There is some debate on whether Black and White audiences can be reached equally well with the same racially-targeted media. Research on Blacks' and Whites' response to race-specific messages suggests that whites respond no differently to media messages targeted to Blacks or Whites (Appiah 2001a, 2001b, 2002; Bush, Gwinner, & Solomon, 1974; Bush, Hair, & Solomon, 1979; Greenberg, 1986; Pitts, Whalen, O'Keefe, & Murray, 1989; Schlinger & Plummer, 1972; Soley, 1983; Whittler, 1991; Whittler & Dimeo, 1991), while Blacks respond to media messages more favorably when the messages are targeted to Blacks (Appiah 2001a; 2001b, 2002; Choundhury & Schmid, 1974; Greenberg & Atkin, 1982; Whittler, 1989, 1991). Despite the evidence, there still appears to be some disagreement on which racially-targeted media best appeal to audiences, particularly Black audiences. Some argue that White-targeted media is just as effective in reaching Blacks as Black-targeted media (e.g., Askey, 1995; Gadsden, 1985) while others hold that the best way to reach Blacks is through Black-oriented media (Appiah & Wagner, 2002; Fannin, 1989; Harris, 1981). Much of the work used to resolve this debate has focused on audiences' responses to traditional media rather than new media. This study explores whether Black and White audiences respond to race-specific messages from Internet Web sites as they do to messages from more traditional media.
Previous studies have been useful in understanding and highlighting how Black audiences respond to race-specific media messages and characters from television and print. Much of this research finds that Black consumers heavily rely on print and television for information and use that information when making purchases (Miller & Miller, 1992; Soley, 1983), but they often ignore television and advertising that is perceived to be targeted to primarily White audiences (Appiah, 2002; Rossman, 1994, "Where blacks," 1993). Instead Black audiences are more attracted to media with Black characters (Appiah, 2001b; Dates, 1980), rate Black characters more positively (Greenberg, 1986; Whittler, 1991 ; Appiah, 2002), and show an increased likelihood of purchasing products promoted by Black characters (Whittler, 1989). Studies also indicate that Blacks are more likely to identify with Black television characters (Greenberg & Atkin, 1982), recall more content from Black characters (Appiah, 2002; Whittler, 1991), and more likely to trust ads and editorial content in Black-targeted media ("Study Reveals Blacks," 1998).
Few, if any, empirical studies, however, have explored whether these same findings are consistent when examining media such as the Internet. As the Internet becomes more crowded and as Blacks increase use of and access to the Internet, content providers will need to know how best to reach both Black and White audiences. The purpose of this study is to determine whether Black and White Web users display differences in their surfing and evaluation of Web sites based on the target race of the site. Conceptually, it is expected that race-targeted Web sites will not affect White viewers' responses to a Web site. In contrast, it is expected that race-targeted Web sites will influence Black viewers' responses to a site.
Internet Access and Usage
Current Internet research on Blacks and Whites has focused a great deal on differences in Blacks' and Whites' access to and use of the Internet (see digital divide, Hindman, 2000; Beaupre & Brand-Williams, 1997; Henry, 1999; Joyce 1997; McConnaughey & Lader, 1998; Hoffman & Novak, 1998; Nie & Erbring, 2001; Schement, 2001; Walsh, Gazala, & Ham, 2001). For example, although Blacks with incomes below $40,000 were far less likely than Whites to own a computer and go online (Raney, 1998; Hoffman & Novak, 1998), there is evidence that Blacks with higher incomes use computers and go online at the same or greater rate than their White counterparts. Studies indicate that Blacks with incomes above $40,000 are as likely or more likely than Whites to own a computer, have computer access at work, and use the Internet during office hours (Hoffman & Novak, 1998; Kaiser Family Foundation, 2001). Other empirical studies indicate that there is virtually no difference in Internet use among Blacks and Whites with middle to upper level incomes (Hoffman & Novak, 1998; Hubbard, 2000), and that Blacks are among the fastest growing segments on the World Wide Web (DePriest, 2000; Sutel, 1999; Walsh et. al., 2001).
The focus of this study, however, is not whether Blacks have access to or use of the Internet but rather what do Blacks and Whites access when on the Internet, and how does the content affect their Web browsing and evaluation. This study provides a better understanding of the theoretical underpinnings and psychological mechanisms at work when online users are exposed to race-targeted Web sites on the Internet.
Identification theory (Kelman, 1961) maintains that during an interaction people automatically assess their level of similarity with a source and make similarity judgments (Hovland & Weis, 1951; Kelman, 1961). This process drives individuals to choose models to identify with based on perceived similarities between themselves and the model (Basow & Howe, 1980; Kelman, 1961). When individuals perceive that a source possesses a specific characteristic similar to their own, they begin to infer that the source will also share other characteristics, all of which lead to greater identification (Feick & Higie, 1992). Studies have shown that individuals who identify with media characters are more affected by the media content in which those characters are engaged (Huesman, Eron, Klein, Brice, & Fischer, 1983). For example, Black viewers who identify with Black characters have been found to more likely believe they are the target audience of media with Black characters, and evaluate more favorably media with Black characters than they do media with White characters (Aaker, Brumbaugh, & Grier, 2000; Appiah, 2001a, 2001b; Whittler, 1989, 1991; Whittler & Spira, 2002).
Research has shown that Blacks and Whites differ in the attributes used to determine similarity. Blacks often use race as their primary criterion to evaluate people (Smedley & Bayton, 1978). Blacks do not seem to consider both races equally relevant when learning about and evaluating content in race-targeted media (Appiah, 2001b, 2002). Instead, evidence suggests Blacks regard White characters and content as less relevant and less worthy of attention. In a study examining White and Black viewers' recall of Black and White television characters, Appiah (2002) found that Black audiences recalled substantially less information from White television characters than they did from equivalent Black television characters. Appiah maintained that Black viewers may have displayed a conscious and active form of "dis-identification" with White characters. That is, because Black audiences may not identify with White models, Black viewers may have believed the content from White media and characters was personally irrelevant and not directed at them, causing them to tune out. In viewing targeted Internet Web sites, this dis-identification process could similarly lead Black viewers to find White-targeted Web sites less personally relevant, causing Blacks to respond more favorably to Black-targeted Web sites.
