Consumer Perception of Advertising Effectiveness and Life-Style-Congruency

D. W. Greeno, University of Toronto
G. H. Haines, Jr., University of Toronto
M. S. Sommer, University of Toronto
ABSTRACT - A sample of Canadian housewives was exposed to three levels of life style advertising message congruency and then asked to rate the advertisements along a number of traditional additudinal pre-test rating scales of effectiveness. The results indicate that subjects do not rate the most congruent communication message as the most effective. In fact, the one life style "type" investigated here rated the most "incongruent" advertisement as being the most effective. On the basis of pretest rating scales one might conclude that message-receiver congruence may not be a valid marketing communications strategy in all cases.
[ to cite ]:
D. W. Greeno, G. H. Haines, Jr., and M. S. Sommer (1976) ,"Consumer Perception of Advertising Effectiveness and Life-Style-Congruency", in NA - Advances in Consumer Research Volume 03, eds. Beverlee B. Anderson, Cincinnati, OH : Association for Consumer Research, Pages: 328-333.

Advances in Consumer Research Volume 3, 1976      Pages 328-333

CONSUMER PERCEPTION OF ADVERTISING EFFECTIVENESS AND LIFE-STYLE-CONGRUENCY

D. W. Greeno, University of Toronto

G. H. Haines, Jr., University of Toronto

M. S. Sommer, University of Toronto

ABSTRACT -

A sample of Canadian housewives was exposed to three levels of life style advertising message congruency and then asked to rate the advertisements along a number of traditional additudinal pre-test rating scales of effectiveness. The results indicate that subjects do not rate the most congruent communication message as the most effective. In fact, the one life style "type" investigated here rated the most "incongruent" advertisement as being the most effective. On the basis of pretest rating scales one might conclude that message-receiver congruence may not be a valid marketing communications strategy in all cases.

INTRODUCTION

The research tradition in consumer behavior has evolved from one that was molecular in nature to a current one that is more molar. Research focused narrowly on such individual constructs as personality, social class, reference groups, or self image in the molecular tradition These constructs, however, were not integrated in an operational sense until the life-style analytic format (Wells and Tigert, 1971; Tigert, 1972), a more molar approach, was developed.

However, whether the research tradition was molecular or molar, managerial prescriptions were offered. These prescriptions, while apparently reasonable, were rarely empirically validated. For example, one implication for marketing communications that has often been stated is that the message sender should reinforce, in a congruent fashion, a receiver's personality, social class, reference group, self image, or life-style profile. The marketing communications literature (Kernan, Dommermuth and Sommers, 1970) however does not directly demonstrate that message-receiver congruence is a valid communication strategy. One previous study of the congruence principle which worked with personality variables (need for achievement and affiliation) found very mixed results indeed (Whittaker, 1972). This paper reports on an experiment designed to test the validity of the message congruence prescription. The research was designed to address the question: would information be processed in a more favorable manner if the congruency rule with respect to life style is followed? That is, is the prescription of product-buyer congruency valid?

Theory

Even a cursory view of the task of extracting information from an advertisement would suggest that a major component of such a task is the processing of sensory, particularly visual, information. The advertisement perception and processing task can be characterized as consisting of three elements:

1. detection - the environmental information necessary to notice the advertisement must be available to the consumer. Available evidence (Ackoff and Emshoff 1975, p. 98) suggests this is not always the case

2. diagnosis - the consumer must formulate a set of alternative actions available, and

3. decision - the consumer must select and accurately execute the optimum (in terms of the consumers' own goals) alternative within a specific time span.

This section describes briefly a theory of perceptual information processing, due to Broadbent (1971) and Sokolov (1969), which has been successfully applied to predicting effective driver decision making by Mihal (1974). This theory can be used as a conceptual basis for the identification of a number of predictions of consumer reactions to advertising, and also as a basis for explaining how these predictors function. It should be noted other theories are available, such as that proposed by Wright (1974). The justification for using the Broadbent-Sokolov theory is a previous research base (see Sokolov 1969 and Mihal 1974 and references therein) which supports its validity as a general model of perceptual information processing.

