The effects of actors' race in commercial advertising: review and extension.

In the mid-1960s, advertisers faced a sensitive social and economic issue, the possibility of including black actors with white actors in commercial messages. This new form of promotion, called "integrated" advertising, became a societal concern, and a focus of marketing strategy for the next

One purpose of this paper is to evaluate previous studies of white and black viewers' responses to advertisements featuring actors of different races. Unfortunately, because much of this early research does not meet today's standards of methodological rigor, the conclusions of these studies must be questioned. The second purpose of this paper, then, is to present results of two recent studies that examine the effects of actors' race in advertising messages.

Limitations of Previous Research

Table 1 provides a summary of previous studies which have examined the effects of actors' race in advertising messages. By and large, findings of these studies suggest that white consumers do not show extremely negative reactions as a result of exposure to black models in promotional material. Furthermore, when black actors were included in advertisements, black consumers were better able to recall the advertisement's content, and had more positive affect toward the advertisement and the actors.

Unfortunately, methodological problems render the majority of these studies' findings uncertain. Many of the studies failed to consider issues of 1) demand characteristics, 2) carry-over effects, 3) problems using real advertisements (i.e., variability in exposure, uncontrolled sources of variation), and 4) the necessity of using a control group. The two experiments discussed here used traditional experimental design methods to examine the effects of actors' race on viewers' attitudes and purchase intentions. Current theories of persuasion are used to explain the effects of source similarity or dissimilarity in advertising stimuli.

Viewers' Processing

of Communicator

Characteristics

Chaiken (1980) proposed that message recipients may used systematic or heuristic strategies when considering the validity of a message's conclusion. She argued that when recipients engage in systematic processing, they devote considerable effort to attending to, comprehending, and evaluating the message's arguments. In contrast, when recipients engage in heuristic processing, they exert little cognitive effort in determining the validity of a message; recipients accept the conclusion of a message based on simple rules or heuristics that they have learned from past experiences. For instance, when viewing an advertisement featuring an actor of differing race, racially conscious viewers may employ a dissimilar-disagree heuristic such as "he is not like me so it is hard to agree with him."

Kelman (1961), in his theory of social influence, argued that the identification process occurs when individuals change their opinion because they want to maintain or establish a satisfying, self-defining relationship with the communicator. Kelman's (1961) identification model suggests that racially sensitive viewers (e.g., whites who are prejudiced against blacks and blacks who identify strongly with black culture) would experience a negative identification process in which their overall feelings toward an actor of a different race would be unfavorable. These racially sensitive viewers would probably not accept the actor's testimony about a product or service.

The present research (Whittler 1989; Whittler and Scattone, forthcoming) is based upon the notion that people will often use simple decision rules when presented with salient communicator cues that are generally processed in heuritistic fashion (see Chaiken 1987; Chaiken, Liberman, and Eagly 1989). Since black actors appear primarily in minor and background roles (Wilkes and Valencia 1989), the black actor's race may be a salient communicator characteristic, especially for persons concerned with racial issues. For these individuals an actor's race should be a positive cue for racially similar viewers but a negative cue for racially dissimilar viewers.

Hypotheses

Based upon the preceding discussion, two experiments were conducted to examine viewers' processing of racial cues in advertising stimuli. In both studies it was predicted that high-prejudice whites (and high-identification blacks in Study 1) would react more strongly to racial cues than low-prejudice whites (and low-identification blacks) on source-processing measures (i.e., perceived similarity, symbolic role identification). Specifically, the following hypothesis was tested:

[H.sub.1]: Two-way interactions are predicted (i.e., Actor's Race X Racial Attitudes Toward Blacks). High-prejudice whites (versus low-prejudice whites) will perceive themselves as less similar to, and identify less strongly with, black than white actors. Also, high-identification blacks (versus low-identification blacks) will perceive themselves as more similar to, and identify more strongly with, black than white actors.

Moreover, if the first hypothesis based upon the systematic/heuristic framework is accepted, a second issue is whether low-prejudice whites and low-identification blacks will rate (and remember) message arguments (e.g., comprehension of claims, product and brand awareness) more strongly (and more easily) than will high-prejudice whites and high-identification blacks. Specifically, the following hypothesis was tested:

[H.sub.2]: Main effects are predicted. Low-prejudice whites and low-identification blacks (versus high-prejudice whites and high-identification blacks) will show larger differences in ratings about the message itself (i.e., product and brand awareness, recall and recognition of message claims, comprehension, and strength of message claims).