Among Whites, racial similarity plays little if any role in how Whites respond to targeted media. Whites do not generally think of themselves as distinctly part of a specific ethnic group (Phinney, 1992; Royce, 1982), and as such have been found to consistently place significantly lower importance on their racial and ethnic identity (Phinney, 1992). Although Whites may not use race as a cue to assess similarity with a source, identification theory postulates that identification can still occur when individuals infer that their tastes and preferences are similar to those of the source (Eagly, Wood, & Chaiken, 1978). White audiences have been found to use occupational status or social class cues rather than racial cues to determine perceived similarity between themselves and a source. In fact, researchers (e.g., Coleman, Jussim, & Kelley, 1995) argue that characteristics such as personal appearance, dialect style, and socioeconomic status have a greater impact on Whites' evaluations of a source than does the race of a source. In support, studies on race-class stereotypes reveal that the dominant criterion used by White subjects to evaluate people is occupational (Feldman, 1972) or social class status (Smedley & Bayton, 1978), and not race. For example, Smedley and Bayton discovered that White subjects rate middle-class Blacks and Whites equally favorable, and rate lower-class Whites and Blacks equally less favorable. Similarly, in a study on perceptions of Black and White television characters Appiah (2002) found that White viewers evaluated comparable Black and White television characters equally favorable.
Not only is it possible for Whites to identify with Blacks who are perceived to share their tastes and preferences, evidence shows that Whites can be more attracted to Blacks with beliefs, values, and personalities similar to their own than they are to Whites who are different with respect to these characteristics (Appiah, 2001a). For instance, in a study examining White and Black adolescents' responses to Black and White character ads, Appiah found that White adolescents were more likely to identify with Black characters in ads than they were with White characters in ads. This led White subjects to not only believe they were the target of Black character ads but to also rate Black character ads more favorably.
These findings reveal that when White and Black media models possess comparable traits, White viewers identify with them at least equally and will be similarly affected by the media content in which these comparable White and Black characters appear. Further, these findings imply that when White viewers are exposed to equivalent Black and White characters and sites, Whites should show no difference in their browsing time on or evaluation of a site based on the racial target of the site.
Another theory used to explain the psychological mechanisms at work when online users are exposed to race-targeted Web sites is distinctiveness theory. This theory suggests that a person's own distinctive traits (e.g., African American, red-headed, left-handed) will be more salient to him or her than more prevalent traits (e.g., Caucasian, brunette, right-handed) possessed by other people in his or her environment (McGuire, 1984; McGuire, McGuire, Child, & Fujioka, 1978). The theory predicts that ethnicity will be more salient for people whose ethnic group is part of a numeric minority in a social environment than it will be for members of an ethnic majority in a particular social environment. Black people, for instance, are a numeric minority in the United States and in the media. As a result, they should be highly aware and mindful of their race in personal and mediated situations.
Given their absence in social and mediated environments, Blacks are particularly mindful of their ethnicity. Conversely, Whites' ethnicity is not readily available in their self-concept and is unlikely to grow in importance until they are no longer in the majority in specific settings (Phinney, 1992). As a given ethnic group becomes numerically more dominant in a social environment, ethnicity becomes progressively less salient in the self-concept of its members (McGuire et. al., 1978). This is evident by research that shows majority Whites are significantly less likely than minority Blacks and Hispanics to mention their ethnicity when asked to define themselves (McGuire et al., 1978; Phinney, 1992). In fact, altering a person's social environment such that different physical characteristics become distinctive will lead to a change in a person's self-concept (McGuire et al., 1978). As McGuire and colleagues (1978) indicated, a Black woman in a large group of White women will be acutely aware of her race. When the same Black woman moves to a large group of Black men, her Blackness loses salience and she becomes more conscious of being a woman. Similarly, a White man entering a room full of Black men suddenly becomes mindful of his Whiteness. However, when the same White man enters a room full of White women, he is more conscious of his gender and not his ethnicity.
How might being a member of a distinctive group or nondistinctive group influence Black and White viewers' responses to targeted media? Studies using distinctiveness theory demonstrate that targeted media would be most effective in contexts in which the target market is a numeric minority (Aaker et al., 2000; Appiah, 2001a, 2001b, 2002; Grier & Deshpande, 2001). Blacks represent a numeric minority group and, compared to Whites, have few media messages targeted specifically at them. Grier and Brumbaugh (1999) argue that, unlike Whites, Blacks as ethnic minorities appreciate the acknowledgement associated with being a target market and are more likely to use targeting cues based on their racially distinctive trait in attending and evaluating media than White majority members are to use targeting cues based on their nondistinctive trait. They also mention that, due to their increased awareness of the trait that makes them unique, Blacks are more likely to connect with targeted media and make links between the targeted media and themselves. Studies have shown that this leads Blacks to develop more favorable attitudes towards Black-targeted media and its content than towards White-targeted media and content (Aaker et al., 2000; Appiah, 2001 a, 2001b, 2002).
Accordingly, Blacks' favorable attitudes toward Black-targeted media should be evident as they surf the Internet, given less than one half of one percent of Web sites are targeted to Blacks (Hoffman, Novak, & Schlosser, 2001). Media such as Internet Web sites that target a Black distinctive market segment on the basis of race should have a greater impact on Black viewers' attention to and evaluation of the Web sites than Web sites targeting a White nondistinctive market segment. Conversely, White audiences targeted based on their "whiteness," a numerically common trait in the U.S., should use that characteristic less in assigning importance to media than Black viewers targeted because of their "blackness" (Grier& Brumbaugh, 1999).