Overview of the Model. Any given possible state of the environment can be characterized by various stimulus features. Stimulus features combine systematically to produce a stimulus event. Incoming data from stimulus events are transmitted to an individual's sensory register, where the process of information analysis begins. The analysis at this point appears to be primarily on physical and/or descriptive dimensions, such as color, spatial relations, and time. Sensory data are then transferred to short-term memory. The sensory data is then further analyzed and categorized in short-term memory. Next it is held in category states which, roughly, correspond to response alternatives. Then it is transferred to long-term memory or lost from the system. Loss is strongly influenced by the limited capacity of short-term memory. Attention, stress, fatigue, visual interference, prior knowledge, and the number of perceptual categories available all influence the transfer of information into short-term memory.

Three major processes described by Broadbent as occurring in this system appear to be important. First, the system appears to continuously monitor incoming data for novelty or change. The intensity of the monitoring depends on the amount of data coming in, and-on the expectation that a change will occur. Thus, if a person watching television does not expect, on the basis of past experience, to see an advertisement, it is quite possible he or she "won't even see" the advertisement even though he or she is "looking right at it". This process appears to be based primarily on an individual's behavior perceptual skills, that is, his or her ability to detect relevant data in the environment even when the event rate (i.e., the rate of incoming data) is excessive. Therefore, one crucial element at an individual level is perceptual ability, or effectiveness in separating relevant from irrelevant data.

The second process of interest is that the speed or ease of categorizing information in either the sensory register or short-term memory is a function of both the number and availability of category states. This is equivalent to Bruner's (1957) concept of "perceptual readiness". Thus, as perceptual experience with a variety of advertising exposure situations is gained, less information of greater relevance will require processing for the categorization of events. Also, more categories could be available for such classification earlier in the perceptual process. This second process relates to the diagnosis stage of the perceptual task, and suggests a second individual difference that might be related to advertising perception.

FIGURE 1

SCHEMATIC OF BROADBENT-SOKOLOV PERCEPTUAL INFORMATION PROCESSING MODEL

Finally, reaction time appears to be not so much a function of the amount of incoming data, but rather of the transmission of evidence about the nature of environment. Latency is regarded as the result of the time necessary for unreliable evidence to accumulate to a point where the individual believes the probability of an error in response falls to an acceptably low level. If criteria for reaction are set to allow responses on relatively little evidence, reaction will be fast but inaccurate. It is not simple physical motor reaction speed alone which accounts for the speed of decision making.

Given the effect on decision making of the three factors Broadbent discusses, an experiment was designed to attempt to test the congruency hypothesis as well as to explore the number of category states consumers typically use in perceiving advertisements.

Research Design

A national probability sample of 686 English-speaking Canadian housewives completed a 67 item life style instrument. The data were analyzed to discover life style differences among users and non-users of convenience foods. Because of the product market being studied, the instrument represented items which reflected food preparation, homemaking, family life, out of home social behavior, and role-related items. An R factor analysis yielded 18 factors from the 67 items explaining 53.9% of the variance A Q factor analysis along the 18 factors created three life style groups (an arbitrary 0.30 cutoff point on a group factor loading was used to allocate

subjects). A multiple discriminant analysis of the three groups was then performed to test the reliability of the classification system. 84% of the subjects were correctly classified in this procedure, which is a very acceptable level. The groups as constructed were viewed as potentially unique market segments for purposes of differentiating product image. The groups, labeled Specialty, Simplicity, and Basic, were comprised of subjects who had in common such R factors as:

Specialty enjoyment of outside activities, enjoyment of cooking and entertaining, enjoyment of dining;

Simplicity making cooking easier, liking prepared foods, lacking in organization;

Basic economy consciousness, enjoyment of family life, enjoyment of traditional meals.