Study 1

A total of 160 white and 140 black undergraduates rated a professionally prepared storyboard featuring a white or black actor promoting a portable word processor or a liquid laundry detergent. A 2 X 2 (Actor's Race: white or black X Product Category: liquid laundry detergent or portable word processor) between-subjects analysis of variance design was employed in which participants were randomly assigned to one of four experimental cells. In addition, whites' racial prejudice and blacks' identification with black culture were assessed and included in the analysis.

Stimulus Material. Full-color storyboards were professionally prepared to promote two fictional products (a portable word processor and a liquid laundry detergent) and three familiar products (a cordless telephone, a popcorn popper, and a frozen fruit beverage). All advertisements were photographically reduced, placed in 8 1/2-by-11-inch plastic covers, and inserted in a booklet. Two versions of the portable word processor and the liquid laundry detergent storyboards were created; one featured a white male, and the second, a black male. The white and black actors for each product received comparable mean attrativeness and intelligence ratings. The actors were also similar in dress and appearance, including facial expressions and gestures. In short, these advertisements were identical except for the race of the actor.

Procedure. Each participant was given a booklet containing the storyboards and the questionnaire. After viewing all of the storyboards, participants

[TABULAR DATA OMITTED]

answered questions about one of the irrelevant storyboards (e.g., the cordless telephone). These questions included, for instance, general impressions of the storyboards and the product, and likelihood of purchase. The next set of questions concerned the critical storyboards for the fictitious products, the portable word processor or the liquid detergent. The remaining questions concerned the filler storyboards, which featured a popcorn popper and a fruit punch beverage. Participants then completed the measure of attitudes toward blacks. Finally, a post-experimental inquiry assessed whether participants were aware of the true purpose of the experiment. Participants were then debriefed and thanked for their participation.

Measuring Instruments. Participants responded on 15-point Likert scales for all measures except thought-listing, product- and brand-awareness measures, recall and recognition of message claims, and attitudes toward blacks.

Thoughts. After each storyboard was shown, participants were given two and one-half minutes to list their thoughts and ideas. Thoughts produced for the critical storyboards were subsequently scored by two independent raters as product- or execution-oriented and as positively, negatively, or neutrally valenced. Interrater reliability coefficients (Pearson r) for these categories ranged from .85 to .98.

Source Perception. Participants rated their degree of similarity to the actor in terms of overall lifestyle, personality, cultural background, dress and appearance, and basic values (see McKirnan, Smith, and Hamayan 1983). Responses were averaged over the four scales (alpha = .72 in Study 1; .86 in Study 2). Participants also expressed their ability to identify with the actor in the storyboard by indicating the extent of their agreement with the following statements: (1) a person whom I want to be like; (2) my type of person; and (3) a person who speaks for a group of which I am a member (see Kelman and Eagly 1965). Responses were averaged over the three scales (alpha = .79 in Study 1; .82 in Study 2).

Message Evaluation. Participants were asked to list all of the product categories for which they saw storyboards, and to recall the brand name of the produce in that category. Further, as a recognition measure, participants were given descriptions of the three product categories and were asked to choose the correct brand name of the product from seven choices provided.

Participants also rated the validity of the message claims on four scales (strong-weak, persuasive-unpersuasive, important-unimportant, believable-unbelievable). Responses were averaged over the four scales (alpha = .87 in Study 1, .92 in Study 2). Message evaluation was also assessed by an eight-item, multiple-choice test on the details of the storyboard copy (i.e., "true," "false," or "not mentioned in the ad").

Outcome Measures. Participants indicated how likely it would be that they would purchase the laundry detergent (or portable word processor). Then, participants indicated one of four degrees of interest in obtaining additional information about the advertised product: (1) have no interest at all; (2) would like to receive information in the mail; (3) would talk with the salesperson over the phone; or (4) would meet with a representative in person about the product (Busch and Wilson 1976).

Participants gave their overall impression of the product on three semantic-differential scales (bad-good, satisfactory-unsatisfactory, low quality-high quality; see Petty, Cacioppo, and Schumann 1983). Responses were averaged over the three scales (alpha = .91 in Study 1; .88 in Study 2). Participants gave their their overall impression of the storyboard on three semantic differential scales (unattrative-attractive, fatorable-unfavorable, persuasive-unpersuasive). Responses were averaged over the three scales (alpha = .91 in Study 1; .84 in Study 2).