Consistent with distinctiveness theory, Whites as a majority group are relatively unaffected by White-targeted media. White racial majority members seem to be less mindful of a character's race and focus on similarities between themselves and the source that are less race-specific (e.g., values, dress, lifestyle, appearance) as evidenced by studies which show that White audiences respond just as favorably to ads and targeted media with Black models as they do to ads and targeted media with White models (Appiah, 2001a, 2001b, 2002; Bush et al., 1979; Schlinger & Plummer, 1972; Whittler, 1991). Grier and Brumbaugh (1999) summarize such findings by concluding, "Apparently whiteness is neither a salient nor a meaningful characteristic for those non-distinctiveness viewers, and targeting on the basis of membership in the dominant culture does not factor into the meaning they create. Unless an advertisement speaks to them on some other dimension, such as gender or psychological profile, targeting on the basis of membership in the dominant culture is likely to be ineffective" (p. 90).
This discussion leads to the overall expectation that Black Web users' navigation time, attitudes, and recall will be more strongly affected by the racial target of the site than will White Web users. More specifically:
H1 : Black Web users--as members of a minority and distinctive group--will spend more time overall navigating on a Black-targeted Web site than they will on an equivalent White-targeted Web site. H2: Black Web users--as members of a minority and distinctive group-will spend more time viewing each story on a Black-targeted Web site than they will on an equivalent White-targeted Web site. H3: Blacks--as members of a minority and distinctive group--will rate a Black targeted Web site more positively than they will an equivalent White-targeted Web site. H4: Blacks--as members of a minority and distinctive group--will rate stories on a Black-targeted Web site more favorably than they will the same stories on an equivalent White-targeted Web site. H5: Blacks--as members of a minority and distinctive group--will recall more banner ads from a Black-targeted Web site than they will from an equivalent White-targeted Web site. H6: Whites--as members of a majority and nondistinctive group--will show no difference in their navigation time, attitudes, or recall based on the racial target of the Web site.
Subjects and Design
Two hundred three undergraduate and graduate students (ages 18 to 34, median age 20) from a large, public Midwestern university participated in this study: 106 Whites and 97 Blacks. Sixty percent were female and 40% were male. These subjects were recruited from courses in the School of Journalism and Communication and voluntarily participated in the study for either extra course credit or to have their name included in a drawing to win $200. Due to the difficulty of selecting only Black and White students, all ethnic groups were included during data collection. However, only Black and White subjects were included in the study analyses.
The experiment employed a 2 (subjects' race: Black or White) x 2 (Web site target: Black or White) between-subjects design to test the hypotheses for both Black and White Web site users. The five dependent variables were: 1) total time (in minutes) spent navigating on a Web site; 2) total time (in minutes) spent viewing each story; 3) overall attitude toward the Web site; 4) overall attitude toward the stories; 5) and recall of banner ads.
A professional Web designer created two identical Web sites, one site targeted to Black consumers and the second designed for White audiences. Upon entering a Web site users were greeted with the name of the Web site, which was animated to scroll across the page. "Community in Motion: The Essence of America online 1.1," was the name of the general market site while "Community in Motion: The Essence of Black America online 1.1," was the name of the Black-targeted site. This simple addition of the word "Black" in the name of the Black-targeted site was done to provide users with a racial cue that would reinforce the racial orientation of the Web site.
Each Web site allowed the user the opportunity to enter into one of five subject areas from the orange colored home page. The subject areas included the following: 1) entertainment; 2) health; 3) news; 4) sports; and 5) travel. Each of the 5 subject areas had 4 stories through which users could navigate--a total of 20 stories on each site. For example, when users clicked on the subject heading "sports," they were linked to a page that contained a sports story, the titles, and links to each of the four sports stories, and the links to each of the five subject areas. The stories were all real stories about real events. Each of the 20 stories was approximately the same length.
The stories were chosen by a group of 30 college students who reviewed various online magazines and selected more than 100 stories using the guidelines that the stories be interesting, not time-sensitive, not widely known, and not from the local news. The four stories in each subject category that were deemed most interesting, least time sensitive, and least familiar to the campus student population were selected for the study.
Both the Black-targeted and the White-targeted Web sites had identical stories, layout, and banner ads. Other than the title, the only difference between the sites was that Black pictures accompanied the stories on the Black-targeted site and White pictures were used on the White-targeted site. Effort was made to use equivalent Black and White pictures alongside each story. For example, for the sports story discussing female athletes' anterior cruciate ligament (ACL) injuries, a picture of a Black female doctor was used for the Black-targeted site and a picture of a White female doctor was used for the general market site. To better ensure that Black and White pictures were equivalent, many of the pictures were digitally manipulated to vary only the race of the characters while holding constant all other visual features of the pictures. Using this technique, any differences in Web users' responses to Black-or White-targeted sites must be solely attributed to the race of the characters in the site.
Each subject participated individually in a small lab located in the School of Journalism and Communication. Each subject was seated at a desktop computer and randomly assigned to navigate through one of the two race-targeted Web sites.
Prior to navigating on a Web site, participants were told that the purpose of the Internet study was to learn more about the types of Web sites they like best, which would enable researchers to improve the look, style, and content of Web sites. They were told that the Web site was a prototype that may soon appear on the World Wide Web and were asked to spend up to 45 minutes navigating through and examining the site. They were told that they would complete a questionnaire pertaining to the Web site once they finished navigating through the site, and they were asked to provide their honest evaluation of the site when completing the questionnaire. Participants were guaranteed anonymity.
As part of the instructions, students on the Black-targeted site were told that the Web site they would view was targeted primarily to "African Americans." Students on the White-targeted site were told specifically that the site was targeted primarily to a "general White mainstream audience." This was done to provide users with a racial cue that would reinforce the racial orientation of the Web site. Five percent of the subjects indicated at least some knowledge of the study purpose and were excluded from the overall analysis.