The next step in preparation for the study of the congruency principle consisted of creating three print advertisements. An advertising agency was instructed to develop three communications, each congruent to the lifestyle groups above. The content of the product concept in life-style format was based on the four or five most heavily loaded factors which characterized each of the three market segments. Creating advertisements and/or product concepts in this manner conforms with the recommended and/or ordinary "creative" practice for the implementation of life-style research. Illustrated below in Figure 2 are the advertisements with the three lifestyle treatments. These advertisements illustrate how the life-style factors were operationalized by the creative people. Note the emphasis on outside activities and enjoyment of cooking and entertaining in the Specialty treatment; the simplicity and ease of preparation for the Simplicity treatment and the family and economy orientation in the Basic treatment. The most important factors characterizing that segment provide the lifestyle context of the product concept for each treatment.

The three treatments were then presented to 99 housewives who fell into the Specialty life-style category. These subjects were classified by submitting their 67 lifestyle scores from the original research instrument to the original discriminant function. Each subject was asked to scale for the experiment: (1) each of the advertisements along ten dimensions (see Table 1) on a 6 point Likert type scale, and (2) to rank the advertisements in terms of their perceived effectiveness to get the subject to try the product. No time constraint was placed on this task in an effort to eliminate reaction time as a variable in the perceptual data processing task. For comparison purposes, the subjects were then exposed to all three advertisements simultaneously. The specialty ad was held to be congruent to the subjects life-style. The simplicity advertisement was judged neutral; the basic value advertisement was judged incongruent.

FIGURE 2

PRODUCT CONCEPT TREATMENT

Hypotheses

It can be seen that message congruency does not enter in any obvious way into the perceptual information processing theory. Thus, the theory predicts Hypothesis 1: Specialty seeking life style consumers will not rate advertising messages designed to be congruent with their life style as being superior to messages that are incongruent or neutral.

TABLE 1

PRE-TEST ATTITUDE AND OPINION STATEMENTS FOR THE ADVERTISEMENTS

The second process of interest relates to category states. Previous research (Lucas and Britt 1963) suggests there is, in a perceptually experienced consumer, only one category state, which seems to roughly correspond to a liking-disliking categorization. This suggests Hypothesis 2: Specialty seeking life style consumers will be found to have only one category state for the advertisements they view.

Analysis and Findings

Hypothesis one was tested by performing a one way hypothesis of variance on the attitude ratings for the 99 subjects taken when they were looking at all three advertisements. Table 2 presents the results. The null hypothesis of similarity is rejected in the seven dimensions (out of the ten) shown in Table 2. However, it is evident that these results do not support the conventional wisdom. The Basic food and value message is rated considerably higher on all seven of these attributes, compared to the Specialty oriented (congruent) and Simplicity directed messages. These results suggest that an incongruent advertisement would be best.

The number of attitude scales that are significantly different for these messages may be quite deceiving. On first glance it is possible to be tempted to conclude that the incongruent messages are superior along several major underlying dimensions (category states) of pretest advertising effectiveness. Therefore, a factor analysis of the items was carried out on the ten attributes. This analysis revealed only one factor with an eigenvalue greater than one.

This eigenvalue, which had a value of 5.12, explains 51.1 percent of the variance. There is no evidence to reject Hypothesis 2.

Discussion and Further Analysis

The main hypothesis, that specialty seeking life style consumers would not rate specialty advertising messages as being superior, cannot be rejected. However, such consumers appear to rate a basic value advertising message as being superior - an unexpected result. Further, the rejection of the statistical null hypothesis appears to be on the basis of a single underlying dimension or category state. It appears that consumers do not rate or rank an advertisement along several independent category states, but that an overall "halo" effect exists. The overall impression of the family-economy orientation was rated more effective in this case than the congruent "out-and-about" specialty oriented message by consumers who were specialty types in terms of their self-reported lifestyles! The notion of congruency must be questioned.

This result can certainly be rationalized, once it has been observed, in terms of the underlying theory. An incongruent or neutral advertisement represents novelty or change to the consumer, who is therefore more likely to detect it. However, further experimental work is clearly in order. One set of data is not enough.

