***Must have an understanding of Social Science/Psychology***
(use the Portrayals of Minorities and Women in Super Bowl Advertising article). How have women been portrayed? How have people of a minority background been portrayed? What theory links advertising to stereotyping or negative attitude acquisition?
I have provided you with the article
(you don’t have to search for the article)
- I have provided you with the video lecture for the paper
***Must have an understanding of Social Science/Psychology*** (use the Portrayals of Minorities and Women in Super Bowl Advertising article). How have women been portrayed? How have people of a minori
Social Psychology Observational Research Project Assignment Final Paper The paper is organized into different sections (3-5 pages maximum). Answer each of the bulleted points as thoroughly as you can. Don’t skip any bullets. Organize your response into paragraph form and use the sections headings in your paper. Please cite all sources and review the plagiarism policy in the syllabus. Paraphrase and cite material or use quotes and a citation when taking material word for word. This course uses SafeAssign to check for plagiarism. A. Create a title page. Have a running head and number the pages. B. Literature review (please cite all sources in APA style) Provide an introduction to issues surrounding the use of women and men in advertising (use the Portrayals of Minorities and Women in Super Bowl Advertising article). How have women been portrayed? How have people of a minority background been portrayed? What theory links advertising to stereotyping or negative attitude acquisition? What were the important results of the study you read as part of the Information Literacy assignment? If there are many, focus in on just the main results you think are important to your project. What are the implications of the research in relation to your own project topic? What is your hypothesis(es)? (Can you make a prediction, based on the literature you read, as to how the two variables in your selected topic will be related; for example, how will gender be displayed across the different webpages? C. Methods Describe our sample. What was observed and under what conditions? Describe your procedure. What were the major categories coded in the study? Describe methods for collecting data? (what the coding sheet looked like and how it worked) D. Results What were the main findings from the class data set for your selected topic? Describe and summarize the data and analyses that relates to your topic. E. Conclusion (please cite all sources in APA style) In regards to your hypothesis, does the data support or refute your hypothesis/ prediction? How did things turn out? How do the results from the class data set on your topic relate to the articles you read? Does the data support or refute the prior studies’ findings from your literature review (be specific)? What are the implications and /or applications of your research? What were the study limitations? What would you do differently next time? F. Reference Please put all the articles used in the paper into an APA formatted reference page.
***Must have an understanding of Social Science/Psychology*** (use the Portrayals of Minorities and Women in Super Bowl Advertising article). How have women been portrayed? How have people of a minori
WINTER 2019 VOLUME 53, NUMBER 4 1535 CHARLES R. TAYLOR, ALEXANDER MAFAEL, SASCHA RAITHEL, CARISSA M. ANTHONY, AND DAVID W. STEWART Portrayals of Minorities and Women in Super Bowl Advertising Portrayals of women and minorities advertising have long been of inter- est to advertising scholars. While research has found that the overall representation of these groups has increased, some stereotypes persist, and so do questions about the quality and prominence of portrayals. This study examines portrayals of minorities and women in Super Bowl advertising, the main “pop culture” showcase for US advertising. A con- tent analysis of 10 years of Super Bowl ads is conducted and a multino- mial logit regression model is employed to delve deeper into the content analysis results. Findings show that while the overall representation of women and various minority groups is strong, a deeper analysis shows that these groups are seldom depicted as primary characters by them- selves and that some subtle stereotypes persist. We also find that ads featuring female principal characters are more likely to feature home settings, sexual appeals, emotional messages, and music as a major ele- ment and that it is less likely for female (vs. male) celebrities to be used. Minority principal characters are more likely to be celebrities and be included in ads with music or for technical products but are less likely to be included in ads featuring corporate social responsibility messages. Scholars have been studying race and gender stereotypes in advertising portrayals for many decades. Historically, this