Data Collection and Measurement Instruments
Two techniques were used to collect data for this study. First, online computer tracking software was used to track and time Black and White subjects' movements throughout the Web sites. This is an unobtrusive means of collecting data and provides a number of benefits that have been mentioned in previous research (see Ettema, 1985; Eveland & Dunwoody, 1998; Rice & Borgman, 1983). In particular, given the problem of social desirability that may sometimes influence the validity of self-report measures in studies on race and ethnicity (see Gaertner & Dovidio, 1986; McConahay, 1986), online tracking software was used to collect data unobtrusively to eliminate the potential problem of socially desirable responses. As the subject moves throughout the site, the Web server makes a record of the time, the page, and a host of other information from the user. The information from each user is saved on the Web server and can later be analyzed. For this study, this software was used to measure two specific dependent variables: 1) subjects' overall navigation time on a Web site; and 2) their overall time viewing each story on a Web site. For example, subjects spent an average of 24.16 minutes navigating on a site, 25.87 minutes on the Black-targeted site, and 22.27 minutes on the White-targeted site. Subjects spent an average of just under a minute (M = 0.92) viewing each story on a site, 0.98 minutes viewing each story on the Black-targeted site, and 0.85 minutes viewing each story on the White-targeted site.
The second method used to collect data was a questionnaire that was given to participants after they finished navigating on the Web site. This questionnaire gathered information on two other dependent variables--users' overall attitude toward the Web site, and their overall attitude toward the stories. An attitude toward the Web site index was created by averaging the mean scores of eleven 7-point semantic differential scales: boring/interesting, bad/good, negative/positive, useless/ useful, worthless/valuable, poor/outstanding, not for me/for me, weak/strong, not appealing/appealing, not attractive/attractive, and not likable/likable. These scales have been used successfully in other studies and have shown strong evidence of reliability (e.g., Appiah, 2001a; Bush et al., 1979; Desphande & Stayman, 1994; Green, 1999). The coefficient [alpha] for this index was .95.
An index that measured subjects' attitude toward the Web stories was developed by averaging the mean scores of nine 7-point semantic differential scales: inaccurate/ accurate, not credible/credible, biased/unbiased, unclear/clear, unfair/fair, not informative/informative, unimportant/important, not persuasive/persuasive, and poorly written/well-written. These scales were modeled after those used by Sundar (2000). The coefficient [alpha] for this index was .86.
The final dependent variable, recall of banner ads, was used to measure an aspect of users' attention to the Web site. Users who are attentive as they browse a Web site are likely to notice peripheral items on a site such as banners ads. Recall of banner ads may be a particularly important measure of attention, given that Web browsers generally are unaware, ignore, or simply cannot recall banner ads (Davis, 1999; Thorson, Wells, & Rogers, 1999). The unaided recall procedure for this study resembles those in other studies on ad recall (see Beattie & Mitchell, 1985, for a review). Subjects were specifically asked to, "Please list each banner ad you remember seeing on the Web site." In order to provide a correct response, subjects could not just describe the banner ad, they had to accurately name the brand. One banner ad was used for each of the five subject areas for a total of five. For example, a banner ad for Poland Spring Water was used for each story on sports whereas a banner ad for Hollywood Video was used for each story on the entertainment subject area. Each of the five banner ads was located in the same place on the page and was approximately the same size.
The questionnaire also measured race/ethnicity of each Web user. Subjects were given a list of racial and ethnic groups from which to choose. Subjects who did not designate themselves as being either Black or White or identified with more than one racial or ethnic group were not included in the analysis.
The results of the experiment are presented and discussed according to the hypotheses presented earlier. A series of two-way analyses of variance for all hypotheses is provided below with follow-up analyses conducted to examine significant findings. The same analyses were conducted for all five dependent variables.
Overall Time Spent on Web Sites
It was predicted that Blacks would spend more time overall viewing a Black-targeted Web site than they would viewing an equivalent White-targeted site, while White viewers' navigation time would not be affected by the racial target of the sites. A marginally significant interaction between subjects' race and Web site target race (F (3, 171) = 2.91, p < .10) was found. Further examination of the means (see Table 1) showed that Blacks spent significantly (F (1, 80) = 5.41, p < .05) more time viewing a Black-targeted site (M = 28.01) than they did a White-targeted site (M = 21.81).
Overall Time Spent on Each Story
This hypothesis predicted that Blacks would spend more time viewing each story on a Black-targeted site than they would a White-targeted site, while Whites' time on each story would not be affected by the racial target of the sites. A significant interaction between subjects' race and Web site target race (F (3, 1 71) = 4.34, p < .05) was found. Follow-up analyses indicated that Blacks spend significantly (F (1, 80) = 5.71, p < .01) more time viewing a story on a Black-targeted site (M = 1.14) than they do a White-targeted site (M = .84). These results support the hypothesis (see Table 1).
Attitude Toward Web Sites
It was expected that Blacks would rate a Black-targeted Web site more positively than they would a White-targeted Web site, while Whites' rating of Web sites would not be affected by the racial target of the sites. No significant interaction or main effects were found.
Attitude Toward Stories
It was hypothesized that Black subjects would rate stories on the Black-targeted Web site more positively than they would stories on the White-targeted Web site, while Whites' rating of stories would not be affected by the racial target of the sites. No significant interaction or main effects were found.
Recall of Banner Ads
It was hypothesized that Blacks would recall more banner ads from the Black-targeted site than they would from the White-targeted site, while Whites recall would not be affected by the racial target of the sites. A significant interaction between subjects' race and Web site target race (F (3, 201) = 7.07, p < .01) was found. An examination of the means indicated that Black users recalled (F (1, 97) = 4.18, p < .05) more banner ads from the Black-targeted site (M = 1.64) than from the White-targeted site (M = 1.04). In contrast, White viewers' recall was not affected by the racial target of the site.
Previous studies have been useful in understanding and highlighting how Black and White audiences respond to race-specific media messages and characters from television and print (e.g., Appiah 2001a; 2001b, 2002; Pitts et al., 1989; Whittler, 1991). However, few, if any, empirical studies have examined whether Black and White subjects respond the same to race-specific messages on the Internet as they do to messages from traditional media. This study helps determine whether Black and White Web users display differences in their surfing and evaluation of Web sites based on the target race of the site.
