An Experimental Study of the Effects of Commercial Tv Advertising and Pro-Consumer Product Test Results on Tv

Emmanuel J. ChTron, UniversitT du Quebec, Rimouski
Jean Perrien, UniversitT de Sherbrooke, Quebec
ABSTRACT - The relative effect of commercial TV advertising was experimentally tested in comparison with pro-consumer TV program giving product test results. The experiment involved 125 consumers who were on a shopping trip to a local store. Subjects who volunteered were asked to view a 10 minute videotape arrangement and to complete a short questionnaire. Product brand evaluation and intention to buy were the two main dependent variables. Results indicated that pro-consumer TV program had greater influence than commercial advertising. Significant interaction effects were also observed between the two sources of communication.
[ to cite ]:
Emmanuel J. ChTron and Jean Perrien (1981) ,"An Experimental Study of the Effects of Commercial Tv Advertising and Pro-Consumer Product Test Results on Tv", in NA - Advances in Consumer Research Volume 08, eds. Kent B. Monroe, Ann Abor, MI : Association for Consumer Research, Pages: 423-427.

Advances in Consumer Research Volume 8, 1981      Pages 423-427

AN EXPERIMENTAL STUDY OF THE EFFECTS OF COMMERCIAL TV ADVERTISING AND PRO-CONSUMER PRODUCT TEST RESULTS ON TV

Emmanuel J. ChTron, UniversitT du Quebec, Rimouski

Jean Perrien, UniversitT de Sherbrooke, Quebec

ABSTRACT -

The relative effect of commercial TV advertising was experimentally tested in comparison with pro-consumer TV program giving product test results. The experiment involved 125 consumers who were on a shopping trip to a local store. Subjects who volunteered were asked to view a 10 minute videotape arrangement and to complete a short questionnaire. Product brand evaluation and intention to buy were the two main dependent variables. Results indicated that pro-consumer TV program had greater influence than commercial advertising. Significant interaction effects were also observed between the two sources of communication.

INTRODUCTION

With the ever increasing influence of consumerism, the Canadian Broadcasting Corporation (C.B.C.) recently sponsored a new type of TV program designed to present test results on various products and inform consumers about their rights. This new form of communication to consumers has been an immediate success as audience figures indicated as high ratings for this program as for the most popular series. The present research is designed to test the effects of this new source of communication on consumers in contrast with traditional commercial advertising.

BACKGROUND

Previous research on persuasive communication has focused on influence of source credibility (Hovland 19S2), learning without involvement (Krugman 1965) and various types of effects (message, presentation, communicator, audience and risk level) as studied by Levitt (1965) in the context of industrial selling. The nature of these studies was to test identical messages from different sources with similar audiences. Although the Levitt study manipulated the quality of message presentation and type of audience, sources of communication were in fact identical in message content and all under control of the marketer. As stated by Levitt: "In commercial communications particularly, no two sources selling the same product to the same audience are ever likely to employ the same message".

A research by Hempel (1966) tested the effects of printed pro-consumer information such as Consumer Reports against a tape recorded commercial sales presentation. The product involved in the study were two brands of men's white dress shirts. Respondents were college students and re-suits indicated that Consumer Reports effects were greater than the commercial sales presentation. Contrary to expectation, no significant interaction was observed between the two sources of communication. The nature of the respondents and product (low ego-involvement) imposed limitations to the validity of the experiment, further, sources of communication differed as to their form (printed and verbal tape recorded). More validity could have been attained with a printed commercial advertising rather than a tape recording of a salesman.

Recent studies (Settle and Golden 1974, Smith and Hunt 1978) have linked the effects of commercial communication and attribution theory. These studies tended to focus on the processes by which individual perceive and explain causal relationships and give meaning to events. Hence these studies indicated that the effect of commercial communication was greater when product claims where not all favorable. The notion of causal schemata (Kelley 1971) introduced the idea of causal inference by which individual explain their own behavior or that of others. Individuals tend to interpret the situation (the communication) in terms of intrinsic or extrinsic factors. This process has a direct effect on inferences made by the individual about communicators' intentions. When a behavior is perceived as being caused by an external factor such as a proper payment from an advertising agency, it is most likely that there will be less confidence in the communication. However if the behavior is perceived under the influence of internal factors controlled by the actor, there is probably more confidence in the presentation.

