously investigated or currently under consideration by the FTC for possible liti- gation. These products are mouthwashes and footwear. Actual advertisements for these products were used in the initial phase of the study, while actual advertisements with a corrective advertising mock up were used in the second phase of the study. Actual products and advertisements rather than fictitious ones were chosen because consumers make decisions based on information and experience. Sawyer” urges the use of real products in corrective advertising research because only real products allow the measurement of corrective advertising effects on eradicating residual beliefs of previous exposure to advertising of the products.

The specific brands used in this study and the evaluative criteria were determined through focus-group interviews with 31 students. The focus groups consisted of six to eight participants who discussed the attributes they considered important when buying the product. These discussions were tape recorded and later tabulated for response frequencies and evaluative criteria saliency. In addition to the focus groups, FTC rulings on the product attributes related to deception were used to establish the salient evaluative criteria.

Operationalization of the variables

Behavioural intentions (BI) were measured by subjects’ responses to questions of whether or not they intended to buy a particular brand. Respondents indicated their intention on a scale from ‘I will not buy this brand’ (0) to ‘I will buy this brand’ (61, with 3 being the neutral response. Subjects were asked to respond to intentions to buy at the brand level, rather than at the product class level, since Dulany’’ found that the more specific the intention becomes, the higher the correlation with a specific behaviour.

51


Affect (A) was measured by scaling the importance of the evaluative criteria. The respondent was asked to scale the importance of the various criteria from ‘Very unimportant’ (0) to ‘Very important’ (6). Since the importance of the evaluative criteria would be generalized across the product level, the response was solicited at that level.

Beliefs (B) were measured by a response to the likelihood that specific brands would contain each attribute. Responses at the brand level were from ‘Extremely improbable’ (0) to ‘Extremely probable’ (6).

Normative belief was measured by asking the subjects’ estimates that other people who are important, such as friends, parents and doctors, would recom- mend buying each brand. Responses ranged from ‘Extremely improbable’ (0) to ‘Extremely probable’ (6). In this study a general referent was used.

The motivation-to-comply component was measured by asking the respondents to rate the importance they placed on others’ advice when selecting the product. Responses ranged from ‘Don’t care at all’ (0) to ‘Care a great deal’ (6).

Experimental design

The sample consisted of public high-school students from two cities in Michigan. A two-stage non-equivalent pre-test-post-test design was used in the study. The pre-test-post-test component of the design allows for measurement of differences in attitude scores, while the control group provides increased internal validity. Campbell and Stanley13 consider this design to be acceptable when random assignment of groups is not feasible; in this study the groups constitute naturally assembled classrooms rather than randomly assigned assemblages.

Only female students were assigned to the footwear product group because the brands used in the study were marketed to teenage girls. Both sexes were in- cluded in the mouthwash study. Subjects only responded to one product group.

Members of the sample were assigned by classroom to one of two groups - experimental or control. The pre-test questionnaire was administered to both groups. One week later, the experimental group was shown an advertising treatment, referred to as Experimental Treatment Number 1, and the first post-test questionnaire was administered to both groups.

One week after the first post-test, a second advertising treatment, Experimen- tal Treatment Number 2, was given to the experimental group. The second post- test questionnaire was administered to both groups.

Advertising treatment The advertisements used in Experimental Treatment Number 1 consisted of actual advertisements for four different brand items of one product class. The advertisements were obtained from magazines designed to be sold to the

52


sample group. One advertisement contained an allegedly deceptive statement. The allegedly deceptive statements by product class were:

Mouthwash. ‘Two-fisted. A real fighter. Kills germs by millions. So strong it keeps breath fresh for hours . . . and hours. (Brand Name) antiseptic. Kills germs by millions on contact. For bad breath, colds and resultant sore throats.’

Footwear. ‘What you don’t feel is the toe grip at work. It makes your toes “hang on.” Which exercises your legs to help firm and tone them while you walk.’

Experimental Treatment Number 2 also consisted of four advertisements. Three of the advertisements were identical to the three ‘non-deceptive’ advertisements in Experimental Treatment Number 1 , while the allegedly deceptive advertisement was replaced by a ‘corrective advertisement’. The corrective advertisements were identical to the original advertisements except that the allegedly deceptive portion was omitted and a corrective portion inserted. The corrective portion was modelled in format after the Listerine corrective advertisements. The corrective messages were:

Mouthwash. ‘(Brand name) will not help prevent colds or sore throats or lessen their severity.’

Footwear. ‘(Brand name) will not firm or tone legs.’ Each participant was provided with a black and white photostatic copy of the advertisements, while a colour slide of the same advertisement was projected on a screen. They were allowed to refer freely to the advertisements as they completed the questionnaire. This eliminated retention of material as a variable in the study.