In addition to adding to the literature on the effects of racially targeted new media, this study helps to add to the understanding of whether Blacks can be just as effectively reached with general, White-targeted media as they can with Black-targeted media. This is particularly important given the debate on racially-targeted media placement where some argue that White-targeted media are just as effective in reaching Blacks as Black-targeted media (e.g., Askey, 1995; Gadsden, 1985), while others hold that the best way to reach Blacks is through Black-oriented media (Appiah & Wagner, 2002; Fannin, 1989; Harris, 1981). This study also contributes to the field of mass communication by demonstrating the usefulness of online tracking software to collect data unobtrusively. This software was used to track and time subjects' movements as they navigated through either a Black-or White-targeted Web site.
It was expected that race-targeted Web sites would not affect White viewers' browsing and evaluation of a Web site. In contrast, it was expected that race-targeted Web sites would influence Black viewers' responses to a site. In particular, it was hypothesized that Blacks would respond more favorably to a Black-targeted Web site than they would to a White-targeted site. The findings from this study reveal that Blacks were more attentive to the Black-targeted site while Whites were equally attentive to the White-and Black-targeted sites. This discrepancy can in part be explained and understood by identification and distinctiveness theories.
Identification theory postulates that people automatically assess their level of similarity with a source during an interaction and make similarity judgments (Hovland & Weis, 1951; Kelman, 1961). Given that Blacks use race as their primary criterion to assess similarity and evaluate people (Smedley & Bayton, 1978), Blacks should identify more with Black media characters than White media characters. This is evidenced by research that shows that Blacks who identify with Black characters in media are more likely to believe they are the target audience of media with Black characters (e.g., Aaker et al., 2000; Appiah, 2001b) and evaluate more favorably media with Black characters than they do media with White characters (Aaker et al., 2000; Appiah, 2001b; Whittler, 1989, 1991; Whittler & Spira, 2002). This greater identification with Black characters should similarly lead Black Internet users to be more affected by Web content with Black characters than Web content with White characters. Although identification was not specifically measured in this study, as expected, Black subjects' time browsing a site and their average time viewing each story were significantly affected by the Black-targeted Web site with Black pictures. Blacks spent 28 minutes browsing the Black-targeted site compared to less than 22 minutes browsing the White-targeted site. Moreover, Black users spent 26% less time (i.e., 50 seconds v. 68 seconds) viewing each story on the White-targeted site than on the Black-targeted site.
Blacks also recalled more banner ads from the Black-targeted site than they did from the White-targeted site. This is consistent with other research that shows Blacks recall more information from Black-targeted media and content than from Whitetargeted media and content (Appiah, 2002; Whittler, 1991). Given research that consistently points to users ignoring or being blind to banner ads (e.g., Benway, 1999; Davis, 1999), it is reasonable to wonder whether exhibiting greater attention to and recall of banner ads by Blacks could possibly lead to less attention to and comprehension of news stories. The amount of time viewing each story might be a good predictor of Black users' overall memory of site content. In this study, Blacks spent more time viewing stories on the Black-targeted site, so it could follow that they also exhibit deeper comprehension of news story content from the Blacktargeted site vis-a-vis the White-targeted site. Future research should measure the relationship among time on each story, recall of banner ads, and comprehension of news story content to more adequately test overall memory of Web site content.
As with Black Internet users, the identification process should also explain, in part, how Whites' surfing and evaluation of sites may be affected by the racial targeting of the media. Unlike Blacks who use race as their primary cue when making similarity and evaluative judgments (Smedley & Bayton, 1978), Whites do not use race as a cue to determine similarity and are more guided by nonracial cues such as occupational (Feldman, 1972) and social class status (Smedley & Bayton, 1978). This might be expected given research showing that Whites do not tend to think of themselves as distinctly part of a specific ethnic group (Phinney, 1992; Royce, 1982) and place significantly low importance on their racial identity (Phinney, 1992).
Consistent with this literature and the hypotheses, White Web site users displayed no difference in their overall navigation time on a site and displayed no difference in their time spent viewing each story based on the racial target of the Web site. White subjects spent an average of 24 minutes on the Black-targeted site and nearly 23 minutes browsing the White-targeted site. Similarly, White subjects spent just under a minute viewing each story regardless of the racial target of the site. This same pattern was found for White subjects' evaluations of the Web sites and stories, and recall of banner ads.
Distinctiveness theory may provide an explanation as to why White viewers displayed no preference for the White-targeted site. The theory posits that an individual's distinctive traits will be more salient to him or her than more prevalent traits possessed by other people in the environment (McGuire, 1984). The theory also maintains that race or ethnicity will be more prominent in the self-concept of people whose racial or ethnic group is in the minority in their social environment than in people of the majority group (McGurie et al., 1978). For instance, Blacks as ethnic minorities appreciate the acknowledgement associated with being a target market and are more likely to use targeting cues based on their racially distinctive trait in attending and evaluating media than are White majority members based on their nondistinctive trait (Grier & Brumbaugh, 1999). Because of their increased awareness of the trait that makes them unique, Blacks are more likely to connect with targeted media and make links between the targeted media and themselves (Grief & Brumbaugh, 1999). This leads Blacks to develop more favorable attitudes towards the Black-targeted media and their content than towards White-targeted media and content (Aaker et al., 2000; Appiah 2001a, 2001 b, 2002).
Since Blacks represent a numeric minority group with few targeted Web sites available (Hoffman et al., 2001), Web sites that target a Black distinctive group on the basis of race would have a greater impact on Black viewers' navigation time and recall of information than a White-targeted site. In contrast, White Web users who are targeted based on their Whiteness, a numerically common trait in the United States, would use that characteristic less in assigning importance to media and would be relatively unaffected by White-targeted Web sites vis-a-vis Black-targeted Web sites. This was evident in the findings from this study. Since Whiteness is neither salient nor a meaningful trait for Whites, targeting them based on their membership in the dominant culture is likely to be ineffective unless the media "speak" to Whites on some other dimension like gender or psychological profile (Grier & Brumbaugh, 1999).