The objective of the present research was to assess the relationships among the effects of two impersonal communication sources on consumer product brand evaluation and intention to buy. From the consumer perspective, information on products is available through personal (sales presentations, personal contacts, friends) and impersonal sources (advertising, product test results, editorials). The study was intended to experimentally compare a pro-consumer TV broadcast and commercial TV advertising both prepared on video-tapes. The design of the experiment was adapted from Hempel (1966).

PROBLEMS

This study concerns the relative effect on consumers of commercial TV information as compared with product test results such as given by consumer protection broadcasts.

More precisely, a controlled experiment was designed to answer such questions as: How do consumers react to information sources? How information sources influence product evaluation and brand choice? Does the effect of information sources vary with demographic and socioeconomic differences?

For the business community, the impact of product test results on TV may be of much consequence. Previous research by Perrien and ChTron (1979) indicated a large audience familiarity with such a program (75.3% of the respondents on an unaided recall). Further, the audience was not restricted to a socially advantaged group of consumers. Hence, if present, the effect of such information will involve more consumers than was the case with pro-consumer printed information. Following a test comparison on TV, businesses could have a strong incentive to adjust their advertising strategies and marketing mix. Facing a favorable test result, the seller could anticipate increasing demand and should adjust inventories, physical distribution arrangements and/or other marketing-mix elements. In presence of unfavorable test result, the seller could consider various advertising and/or other marketing-mix strategies to counterbalance a probable negative effect on demand.

HYPOTHESES

Formally, the following hypotheses were tested:

H1 - Pro-consumer TV broadcast information on product test results will have greater influence on product evaluation and intention to buy than a similar commercial TV advertising.

The study by Hempel {1966) supported this hypothesis, further, attribution theory suggested that when behavior is perceived under control of the actor, there is more confidence in the communication. Results of a previous survey (Perrien and ChTron 1979) indicated high confidence of respondents (mean score of 5.32 on a seven point scale) in the independent nature of information as given by pro-consumer TV programs.

H2 - There will be significant interaction between the effects of the two sources of information. The influence of one source of information will depend upon the other source of information to with the respondents is exposed.

Although the previous study (Hempel 1966) expected interaction to be significant, analysis of variance results indicated no significant departure from additivity of effects. With the expected greater influence of product test results over commercial advertising it is likely that product evaluation and intention to buy will be mainly determined by this source of communication and thus presents significant interaction with commercial advertising.

H3 - The effect of information sources will vary with demographic and socio-economic variables. Since previous research has dealt with homogeneous student subjects, this hypothesis is proposed to explore the issue of different responses with consumers of varying age, marital status, sex and education.

METHOD

The relative effect of the two sources of information were experimentally tested involving the evaluation of two brands of microwave ovens (high ego-involvement products). Subjects were one hundred and twenty five real consumers who agreed to take part in an experiment on consumption while on a shopping trip to a local food and hardware store. Arrangements were made with the store manager to reserve an enclosed area with twenty seats where videotape viewing could be organized. About one out of four consumers accepted to participate, coffee and cookies were served until about fifteen consumers were recruited for each of the nine parts of the experiment.

The experiment consisted of videotape arrangements showing typically:

1 - A commercial TV advertising for one or two fictive brands of microwave oven (TORIKO or CGX).

2 - A short cartoon unrelated with the topic.

3 - A TV product-test as given by pro-consumer TV broadcast comparing the two fictive brands of microwave oven and favoring brand TORIKO or CGX.

Special care was involved in the preparation of the communication massages. The commercial TV advertising was of the testimonial type showing two "slices of life" scenes involving a career woman, and a secretary. Motive to use microwave oven were time, money saving and safety. Form and contents of this commercial were based on an analysis of various advertising materials for different brands of microwave ovens. Production of the commercial was done by the audio-visual department at a major university and TV product test results were prepared with participation of the host of the C.B.C. TV broadcast.