research has observed that minorities and females are underrepresented and, in many instances, portrayed unfavorably or in a stereotypical manner (Belkaoui and Belkaoui 1976; Eisend 2010; Taylor and Stern 1997). Generalizations made from stereotypical portrayals may contribute to unrealistic expectations and reduce perceived opportunities for these groups (Grau and Zotos 2016; Taylor, Grau and Bang 2005). Although recent studies suggest that extreme forms of minority and female stereotyping in ads have decreased and that Charles R. Taylor ([email protected]) is the John A. Murphy Professor of Marketing at Villanova University. Alexander Mafael Germany ([email protected]) is a postdoctoral researcher in Marketing at Freie Universität Berlin. Sascha Raithel ([email protected]) is Professor of Marketing at Freie Universität Berlin. Carissa M. Anthony ([email protected]) is a Graduate Research Fellow at Villanova University. David W. Stewart is the President’s Professor of Marketing and Business Law at Loyola Marymount University ([email protected]). The Journal of Consumer Affairs, Winter 2019: 1535–1572 DOI: 10.1111/joca.12276 Copyright 2019 by The American Council on Consumer Interests 1536 THE JOURNAL OF CONSUMER AFFAIRS inclusion has been on the rise, some stereotypes persist (Furnham and Lay 2017; Smith, Choueiti, and Stern 2011). In the face of demographic changes, it is timely to examine the degree to which stereotypes persist in advertising for multiple reasons. First, individuals currently categorized as minorities will comprise more than 50% of the US population by the year 2044 (Colby and Ortman 2017). Second, the perceived role of women in the workplace and home has shifted considerably in recent years (Grau and Zotos 2016). Presently, women constitute 47% of the workforce and control 60% of personal wealth (U.S. Department of Labor 2010). A primary goal of advertising is to resonate with the consumers being targeted. Thus, it is important to examine whether portrayals are also evolving and reflect these changes. Moreover, if stereotyped advertising reflects an outdated worldview it may no longer resonate with modern consumers. Given the historically low frequency of stereotypes present in the portrayal of minorities and women and in advertising, the purpose of this study is to examine and assess the depiction of these groups during the Super Bowl. The Super Bowl represents a rare “big event” in a fragmented media environment that is watched with heightened attention. Thus, these portrayals are a component of US popular culture, reflective of current trends and attitudes. To meet our research objectives, a content analysis was performed on Super Bowl ads for the 10-year period of 2008–2017. The longitudinal nature of the study allows us to look not only at averages over this 10-year period, but also to investigate the occurrence of any trends in portrayals over that time. Specific variables investigated include the frequency of portrayals of women and minorities as principal characters in combination with advertising content features such as product type, setting, message appeal, use of celebrities, CSR messages, and music. Thus, we analyze a comprehensive set of content elements that shape the appearance of the advertisement. The study contributes to the literature in the following ways. First, at a societal level, concerns have been raised as to whether stereotypes of women and minorities persist in advertising. Our data allows for an evalu- ation of the degree to which stereotypes exist that may create expectancies that pressure a group or cultivate a distorted view of a group. Second, the study examines whether managerial practices reflect societal trends and attitudes toward ethnicity and gender. Our analysis allows for an assess- ment of whether advertisers have adapted to changing demographics and other societal changes over time in an effort to better appeal to female and minority consumers. Finally, this research contributes by expanding and updating the longstanding literature on portrayals of women and minorities WINTER 2019 VOLUME 53, NUMBER 4 1537 in advertising while including additional content variables in the analysis. Given the continuous societal change, it is important to explore additional variables in order to more fully comprehend stereotyping in advertising. This, coupled with updated findings on previously studied variables, allows us to examine changes over time and maintain a contemporary body of knowledge about these issues. The remaining paper is structured as follows. We begin with a brief summary of prior studies of the portrayals of minority groups and women in advertising. Next, we describe two theoretical perspectives that provide the background for our investigation. Further, we describe our empirical setting, provide details of the data collection and analysis and discuss our results. Finally, we outline