At this stage of research it is too early to conclude that the congruency principle should never be followed. Aside from the fact that the above experiment dealt only with one product in one cultural setting, the validity of opinion and attitude ratings, and order-of-merit rankings, must be questioned (Lucas and Britt 1973). There may be a strong element of "social desirability" manifested in the ratings. The consumers studied here were not family oriented, economy conscious, nor did they care a great deal (relative to the other lifestyles) about nutritional value. As a result when exposed to what they might perceive as "more" culturally acceptable standards they may have rated the advertising along a "desirability" rather than an "effectiveness" dimension.

Another possible explanation for both the above results and the traditional congruency rule is that consumers' may use non-salient (in a choice among products sense) product attributes to adjust or create a product image so it is congruent. That is, consumers may divide product attributes into two category states: attributes that are used to make a marketplace choice and attributes that are used to create a product image congruent to the consumers' lifestyle.

When these same ads were used in another study, certain inferred product attributes were superior for the congruent advertising messages. That is, when asked to project attributes of the product concept represented by the advertisement when seen in isolation, certain perceived product attributes were rated so as to uphold the congruency principle. Table 3 is a summary table of rank correlation coefficients between advertisement

TABLE 2

RELATIONSHIPS BETWEEN MEAN PRE-TEST ATTITUDE RATINGS AND MESSAGE CONGRUENCY

TABLE 3

RELATIONSHIPS BETWEEN PRODUCT CONCEPT CONGRUENCY AND PRODUCT ATTRIBUTES (SPEARMAN RHO)

congruency and perceived product attributes. Specialty was rated one, simplicity, ten, and basic value, three, in computing these statistics. Unfortunately, there is no independent determination of which product attributes were perceived by the consumers to be salient in a choice situation. These results, therefore, could support the image congruence hypothesis that salient product attributes should be presented in an advertisement to be congruent to the receiving consumers' lifestyle.

CONCLUSIONS

The congruity principle has not been supported under the conditions described here. However, when the validity of attitude and opinion ratings is reviewed, the issue of social desirability is confronted, and the fact that certain projected or inferred product attributes are rated superior for congruent messages, it is clear that much more research is required. Still, it does appear that if practitioners are going to use lifestyle in their creative advertising strategies that the congruency principle should be carefully reviewed each time in terms of the particular institutional facts known about the particular product.

REFERENCES

Ackoff, Russell L. and James R. Emshoff, "A Reply to the Comments of Yvan Allaire", Sloan Management Review, 16 (Spring, 1975), 95-98.

Broadbent, D. E., Decision and Stress (New York, N.Y.: Academic Press, 1971.

Brunet, John S., "On Perceptual Readiness", Psychological Review, 64 (March, 1957) 123-152.

Kernan, Jerome B.; William P. Dimmermuth, and Montrose S. Sommers, Promotion: An Introductory Analysis (New York, N.Y.: McGraw-Hill, 1970).

Lucas, Darrell B. and Stewart Henderson Britt, Measuring Advertising Effectiveness (New York, N.Y.: McGraw-Hill, 1963)

Mihal, William L., "Individual Differences in Perceptual-Information Processing and Their Relation To Accident Behavior", unpublished Ph.D. Dissertation, The University of Rochester, Rochester, New York: 1974.

Sokolov, E. N., "The Modeling Properties of the Nervous System", Ch. 23 in: Cole, Michael and Irving Maltzman, eds., A Handbook of Contemporary Soviet Psychology. (New York, N.Y.: Basic Books, 1969).

Tigert, Douglas J., "Are Television Audiences Really Different?", in: Fed C. Allvine, eds., Combined Proceedings 1971 Spring and Fall Conferences, American Marketing Association. Chicago, Illinois: American Marketing Association, 1972).

Wells, William D. and Douglas J. Tigert, "Activities, Interests, and Opinions", Journal of Advertising Research, 11 (August 1971) 27-35.

Whittaker, William, "The Relationship Between Individual Difference Variables, Media Usage, and Product Choice", unpublished Ph.D. Dissertation, The University of Rochester, New York: 1972.

Wright, Peter L., "On The Direct Monitoring of Cognitive Response to Advertising", pp. 220-248 in: Hughes, G. David and Michael L. Ray, eds. Buyer/Consumer Information Processing (Chapel Hill, North Carolina: University of North Carolina Press, 1974).

---------------------------------------