Comparison of the two microwave ovens were presented on criteria such as protection against radiation, efficiency, oven design, price and overall test result was given favoring one of the two brands. This set of criteria was based on real tests presented in an issue of the "Canadian Consumer". Therefore, both types of messages would be regarded as externally valid.

Communication messages were tested with fifteen adult consumers. Subjects were recruited through the local boy' scout organization in exchange for a donation. Comparison of two brands of microwave ovens was presented as the object of the experiment and participants were given a questionnaire to fill in at the end of the viewing. Questions were asked on product brand evaluation, intention to buy, demographic and socioeconomic characteristics and overall impression and comments about the objective of the experiment. Comments indicated that respondents felt they were in a real situation and two out of fifteen indicated that the objective of the experiment was to measure the effect of the C.B.C. TV broadcast. One respondent suspected the objective of the experiment was to measure the influence of advertising.

In the store, consumers were randomly assigned to one of the nine combination treatments shown on figure 1.

FIGURE 1

EXPERIMENTAL DESIGN

As indicated, consumers might be exposed to each of nine combinations of no-message, test results and commercial advertising supporting either one or the other of the two brands of microwave ovens. In the 'no-message" situation, subjects were only shown a slide presenting a microwave oven of the appropriate brand (T0 or CGX0 or both). Commercial advertising was of the testimonial type as mentioned above, contents of the two commercials were identical except for the brand name of the microwave oven. Product test results, as presented by the host of the program, were also identical in contents except for the winning brand name. For example, subjects assigned to cell CGXADTTR would-be exposed to a commercial advertising favoring brand CGX and a product test result supporting brand TORIKO.

Presentation of a short cartoon between the advertising exposure and the test result contributed to add realism to the situation. All nine treatment cells were prepared beforehand on videotape so that presentation to subjects could be made on color TV screen without loss of continuity. The fact that some respondents were accompanied by children added a source of distraction more conform to the real TV viewing situation.

Following the exposure, each respondent completed a questionnaire designed to measure product brand evaluation and intention to buy. Other information requested were: ownership of a microwave oven, viewing and frequency of viewing test results on TV and demographic and socio-economic data. The dependent variables of the experiment were the subject's brand evaluation and intention to buy recorded on a seven-point bipolar rating scale with one of the two brands at each end of the scale.

RESULTS

Frequency data on respondents revealed that 117 out of 125 or 93.6% did not own a microwave oven and that 119 or 95.2% had seen product test result broadcasts on TV. These figures indicated a fairly homogeneous group with respect to these variables. Average age of respondents was 30, 60% were married and 57.6% were females.

Results on product evaluation and intention to buy indicated that the experimental treatment produced significant differences among the nine exposure combinations.

The two-way analysis of variance for the effects of the two types of communications on product evaluation and intention to buy is shown in table 1.

TABLE 1

TWO-WAY ANOVA FOR THE EFFECTS OF THE TWO TYPES OF COMMUNICATION MESSAGES

Hypothesis one suggested that pro-consumer test results will have greater influence on product evaluation and intention to buy than commercial advertising. As both message effects were statistically significant a measure called "omega-square" was computed: (Hays cited by Green 1978). This measure allows to-estimate the relative importance of each type of message in contributing to variation in product evaluation and intention to buy.

TABLE 2

HAY'S OMEGA-SQUARE MEASURES

Results in table 2 show the relative importance of the message variables on product evaluation and intention to buy. These measures can be roughly interpreted as R2 in multiple regression. Product test messages account for 40% of the variation in product evaluation while commercial messages and interaction account respectively for 2.3% and 7.6% of the variation in product evaluation.

Overall, 50% of the variation in product evaluation is accounted for by the two message variables. Results for intention to buy indicate a lower relative importance of product test messages than for product evaluation. However relative importance of commercial messages and interaction is higher for intention to buy than for product evaluation.

Overall, relative importance measures for product evaluation and intention to buy support the hypothesis that product test results has greater influence than commercial advertising.