our contribution and identify opportunities for further research. LITERATURE REVIEW Portrayals of Minority Groups in Advertising In a historical review of the literature on marketing and minorities, Davis (2018, 157) observes that, “Viewed over time, studies suggest that aspects of racism can be overt, covert, or unintended but are systematically embedded in the fabric of marketing organizational cultures such that they often go unrecognized.” Advertising portrayals are a practice that deserves scrutiny, especially since it has been observed that stereotyped portrayals have been present in advertising throughout modern history (Davis 2018). As used here, a stereotype refers to an “oversimplified idea about a group based on some preconceived assumptions” (Morrison 2014, 150). A stereotyped portrayal refers to an advertisement that depicts a stereotype, whether overtly or in a more subtle fashion. In the past, minorities have been found to be underrepresented in tele- vision advertisements in comparison to their population proportion, with minorities being more likely to be portrayed in minor or background roles in comparison to majority models (Greenberg, Mastro, and Brand 2002; Mastro and Tukachinsky 2011; Wilkes and Valencia 1989). In addition, minorities have sometimes been portrayed in a stereotypical manner, such as in limited product categories or specific settings. For example, Asian Americans have appeared disproportionately in ads for technology-based products and in business settings, but only infrequently in family or social settings (Mastro and Stern 2003; Taylor, Landreth, and Bang 2005; Taylor and Stern 1997). Meanwhile, African Americans and Latinos are more likely to appear in ads with social settings as opposed to business setting. 1538 THE JOURNAL OF CONSUMER AFFAIRS Although the frequencies of appearance of minority groups have increased over time, some stereotypes persist. On one hand, African Americans and Asian Americans are well represented in television adver- tising relative to their ratio in the US population and are being shown in more favorable and diverse roles and settings (Taylor and Stern 1997). In contrast, Latinos are still underrepresented in the mainstream media (Mastro and Stern 2003). Recent studies have also found that Asian Amer- icans and Latinos continue to be portrayed in stereotypical roles, though the stereotypes are less blatant than in the past (Mastro and Stern 2003; Tukachinsky, Mastro, and Yarchi 2015). Branchik and David (2018) find that portrayals of African American males have increased dramatically his- torically, but that this group remains overrepresented as athletic figures and celebrities while being underrepresented in any romantic situations. Latinos are likely to be portrayed in home settings and in ads promoting electronics and food (King 2012) while Asians are more likely to be por- trayed in work settings and relationships. In sum, the research suggests that some groups of minorities remain underrepresented and represented in a stereotypical manner and that, while there have been improvements, some propose that such practices can reinforce notions of white superiority (Davis 2018). Portrayals of Women in Advertising Historically, women were underrepresented in advertisements and only portrayed in limited roles and settings (Belkaoui and Belkaoui 1976). Advertisers portrayed women in decorative roles and highlighted youthful and sexual characteristics to spotlight their physical appearance (Uray and Burnaz 2003). Although women were often not associated with an occupa- tion, they were likely to be placed in more family-oriented or home settings to promote household products or play the role of the household shop- per (Ganahl, Prinsen, and Netzley 2003; Uray and Burnaz 2003). Sivulka (2009) observes that increased participation of women in the advertising industry, especially at top levels has led to some advances. However, as observed by Jhally and Kilbourne (2010), concerns about advertising por- trayals of women and their relationship to sexism, gender-related violence, and eating disorders remain. Recent research shows that women are becoming more prominent in advertising portrayals. The proportion of women included in advertise- ments has increased and they are being seen in more settings outside of the home (Hatzithomas, Boutsouki, and Ziamou 2016; Mager and Helge- son 2011). A content analysis of 275 prime time and popular television WINTER 2019 VOLUME 53, NUMBER 4 1539 programs revealed that while the percentage of female characters has increased, it still fell short of achieving gender equality (Smith, Choueiti, and Stern 2011). Another study analyzed prime time TV content for major networks from 2003 to 2008 and revealed that the proportion of men depicted as the main character declined but was still a majority (Fowler and Thomas 2015). While these examples highlight how women’s rep- resentation