Results and conclusions

Beliefs

For each product category, two beliefs were selected for investigation, one being the potentially deceptive belief under study. Another belief closely related to the potentially deceptive belief was selected to examine whether deceptive or corrective statements about one attribute will affect beliefs about other attributes. Mean brand belief scores on the pre-test, first post-test and second post-test were com- puted for each brand for control and experimental groups. Mean scores for the specific brand advertising the potentially deceptive claim were compared with a normative belief score consisting of an average mean score of the other three brands studied. This procedure is termed the ‘normative belief technique’ by Gardner. “’

53


Mouth wash For mouthwash, the cold-prevention belief scores for Listerine were compared with the average of the three other brands. The comparison of the belief scores for the brand under investigation with the means of the three other brands in the product groups is presented in Fig. 1. In the first stage of the study, the mean brand belief of the group viewing the Listerine experimental treatment increased slightly, as did the norm. In both experimental and control groups, the belief about cold prevention power was higher for Listerine than for the other brands of mouthwash. In the second stage, the mean brand belief scores dropped closer to the normative belief score, indicating that the experimental group now perceived Listerine as about as likely to prevent colds as the other brands of mouthwash studied.

The belief that Listerine kills germs may be related closely by some students to the cold-prevention claim. The brand belief for the Listerine control group re- mained higher than the normative belief in all phases of the experiment, while the

( a )

- Experimental Control --_--

6.0 5.5 5.0

Norm* A-A A 3.0

Pre-test Past-test I Post-test 2

;:: 1 3.0

Pre-test Post-test I Post-test 2

Fig. 1. Normative belief technique. (a) Belief that mouthwash prevents colds.

(b) Belief that mouthwash kills germs. Norm* mean represents an average of beliefs in the other three brands.

54


experimental group score increased during the first stage of the experiment and dropped to about the same as the normative belief during the second stage. This pattern follows closely the pattern of the cold-prevention belief.

In the second stage of the experiment, in which all subjects saw the corrective advertisement disclaimer for cold-prevention capabilities, a ‘corrective’ effect was noted for both the cold-prevention belief and the germ-killing belief. The beliefs that Listerine prevents colds and kills germs more closely approximated the beliefs for the other brands studied in the control and experimental groups for germ-killing belief and in the experimental group for the cold-prevention belief. Three other studies’”” have noted this ‘reverse halo’ effect in which a corrective statement for the cold-prevention belief also affected the germ-killing belief. This may be due to a strong association in the minds of consumers between germ- killing power and cold prevention. It may also illustrate the inability of a corrective advertisement to ‘surgically’ remove one deceptive belief without negatively affecting other beliefs about the brand.

Footwear

The comparison of the belief scores for the brand under investigation with the means of the three other brands in the product groups is presented in Fig. 2. The belief that Scholl exercise sandals provide leg exercise was found to be much higher than the normative belief for both the control and experimental groups during the first stage of the study. The belief level dropped only slightly for the experimental groups after administration of the corrective statement focusing on this belief. This may indicate a strongly held a priori belief by the respondents. The differences between this belief for Scholl and the normative belief on the pre- test would lend further evidence to this explanation.

The mean brand belief that Scholl has health value for feet was greater than the normative belief for both control and experimental groups in the first stage of the study. The mean brand belief of the Scholl experimental group fell slightly below the normative belief during the second stage of the study. As in the Listerine germ-killing belief, this would indicate a “reverse halo” effect, since the belief of health value for feet was not addressed in the “corrective” portion of Experimen- tal Treatment 2.

Attitudes and behavioural intentions

Multiple regression was used to analyse the effectiveness of the Fishbein models in predicting attitudes and behavioural intentions, as measured by the questionnaire. The original Fishbein attitude model was found to be a significant predictor of attitudes towards all of the product brands tested. The Fishbein Behavioural Intentions model was also found to be a significant predictor of behavioural intentions to purchase all of the product brands (Table 1).

55


I ( O ) - Experimentol Control ----

6.5 6.0 .A

/---

Scholl -:---- /-*-

Scholl &-I- ________ P--- A

---___ Norm* --- Norm* A

4.5 4.0

3 3*5* *O Post-test I Post-lest 2 Pre-test

6.5 7-0 161 --a ---- -_--

/A------ Schol I &----------

4.5 Schol I - 4.0 Norm‘ A

3.0 3.5 t Pre-test Post-test I Post-test 2

Fig. 2. Normative belief technique. (a) Belief that footwear gives leg exercise.

(b) Belief that footwear has health value for feet. Norm* mean represents an average of beliefs in the other three brands.