Hypothesis two proposed that there will be significant interaction between the effects of the two sources of information. The analysis of variance in table 1 confirm the presence of statistically significant interaction. Graphically this can be seen on figure 2.

FIGURE 2

INTERACTION PATTERNS

Diagrams in figure 2 can be interpreted as follow; exposure to a commercial advertising in favor of brand TORIKO and viewing product test results in favor of brand TORIKO yielded a mean score of 1.83 for product evaluation and 1.6 for intention to buy. These scores were derived by taking the absolute difference between cell means and the mid-point on the seven-point scale featuring one brand at each end of the scale. (Overall grand mean of observations in the experiment was equal to mid-point score). Interaction effect is apparent as response functions depart from parallelism (beyond what might be expected by chance).

These results clearly confirm hypothesis two indicating that responses to one source of information will depend upon the other source of information to which the respondent is exposed.

Hypothesis three was exploratory in nature and proposed to test the effect of variations in demographic and socioeconomic variables on product brand evaluation and intention to buy. The effects of marital status, education and age indicated that although the effect of product-test results remained significant, the effect of TV commercial advertising tended to be less significant (confirming hypothesis one) and interaction was slightly less significant (disconfirming hypothesis two).

An analysis of covariance controlling for age as a concomitant variable resulted in an F ratio of .368 with associated probability of .546 for product evaluation, and an F ratio of .039 with associated probability of .843 for intention to buy. Clearly age was not a significant variable for measuring the effect of the two sources of communication on product evaluation and intention to buy.

The effect of sex as a discriminating variable revealed differences of responses to commercial advertising. Results of the two-way analysis of variance for males and females is shown on table 3 for product brand evaluation.

TABLE 3

TWO-WAY ANOVA FOR MALE AND FEMALE PRODUCT BRAND EVALUATIONS IN RESPONSE TO THE TWO SOURCES OF INFORMATION

As can be seen in table 3, the main effect of the commercial advertising is not significant for male respondents. Similar results were obtained with intention to buy, commercial advertising although significant for female respondents was not significant for male respondents. This result confirms the marginal effect of commercial advertising as compared with product test results. As regards interaction, this effect is still significant for both male and female respondents.

DISCUSSION

The overwhelming influence of product test results confirmed hypothesis one with the implications that businesses would be well advised to take this source of information into account for commercial strategy decisions. Relative importance of product test results tended to decrease when consumer decision was closer to the buying act (intention to buy vs. product brand evaluation) as shown by omega-square measures in table 2. Previous research (Levitt 1965) indicated similar results. In relatively higher risk situations, the audience tended to discount the highly credible source of information.

Independent variables and interaction tended overall to account for less of the variation in intention to buy (4a%) than of the variation in product evaluation (50%). This is consistent with results of a previous survey (Perrien and ChTron 1979) where the effects of product test result programs on TV were measured on the throe components of an attitude. Scores on the conative component were lower than on the cognitive and affective components indicating decreasing effects of product test results when consumer decision was closer to action.

The presence of interaction confirming hypothesis two indicated unanticipated patterns of changes. Effects on product evaluation and intention to buy tended to be higher when the two sources of information were in favor of two different brands rather than both in favor of the same brand {see figure 2). This situation may be accounted for by a defensive reaction in presence of intense persuasive messages such as observed in previous research (Hempel 1966) and by consumer attributional processes and effects in presence of messages with only positive claims (Smith and Hunt 1978).

Exploration of demographic and socio-economic variables revealed that sex was a discriminating variable for responses to commercial advertising. For male respondents the main effect of commercial advertising was not significant. This difference could be accounted for by the type of product chosen. Previous research (Capon and Burke 1977) indicated that microwave ovens scored high on a perceived risk scale (compared to non-durables lower priced items). These products were thought to be rather high ego-involving and implying a substantial money outlay with likely joint husband and wife buying decision. However microwave ovens are likely to be used more often by women and it is suspected that in the experiment, men had more difficulty to identify to the commercial advertising situation showing scenes involving women. Further research will have to be conducted to clarify this issue.

With respect to interaction, hypothesis three was still confirmed for both men and women. However, unanticipated results of lower product evaluation in presence of the two messages favoring the same brand rather than different brands was the results of male rather than female respondents.