Racial Attitudes. After responding to the items in the questionnaire booklet, participants completed a "Survey of Opinions" which included items that measured whites' racial prejudice (for whites in both studies) and blacks' identification with the black culture (for blacks in the first study) embedded in a number of social and political items.

For the version administered to white participants, ten scale items from the Subtle Derogatory Belief Scale of the Multifactor Racial Inventory (Woodmansee and Cook 1967) were included. This scale has been shown to have acceptable reliability and validity as a measure of whites' prejudice toward blacks (see Woodmansee and Cook 1967). The composite prejudice index had high reliability (Cronbach's alpha coefficient = .84).

For the version administered to black participants, thirteen items assessing blacks' identification with black culture were substituted in the questionnaire (see Whittler, Calantone, and Young, forthcoming). The composite identification scale had acceptable reliability (Cronbach's alpha coefficient alpha = .72). A factor analysis (varimax rotation) yielded two rotated factors (cross-race attraction and support of black causes) which accounted for 55.9 percent and 19.8 percent of the variance, respectively.

Major Findings of Study 1

Source Perception. White participants perceived themselves as less similar to black than to white actors, particularly for the portable word processor (p < .001). For black participants, the predicted actor's-race-by-identification-with-black-culture intereaction was significant on perceived similarity (p < .05). Although both low-and high-identification blacks perceived themselves as more similar to black than to white actors, the effect was larger for high- (p < .0001) than for low-identification blacks (p < .001).

For white participants, the predicted actor's-race-by-racial-prejudice interaction was significant for the liquid laundry detergent (p < .001), but nonsignificant for the portable word processor on the ability-to-identify-with-the-actor dependent measure. Essentially, high-prejudice whites found it more difficult to identify with the black actor than with the white actor (p < .01), and found it more difficult to identify with the black actor than did low-prejudice whites (p < .01). For black participants, the predicted actor's-race-by-identification-with-black-culture interaction was significant on the ability to identify with the actor dependent measures (p < .01). High-identification blacks identified more strongly with black than with white actors, whereas low-identification blacks showed no differences in their ability to identify with a white or a black actor.

Table 2 summarizes the findings on these source-processing measures. The general form of the first hypothesis, that high-prejudice whites and high-identification blacks would react more strongly to source cues than low-prejudice whites and low-identification blacks, is supported.

Message Evaluation. Participants did not appear to engage in substantial processing of message arguments. With respect to the study's hypotheses, none of the measures used indicated effects of actor's race or racial orientation toward

[TABULAR DATA OMITTED]

blacks. Hence, the data did not support the second hypothesis that low-prejudiced whites and low-identification blacks would rate (and remember) message arguments more strongly (and more easily) than high-prejudice whites and high-identification blacks.

Outcome Measures. As shown in Table 3, black participants showed an increased likelihood of purchase behavior, perhaps suggesting loyalty to companies that had black actors promoting their products. Finally, high-prejudice whites showed less interest in obtaining additional information about the laundry detergent when a black actor (versus a white actor) was featured in the advertisement (p < .01).

Study 2

Sears (1986) questioned social science's heavy reliance upon college students in laboratory experiments designed to explore social issues. In response, a study conceptually similar to that just described (Whittler and Scattone, forthcoming) investigated the reactions of 160 southeastern white adults to a white or a black actor promoting a fur coat or a liquid laundry detergent. These two product categories were selected because they differed in terms of price and emotional involvement, and hence, information processing may differ (Zaichkowsky 1985). The same procedure and measurement instruments used in Study 1 were used.

Major Findings of Study 2

Source Perception. As shown in Table 2, the predicted actor's-race-by-racial-prejudice interaction was significant for perceived similarity (p < .05), and marginally significant on the ability-to-identify-with-the-actor dependent measures (p < .10). Planned comparisons revealed that low-prejudice whites showed no differences between their similarity or ability to identify with white or black actors, whereas high-prejudice whites perceived themselves as less similar to black than to white actors, and identified more strongly with white than with black actors. In short, the prediction that high-prejudice whites would react more strongly to racial cues than low-prejudice whites was supported.