It should be noted that, unexpectedly, Black subjects' evaluations of the Web sites and the stories were not affected by the racial target of the site. The addition of a Black face alongside a race-neutral story was enough to hold the attention of Black users longer, as evidenced by their increased recall and navigation time on the Black site vis-a-vis the White site. However, the addition of a Black face may not have been enough to get Blacks to evaluate the site and the stories more favorably. One explanation as to why Blacks did not evaluate the Black site and stories more favorably may be in the stories' subject matter. Although Blacks may have found the stories relevant, particularly when Black pictures were alongside the stories, Blacks may not have found the issues uniquely specific to the Black community. For example, stories with a direct association to the Black community (e.g., sickle cell anemia, police racial profiling, affirmative action, gospel music, Black slavery reparations) may have yielded more favorable responses by the Black subjects.
Given none of the stories on the Black-targeted site made specific mention of Blacks or Black culture, a plausible argument could be made that few Black cultural cues on the Black-targeted site led Blacks to spend more time on the site in search of information that was particularly relevant to their race. However, this seems unlikely for two reasons. First, recent empirical work (e.g., Appiah, 2001a, 2001b) has shown that Blacks need few Black cultural cues in media targeted to them in order to summon their racial identities and feel targeted. Evidence shows that Black-targeted media with no Black cultural cues other than the race of the characters were just as effective as Black-targeted media filled with various rich Black cultural cues in getting Black audiences to perceive similarity, identification, and believe that they were the intended audience of the media (see Appiah, 2001a). This suggests that even with additional Black cultural cues on Black-targeted Web sites, even those embedded within the story content, Black navigators may respond no differently.
Second, Blacks could reasonably be accustomed to news stories in Black-targeted media that make no direct mention of race. Black publications such as Ebony and Black Enterprise routinely use Black pictures to accompany articles that are relevant to the lives of their Black audiences although do not mention race in the text. For example, in the September 2002 issue of Black Enterprise, there were ten articles (e.g., "Waiting for a Bears Market," "Social Engineering: Employees Loose Lips Can Sink Your Company Network," "Family Finances: Taking the Next Big Step," "Surviving 9-11") that were accompanied by pictures of Black people although the stories made no mention of Blacks or Black culture. Because the magazines are targeting Blacks and readers are aware of this fact, it may be unnecessary and even redundant to mention Blacks in some news articles. Black readers may then assume that these stories are relevant to them--with or without direct mention of race--because they are in media targeting them. For instance, a news story on BET (Black Entertainment Television) regarding President Bush's stance on disarming Iraq or a story on tips to staying physically fit may make no specific reference to Blacks but because it is on a Black television network and program Black audiences are likely to tune in longer and see this issue as more relevant to their lives, than if these same stories where in general market media like CNN or Headline News Network. To test the impact of culturally relevant content on Blacks' evaluation of Web sites and stories, future research in this area should not only manipulate the race of characters, but also vary the content of the stories to ensure Black-targeted stories explicitly mention and relate to Black culture.
Although the study appears to provide support for the effects of racial-targeting cues on Black users, the content of the instructions may raise questions about the impact of the method on the results. Before browsing the site, subjects were told that the site was targeted primarily to African Americans or told that the site was targeted primarily to White audiences. The use of these instructions as a racial cueing technique may open the results to an alternative explanation. That is, while the differences between Black and White users were statistically significant, they may have resulted in part from the instructions and not the experimental manipulation. Even if true, the results are still informative. The findings provide information on the potential strength of such instructional cues in influencing Black users' navigational behavior. While Blacks might typically skim over information on a Web site, in this case, the instructions, the experimental manipulation, or some combination of the two may have led Blacks to more carefully scrutinize the Black-targeted site, thereby spending more time on the site and processing more information from the site. The implications are that such scrutiny may lead Blacks to comprehend more content from news stories and recall more information about advertised products. In order to adequately clear up any ambiguity associated with racial cueing in the instructions, future research should include an additional condition that does not provide racial cueing in the instructions.
The findings of this study have practical implications for online businesses and news providers struggling to attract larger audiences, particularly ethnic audiences, to their Web sites. First, the results imply that online content providers may more efficiently reach a larger audience by including Black images with their stories, thereby attracting a fast-growing Black online market segment. Without such racial targeting, commercial Web providers run the risk of minimizing the time and attention Blacks exert in navigating through a site. This could have negative consequences on online purchases and repeat traffic. Second, research of this nature seems highly useful in revealing the needlessness of using White-targeted Web sites to reach Whites when Black-targeted Web sites can succeed in simultaneously gaining the attention of both distinct market segments.
Blacks are going online in impressive numbers, and efforts to provide content targeted to Blacks will likely be met with great success (Hoffman et al., 2001). Online businesses and news providers who are intimately aware of differences in Internet use and evaluation among different ethnic groups will be able to better serve the needs and secure the loyalty of these groups (True Colors, 2001).
Table 1 Means for Users' Navigation Time and Attitude Toward Site and Stories White Web Users White- Black- Targeted Targeted Web Site Web Site Average Time on Site 22.70 24.03 (in minutes) Average Time on .86 .86 Story (in minutes) Attitude Toward Site 4.37 4.53 Attitude Toward Story 5.08 5.25 Recall of Banner Ads 2.77 2.21 White Web Users White- Black- Targeted Targeted Web Site Web Site Average Time on Site 21.81 28.01 (in minutes) Average Time on .84 1.14 ** Story (in minutes) Attitude Toward Site 4.25 4.59 Attitude Toward Story 5.15 5.32 Recall of Banner Ads 1.04 1.64 ** Note: * p<.05; ** p<.01 Within the Black Viewers category, asterisk refer to mean pairs that are significantly different based on one-way analysis of variance (N = 203).
Aaker, J., Brumbaugh, A., & Grier, S. (2000). Non-target market effects and viewer distinctiveness: The impact of target marketing on attitudes. Journal of Consumer Psychology, 9(3), 127-140.
Appiah, O. (2001a). Black, white, Hispanic and Asian-American adolescents' responses to culturally embedded ads. Howard Journal of Communications, 12(1), 29-48.