Attitude change and change in behavioural intentions were measured by mean gain scores. Control and experimental groups were compared by the use of t-tests. In no case was there a significant change in attitude or behavioural intention found after exposure to either of the advertising treatments.

Summary and Conclusions

The results of this study indicate that the Fishbein models may be too gross a measure to be used in identifying deception and the effects of corrective efforts. This is consistent with Wilkie’s’’ suggestion that attitudes and behavioural intentions are unsatisfactory levels of response at which to measure deception. He states that these measures are “too general to be attributed to exposures to specific messages” (p. 198). Wilkie suggests that measurement of advertising should occur at the brand-belief level.

56


Table 1. Multiple regression analysis of Fishbein attitude model

Aact Model BI Model Product R2 F R2 F

Mouthwash Scope 0.25 Listerine 0.16 Lavoris 0.23 Listermint 0.30

Footwear Scholl 0.18 Famolare 0.12 Fanfare 0.27 Earth Shoes 0.29

34.0* * * 18.4*** 30.4*** 42.5***

17.2*** 10.8** 38.9*** 31.0* * *

0.28 0.16 0.23 0.32

0.21 0.14 0.28 0.33

38.0* * * 18.81 * * 30.2*** 48.02***

20.1*** 12.0* * * 29.9*** 37.3% * *

***P < 0.001; **P < 0.01.

The use of analysis of beliefs alone is even more important to identify the carry- over effect of a corrective advertisement designed to alter a specific belief, which in reality may affect a related belief. In the case of the mouthwash product group, the corrective message that disclaimed the cold-prevention capabilities for Lister- ine also resulted in a carry-over to the belief about germ-killing.

References 1.

2. 3. 4.

5.

6.

7.

8.

9.

10.

Mann, R.A. & Gurol, M. (1978) An objective approach to detecting and correcting deceptive advertising. Notre Dame Lawyer, 54, 73 -101. 77 FTC 664, 1970. 81, FTC 398, 426, 1972. Mazis, M.B. & Adkinson, J.E. (1975) An experimental evaluation of a proposed corrective advertising remedy. Journal of Marketing Research, 13, 178-183. Moschis, G. & Moore, R. (1982) A longitudinal study of television advertising effects. Journal of Consumer Research, 9, 279-286. Moore, R., Moschis, G. & Lowndes, S. (1975) Some communication and demographic de- terminants of adolescent consumer learning. Journal of Consumer Research, 2, 80-92. Soldow, G. (1983) The processing of information in the young consumer: the impact of cognitive developmental stage of television, radio and print advertising. Journal of Adverfis- ing, 12, 4 1 4 . Fishbein, M. (1967) Attitude and prediction of behavior. In: Readings in Aftifude Theory and Measurement (Ed. by M. Fishbein), pp. 477-492. John Wiley and Sons, New York. Fishbein, M. (1975) Attitude, attitude change, and behavior: a theoretical overview. In: Affitude Research Bridges the Aflantic (Ed. by Philip, Levin, Ogilvy and Mather), pp. 3-16., American Marketing Association, Chicago. Gardner, D.M. (1976) Deception in advertising: a receiver oriented approach to understanding. Journal of Adverfising, 5 , 5-11.

57


11.

12.

13.

14.

15.

16.

17.

Sawyer, A. (1977) Measuring communication effectiveness of deceptive and corrective advertising. In Moving A Head with Attitude Research (Ed. by U. Wind and M. Green- berg), pp. 190-194. American Marketing Association, Chicago. Dulaney, D. (1973) Hypotheses and habits in verbal operant conditioning. Journal of Abnormal and Social Psychology, 63, 9-12. Campbell, D. & Stanley J , (1963) Experimental and Quasi-Experimental Designs for Research. Rand McNally College Publishing Company, Chicago. Mazis, M. & Adkinson, J. (1976) An experimental evaluation of a proposed corrective advertising remedy. Journal of Marketing Research, 13, 3&52. Kuel, P. & Dyer, R. (1976) Brand belief measures in deceptive-corrective advertising: an experimental assessment. In: Marketing: 17761976 and Beyond (Ed. by K. Bernhardt), pp. 373379. American Marketing Association, Chicago. Dyer, R. & Kuehl, F. (1978) A longitudinal study of corrective advertising. Journal of Marketing Research, 15, 39-48. Wilkie, W. (1973) Research on counter and corrective advertising: In Advertising and the Public Interest: Selected Papers f rom the Conference on Advertising and the Public Interest (Ed. by S. Divita), pp. 189-202. American Marketing Association, Chicago.

58

Documents

Corrective advertising information processing by teenagers