Message Evaluation. The effect of racial attitude on the brand-awareness measures sheds some light on message evaluation. Low-prejudice whites exhibited greater brand-name recall than high-prejudice whites (p < .01). The significant actor's-race-by-racial-prejudice interaction for correct brand-name recognition (p < .05) revealed that prejudiced whites' brand-name recognition decreased when the actors were black but was unaffected when the actors were white. Finally, although low-prejudice whites showed greater recall of the products' brand names than did high-prejudice whites, the two groups did not differ in their thoughts, message comprehension, or ratings of the perceived strength of the message claims.

Outcome Measures. The actor's race was the only significant effect on outcome measures. Consistent with the early research (e.g., Barban 1969; Schlinger and Plummer 1972), participants who expressed greater interest in additional information were more likely to purchase the products, and evaluated both the product and the advertisement more favorably when the actors in the advertisements were white (see Table 3).

Summary and Implications

Black households, the largest demographic subgroup targeted by advertisers, watch 23 hours more television than other households each week (Nielsen Media Research 1988). It makes sound economic sense, then, for advertisers to consider including black actors in advertisements for mixed audiences. Wilkes and Valencia's (1989) content analysis of prime-time television commercials reveals that today's advertisers should be applauded for their willingness to include black actors in their messages. Unfortunately, their findings also include that black actors are typically assigned minor and background roles.

Experimental research offers insight into the effects of black actors in major roles or as featured characters. Given the consistency of the findings of previous research (despite methodological limitations), it appears that some white viewers are unaffected by the presence of black actors (as indicated by measures of message and brand name recall, attitudes toward the product, brand, or company, and purchase intentions), while others are not as positive to black actors as they are to white actors. In contrast, black viewers respond more positively to black actors.

Two recent investigations (Whittler 1989; Whittler and Scattone, forthcoming) explored the effects of actor's race in commerial advertising. The findings for the outcome measures of Whittler and Scattone indicated that regardless of their attitudes toward blacks, white adults were less likely to purchase the products, and had less favorable attitudes toward the products and advertisements when the advertisements featured a black rather than a white actor. These findings contradict those of Whittler (1989). Using the same outcome measures, he reported that white college students reacted similarly to advertisements containing a white or black actor, whereas black college students reacted more favorably to advertisements containing a white or black actor. In reconciling the differences between the two studies, Whittler and Scattone (forthcoming) suggest that today's younger whites and blacks

[TABULAR DATA OMITTED]

have more opportunity for interaction than adult whites and blacks. Thus, younger people may be more willing to accept individuals of different races.

These two studies also extend early research by offering an explanation for why the actor's race matters in commercial advertising. Findings from both studies support the notion that viewers with stronger racial attitudes are more likely to be affected by source characteristics (i.e., whether the actors are black or not) than viewers with weaker racial attitudes. However, these studies' findings do not fully support the notion that viewers with weaker racial attitudes are more likely to consider the message claims than viewers with stronger racial attitudes.

Tommy E. Whittler (Ph.D., Purdue University) is assistant professor of marketing, University of Kentucky.

References

Barban, Arnold M. (1969), "The Dilemma of Integrated Advertising," Journal of Business, 42 (October), 477-496.

Barban, Arnold M. and Edward W. Cundiff (1964), "Negro and White Response to Advertising Stimuli," Journal of Marketing Research, 1 (November), 53-56.

Block, Carl E. (1972), "White Backlash to Negro Ads: Fact or Fantasy?" Journalism Quarterly, 49 (Summer), 258-262.

Busch, Paul S. and David T. Wilson (1976), "An Experimental Analysis of Salesman's Expert and Referent Bases of Social Power in the Buyer and Seller Dyad," Journal of Marketing Research 13 (February), 3-11.

Bush, Ronald F., Robert F. Gwinner, and Paul J. Solomon (1974), "White Consumer Sales Response to Black Models," Journal of Marketibng, 38 (April), 25-29.

Bush, Ronald F., Joseph F. Hair, Jr., and Paul J. Solomon (1979), "Consumers' Level of Prejudice and Response to Black Models in Advertisements," Journal of Marketing Research, 16 (August), 341-345.

Cagley, James W. and Richard N. Cardozo (1970), "White Response to Integrated Advertising," Journal of Advertising Research, 10 (April), 35-39.