Appiah, O. (2001b). The effects of ethnic identification on black and white adolescents' evaluation of ads. Journal of Advertising Research, 41(5), 1-16.
Appiah, O. (2002) Black and white viewers' perception and recall of occupational characters on television. Journal of Communication, 52(4), 776-793.
Appiah, O., & Wagner, M. (2002). Differences in media buying by online business in blackand white-targeted magazines: The potential impact of the digital divide on ad placement. Howard Journal of Communications, 13(4), 251-266.
Askey, L. (1995, November 13). Is the medium the message? Adweek, 36, p. 57.
Basow, S. A., & Howe, K. G. (1980). Role-model influence: Effects of sex and sex-role attitude in college students. Psychology of Women Quarterly, 4, 558-572.
Beattie, A. E., & Mitchell, A. A. (1985). The relationship between advertising recall and persuasion: An experimental investigation. In L F. Alwitt, & A. A. Mitchell (Eds.), Psychological Processes and Advertising Effects (pp. 129-155). Mahwah, NJ: Erlbaum.
Beaupre, B., & Brand-Williams, O. (1997, February 9). Gap for middle-class, poor blacks widens: Few resources fuel disparity. The Detroit News, pp. B1-B2.
Benway, J.P. (1999). Banner blindness: What searching users notice and do not notice on the World Wide Web. Unpublished doctoral dissertation, Rice University.
Bush, R. F., Hair, J. F., & Solomon, P. J. (1979). Consumer's level of prejudice and response to black models in advertisements. Journal of Marketing Research, 16, 341-345.
Bush, R. F., Gwinner, R. F., & Solomon, P. J. (1974). White consumer sales response to black models. Journal of Marketing, 38, 25-29.
Choudhury, P. K., & Schmid, L. S. (1974). Black models in advertising to blacks. Journal of Advertising Research, 14(3), 19-22.
Coleman, L. M., Jussim, L., & Kelley, S. H. (1995). A study of stereotyping: Testing three models with a sample of blacks. Journal of Black Psychology, 21(4), 332-356.
Dates, J. (1980). Race, racial attitudes and adolescent perceptions of black television characters. Journal of Broadcasting, 24(4), 549-560.
Davis, J. F. (1999). Effectiveness of Internet advertising by leading national advertisers. In D. W. Schuman, & E. Thorson (Eds.), Advertising and World Wide Web (pp. 81-95). Mahwah, NJ: Erlbaum.
DePriest, T. (2000, June 30). Narrowing the gap. Atlanta Tribune: The Magazine, pp. 38, 40.
Desphande, R., & Stayman, D. (1994). A tale of two cities: Distinctiveness theory and advertising effectiveness. Journal of Marketing Research, 31,(4) 57-64.
Eagly, A. H., Wood, W., & Chaiken, S. (1978). Causal inferences about communicators and their effect on opinion change. Journal of Personality and Social Psychology, 36, 424-435.
Ettema, J. S. (1985). Explaining information system use with system-monitored vs. self-reported use measures. Public Opinion Quarterly, 49, 381-387.
Eveland, W. P., & Dunwoody, S. (1998). Users and navigation patterns of a science world wide Web site for the public. Public Understanding of Science, 7, 285-311.
Fannin, R. (1989). Reach vs. environment. Marketing & Media Decisions, 24, 42-48.
Feick, L., & Higie, R. A. (1992). The effect of preference heterogeneity and source characteristics on ad processing and judgments about endorsers. Journal of Advertising, 21(2), 9-24.
Feldman, J. M. (1972). Stimulus characteristics and subject prejudice as determinants of stereotype attribution. Journal of Personality and Social Psychology, 21(3), 333-340.
Gadsden, S. (1985, December 19). Blowing the whistle on media placement. Advertising Age, 56(3), pp. 26, 28.
Gaertner, S. L., & Dovidio, J. F. (1986). The aversive form of racism. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 61-89). Orlando, FL: Academic Press.
Green, C. L. (1999). Ethnic evaluations of advertising: Interaction effects of strength of ethnic identification, media placement, and degree of racial composition. Journal of Advertising, 28(1), 49-64.
Greenberg, B. S. (1986). Minorities and the mass media. In J. Bryant, & D. Zillmann (Eds.), Perspectives on media effects (pp. 165-188). Hillsdale, NJ: Erlbaum.
Greenberg, B. S., & Atkin, C. (1982). Learning about minorities from television: A research agenda. In G. Berry & C. Mitchell-Kernan (Eds.), Television and the socialization of the minority child (pp. 215-243). New York: Academic Press.
Grier, S. A., & Brumbaugh, A. M. (1999). Noticing cultural differences: Ad meanings created by target and non-target markets. Journal of Advertising, 28(1), 79-93.
Grier, S. A., & Deshpande, R. (2001). Social dimensions of consumer distinctiveness: The influence of social status on group identity and advertising persuasion. Journal of Marketing Research, 38, 216-224.
Harris, R. (1981). Focus groups offer six guidelines for black-oriented ads, Marketing News, 5, p. 14.
Henry, S. (1999, August 12). BET plans site for African Americans: Web offering in works amid talk of 'digital divide' between whites, minorities. Washington Post, p. E1.
Hindman, D. B. (2000). The rural-urban digital divide. Journalism and Mass Communication Quarterly, 77(3), 549-560.
Hoffman, D. L., & Novak, T. P. (1998). Bridging the digital divide: The impact of race on computer access and Internet use. Retrieved August 4, 2003, from Vanderbilt University, Owen Graduate School of Management Web site: http://elab.vanderbilt.edu/research/ papers/html/manuscripts/race/science.html
Hoffman, D. L., Novak, T. P., & Schlosser, A. E. (2001). The evolution of the digital divide: Examining the relationship of race to Internet access and usage over time. In B. M. Compaine (Ed.), The digital divide: Facing a crisis or creating a myth (pp. 47-97). Cambridge, MA: MIT Press.
Hovland, C., & Weiss, W. (1951). The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 15(4), 635-660.