Female respondents tended to show the anticipated pattern, scoring higher on product evaluation when both messages favored the same brand rather than different brands. This result although speculative would be in contradiction with previous research (Krugman 1965). Female respondents were probably more involved, and they were expected to have higher defense reaction than male respondents. However response delays may be too short and real choice situations too distant for such a tentative comment.

Some of this unexpected variation in product evaluation may be accounted for by different connotations associated with the two fictive brand names. In fact, mean cell re-suits indicated larger variations in product evaluation and intention to buy when brand TORIKO rather than brand CGX was involved (see figure 2). In the French Canadian environment, brand name CGX was possibly less appealing to respondents than brand TORIKO.

CONCLUSION

The effects of commercial TV advertising and pro-consumer product test results on Canadian TV program were experimentally compared.

Findings indicated the great influence of this pro-consumer source of information. Presence of significant interaction with commercial advertising implied that the net effect of the two sources of information will be different from a simple addition of the separate effects of each source of communication.

Not only sources of communication, but message contents, seemed to have different influence on male and female respondents. Further research should investigate whether the extent of ego-involvement with the product is the same for men and women and whether fictive brand names do have different connotations. In spite of these limitations it is felt that the experiment was closer to the real situation than research conducted with student subjects within university walls.

From the seller's point of view, pro-consumer product test results on TV programs introduces a new kind of uncontrollable source of sass communication influencing product evaluation by consumers. Marketers should consider this new variable for the necessary adjustments to their marketing-mix.

From the public policy maker point of view, pro-consumer TV programs giving product test results can be viewed as an efficient means to counter-balance the effect of commercial TV advertising among the wide segment of the population less likely to be reached with printed media.

However, our findings imply that magnitude of communication effects is not easily predictable. Another problem is the relative short-tern of these types of communication aggravated by the rapid turnover of brands and products on the market. Communication policy makers with consumer welfare in mind should try to reach an adequate balance between concrete, highly involving product test results and less attractive general consumer education with longer tern effect.

REFERENCES

Capon, Noel and Burke, Marian(1977), "Information Seeking Behavior in Consumer Durable Purchases", Proceedings of the American Marketing Association, 110-115.

Golden, Linda L. (1977), "Attribution Theory Implications for Advertisement Claim Credibility", Journal of Marketing Research, 115-118.

Green, Paul E. and Tull, Donald S. (1978), Research for Marketing Decisions, Fourth Edition, Prentice Hall, 369- 372.

Hempel, Donald J. (1966), "An Experimental Study of the Effects of Information on Consumer Product Evaluations", in Science Technology and Marketing, edited by Raymond Haas, American Marketing Association, 589-597.

Hovland, Carl I. and Weiss, Walter (1952), "The Influence of Source Credibility on Communication Effectiveness", Public Opinion Quarterly, Vol. 15, no. 4, 635-650.

Kelley, Harold H. (1971), "Causal Schemata and the Attribution Process", in Edward E. Jones et al., eds., Attribution: Perceiving the Causes of Behavior, General Learning Press, 151-174.

Krugman, H. E. (1965), "The Impact of Television Advertising: Learning Without Involvement", Public Opinion Quarterly, Vol. 29, no. 2, 349-356.

Levitt, Theodore (1965), Industrial Purchasing Behavior: a Study of Communications Effects, Division of Research, Graduate School of Business Administration, Harvard University, Boston, 5-11.

Perrien, Jean and ChTron, Emmanuel J. (1979), "Pro-Consumer Information on TV: Results and Implications of the Canadian Experience", 8th Annual Meeting of the European Academy for Advanced Research in Marketing, D1-D23.

Smith, Robert E. and Hunt, Shelby D. (1978), "The Effectiveness of Personal Selling Over Advertising: an Attributional Analysis", Educator's Proceedings, American Marketing Association, 158-165.

Smith, Robert E. and Hunt, Shelby. (1978),"Attributional Processes and Effects in Promotional Situations", Journal of Consumer Research, vol. 5, 149-158.

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