Chaiken, Shelley (1980), "Heuristic versus Systematic Information Processing and the Use of Source versus Message Cues in Persuasion," Journal of Personality and Social Psychology, 39 (November), 752-766.

Chaiken, Shelley (1987), "The Heuristic Model of Persuasion," in Social Influence: The Ontario Symposium, Vol. 5, M.P. Zanna, J.M. Olson, and C.P. Herman eds., Hillsdale, NJ: Erlbaum.

Chaiken, Shelley, Akiva Liberman, and Alice H. Eagly (1989), "Heuristic and Systematic Processing Within and Beyond the Persuasion Context," in Unintended Thought: Limits of Awareness, Attention, and Control, ,J.S. Uleman and J.A. Bargh eds., New York: Guilford.

Choudbury, Pravat K. and Lawrence S. Schmid (1974), "Black Models in Advertising to Blacks," Journal of Advertising Research, 14 (June), 19-22.

Guest, Lester (1970), "How Negro Models Affect Company Image. Journal of Advertising Research, 10 (April), 29-33.

Kelman, Herbert C. (1961), "Processes of Opinion Change," Public Opinion Quarterly, 25 (Spring), 57-78.

Kelman, Herbert C. and Alice H. Eagly (1965), "Attitude Toward the Communicator, Perception of Communication Content, and Attitude Change," Journal of Personality and Social Psychology, 1 (January), 63-78.

McKirnan, David J., Charles E. Smith, and Else V. Hamayan (1983), "A Sociolinguistic Approach to the Belief-similarity Model of Racial Attitudes," Journal of Experimental Social Psychology, 19 (September), 434-447.

Muse, William V. (1971), "Product-Related Response to Use of Black Models in Advertising," Journal of Marketing Research, 8 (February), 107-109.

Nielsen Media Research (1988), "Television Viewing Among Blacks," 4th Annual Report, as cited in USA Today, August 12, 1988, D1.

Petty, Richard E., John T. Cacioppo, and David Shumann (1983), "Central and Peripheral Routes to Advertising Effectiveness" The Moderating Role of Involvement," Journal of Consumer Research, 10 (September), 135-146.

Schlinger, Mary Jane and Joseph T. Plummer (1972), "Advertising in Black and White," Journal of Marketing Research, 9 (May), 149-153.

Sears, David O. (1986), "College Sophomores in the Laboratory: Influences of a Narrow Data Base on Social Psychology's View of Human Nature," Journal of Personality and Social Psychology, 51 (September), 515-530.

Solomon, Paul J., Ronald F. Bush, and Joseph F. Hair, Jr. (1976), "White and Black Consumer Sales Response to Black Models," Journal of Marketing Research, 13 (November), 431-434.

Stafford, James E., Al E. Birdwell, and Charles E. Van Tassel (1970), "Integrated Advertising--White Backlash?" Journal of Advertising Research, 10 (April), 15-20.

Szybillo, George J. and Jacob Jacoby (1974), "Effects of Different Levels of Integration on Advertising Preference and Intention to Purchase," Journal of Applied Psychology, 59 (June), 274-280.

Tolley, Stuart and John Goett (1971), "Reactions to Blacks in Newspaper Ads," Journal of Advertising Research, 11 (April), 11-17.

Whittler, Tommy E. (1989), "Viewers' Processing of Actor's Race and Message Claims in Advertising Stimuli," Psychology and Marketing. 6 (Winter), 287-309.

Whittler, Tommy E., Roger J. Calantone, and Mark J. Young (forthcoming), "Strength of Ethnic Affiliation: Examining Black Identification with Black Culture," Journal of Social Psychology.

Whittler, Tommy E. and joan Scattone (forthcoming), "Viewers' Reactions to Racial Cues in Advertising Stimuli," Journal of Advertising Research.

Wilkes, Robert E. and Humberto Valencia (1989), "Hispanics and Blacks in Television Commercials," Journal of Advertising, 18 (1), 19-25.

Woodmansee, John J. and Stuart W. Cook (1967), "Dimensions of Verbal Racial Attitudes: Their Identification and Measurement," Journal of Personality and Social Psychology, 7 (October), 240-250.

Zaichkowsky, Judith L. (1985), "Measuring the Involvement Construct," Journal of Consumer Research, 12 (December), 341-352.