Hubbard, L. (2000, March 3). Is the digital divide a black thing? Salon News. Retrieved September 21, 2000, from http:www.salon.com/news/feature/2000/03/02/digital/ index.html
Huesman, L. R., Eron, L. D., Klein, R., Brice, P., & Fischer, P. (1983). Mitigating the imitation of aggressive behaviors by changing children's attitudes about medial violence. Journal of Personality and Social Psychology, 44, 899-910.
Joyce, C. (1997, June 5). Race matters in cyberspace too. Salon Magazine. Retrieved September 21, 2000, from http://www.salon.com/june97/21st/race970605.html
Kaiser Family Foundation. (2001). Survey of Americans on technology. In B. M. Compaine (Ed.), The digital divide: Facing a crisis or creating a myth (pp. 274-277). Cambridge, MA: MIT Press.
Kelman, H. C. (1961). Process of opinion change. Public Opinion Quarterly, 25(1), 57-78.
McConahay, J. B. (1986). Modern racism, ambivalence, and the modern racism scale. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 91-126). Orlando, FL: Academic Press.
McConnaughey, J. W., & Lader, W. (1998). Falling through the net II: New data on the digital divide. Retrieved August 4, 2003, from National Telecommunications and Information Administration, United States Department of Commerce Web site: http://www. ntia.doc.gov/ntiahome/net2/falling.html
McGuire, W. (1984). Search for the self: Going beyond self-esteem and the reactive self. In R.A. Zucker, J. Aronoff, & A. I. Rabin (Eds.), Personality and the Prediction of Behavior (pp. 73-120). New York: Academic Press.
McGuire, W., McGuire, V., Child, P., & Fujioka, T. (1978). Salience of ethnicity in the spontaneous self-concept as a function of one's ethnic distinctiveness in the social environment. Journal of Personality and Social Psychology, 36(5), 511-520.
Miller, P., & Miller, R. (1992). Trends are opportunities for targeting African-Americans. Marketing News, 26(2), p. 9.
Nie, N. H., & Erbring, L. (2001). Internet and society: A preliminary report. In B. M. Compaine (Ed.), The digital divide: Facing a crisis or creating a myth (pp. 268-271). Cambridge, MA: MIT Press.
Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of Adolescent Research, 7(2), 156-176.
Pitts, R. E., Whalen, D. J., O'Keefe, R., & Murray, V. (1989). Black and white response to culturally targeted television commercials: A values-based approach. Psychology & Marketing, 6(4), 311-328.
Raney, R. F. (1998, May 10). Studies reach contradictory conclusions about the internet populations. New York Times, p. B10. Retrieved November 9, 1999, from http://www. nytimes.com/library/tech/98/05/cyber/articles/10race.html
Rice, R. E., & Borgman, C. L. (1983). The use of computer-monitored data in information science and communication research. Journal of the American Society for Information Science 34, 247-256.
Rossman, M. L. (1994). Multicultural marketing: Selling to a diverse America. New York: American Management Association.
Royce, A. P. (1982). Ethnic identity: Strategy of diversity. Bloomington: Indiana University Press.
Schement, J. R. (2001). Of gaps by which democracy we measure. In B. M. Compaine (Ed.), The digital divide: Facing a crisis or creating a myth (pp. 303-307). Cambridge, MA: MIT Press.
Schlinger, M. J., & Plummer, J. T. (1972). Advertising in black and white. Journal of Marketing Research, 9, 149-153.
Smedley, J. W., & Bayton, J. A. (1978). Evaluative race-class stereotypes by race and perceived class of subjects. Journal of Personality and Social Psychology, 36(5), 530-535.
Soley, L. (1983). The effect black models in magazine ad readership. Journalism Quarterly, 48 (2), 337-339.
Study reveals blacks trust black media most and magazines have highest trust level of all (1998, November 2). Jet, p. 57.
Sundar, S. S. (2000). Multimedia effects on processing and perception of online news: A study of picture, audio, and video downloads. Journalism & Mass Communication Quarterly, 77 (3), 480-499.
Sutel, S. (1999, August 13). Online inclusion: BET and backers aims to bridge Internet's digital divide. Las Vegas Review-Journal. Retrieved August 4, 2003, from http://www.Ivrj.com/ Ivrj_home/1999/Aug-13-Fri-1999/business/11748336.html
Thorson, E., Wells, W. D., & Rogers, S. (1999). Web advertising's birth and early childhood as viewed in the pages of advertising age. In D. W. Schuman & E. Thorson (Eds.), Advertising and World Wide Web (pp. 5-25). Mahwah, NJ: Erlbaum.
True colors: Cultural factors influence the online behavior of ethnic consumers (2001, April). American Demographics, 23(4), 53-60.
Walsh, E. O., Gazala, C. H., & Ham, C. (2001). The truth about the digital divide. In B. M. Compaine (Ed.), The digital divide: Facing a crisis or creating a myth (pp. 265-268). Cambridge, MA: MIT Press.
Where blacks, whites diverge (1993). Brandweek 34(18), p. 22.
Whittler, T. E. (1989). Viewers' processing of actor's race and message claims in advertising stimuli. Psychology of Marketing, 6, 287-309.
Whittler, T. E. (1991). The effects of actors' race in commercial advertising: Review and extension. Journal of Advertising, 20(1), 54-60.
Whittler, T. E., & Dimeo, J. (1991). Viewers' reactions to racial cues in advertising stimuli. Journal of Advertising Research, 31(6), 37-46.
Whittler, T. E., & Spira, J. S. (2001). Model's race: A peripheral cue in advertising messages? Journal of Consumer Psychology, 12(4), 291-301.
Osei Appiah (Ph.D., Stanford University) is an Assistant Professor in the School of Journalism and Communication at the Ohio State University. His research interests include media effects on ethnic minorities, and the impact of ethnic identity on audiences' responses to media.
An earlier draft of this paper was presented at the annual convention of the Association for Education in Journalism and Mass Communication in August, 2002, where it won Top Paper in the Minorities and Communication Division. The author wishes to thank Tracy and Tazmin Appiah for their inspiration.…