and portrayals in more diverse settings have been increasing, recent studies still point to portrayals of women being valued for physical appearance and attractiveness (Furnham and Lay 2017; Mager and Helge- son 2011; Smith, Choueiti, and Stern 2011). Further, women are more prone to appear in advertisements that focus on product categories that promote body, beauty, and personal care or cleaning products (Furnham and Lay 2017; Matthes, Prieler, and Adam 2016). THEORETICAL PERSPECTIVES ON THE IMPACT OF PORTRAYALS In examining potential impacts of stereotyped portrayals, we employ two theoretical perspectives, expectancy theory and cultivation theory. Together, these two theories cast insight on the impacts of stereotypes portrayals both at an individual level and societal level. Expectancy the- ory is effective in helping understand the impacts repeated exposures to stereotypes can have on an individual person. Meanwhile, cultivation theory takes a more collective view, focusing on the overall impact on society. Expectancy Theory Stereotypes grounded in ethnicity and/or gender can elicit negative attitudes when they are held against a group (Ashmore and Del Boca 1981; Esses, Haddock, and Zanna 1993; Vinacke 1957). Expectancy theory (Jussim 1990) proposes that stereotypes are reinforced upon repeated exposure to advertising portrayals, leading to expectations consistent with the portrayal of a group. Women and minorities exposed to stereotypes repeatedly portrayed in advertising may feel pressure to live up to or conform to these stereotypes (Peterson 2007). For example, when a woman is exposed to repeated images of “ideal” models in terms of body shape she may experience pressure to live up to this image. Similarly, when Asian Americans see repeated “model minority” images of Asian models in ads for technical products and business settings/relationships, they may feel pressure to excel at math or as science students, or they may feel left out and 1540 THE JOURNAL OF CONSUMER AFFAIRS not “cool” within society (Phua 2014; Taylor, Landreth, and Bang 2005; Zhang 2010). In general, failing to meet expectations evoked by portrayals may result in stress for minority groups (Graham 1983). Cultivation Theory Cultivation theory (Gerbner et al. 1980) suggests that, over time, exposure to stereotyped portrayals in television programming and advertisements cultivates individuals’ perceptions of reality. In other words, exposure to advertising influences peoples’ perceptions of what other people in the world are like (Morgan and Shanahan 2010). Cultiva- tion theory implies that repeated exposure to stereotypes of minorities and females can result in inaccurate beliefs and judgments about those groups (Mastro, Behm-Morawitz, and Ortiz 2007). If a stereotype is repeated in advertising, viewers process this information (Bandura 2002) and are prone to accept the stereotype as reality. For example, if women are shown repeatedly in subservient roles, this can create a perception that this is the normal state of affairs. Thus, overrepresentation of a group in certain settings, occupations, or relationships, can affect and potentially influence viewer perceptions (Bang and Reece 2003). Similarly, the lack of portrayal of a group can cultivate perceptions. For example, if Latinos are shown very infrequently in managerial posi- tions in ads, this can cultivate the idea that few Latinos can hold such occupations (Bang and Reece 2003). The cultivation of stereotypes not only affects minority groups but also perceptions of the majority. Stereo- types can become reflective of dominant societal attitudes that can promote inequality and favoritism. With this in mind, we put forward research ques- tions based on prior literature. RESEARCH QUESTIONS The variables analyzed in this study primarily represent those most fre- quently analyzed in prior content analysis studies of women and minority groups. Both overall representation of groups, as well as role prominence, have been regarded as important variables for assessing how visible the groups are in society. Additional variables frequently analyzed in portrayal studies (see above) have included product type, settings, sexual appeals, use of celebrities and relationships among characters. In addition to these variables, we also analyzed some executional variables commonly used in Super Bowl ads, including humor, CSR messages, and use of music. Some variables, such as depictions of Lesbian, Gay, Bisexual and Transgender WINTER 2019 VOLUME 53, NUMBER 4 1541 (LGBT) characters were not included due to the limited frequency of these portrayals in the Super Bowl data. Presence and Role Prominence The literature on advertising portrayals of minorities and women gen- erally suggests that there has been an increase in the frequency of appear- ances relative to proportionality (Briggs 2008; Peterson 2007; Stevenson and Swayne 2011; Taylor and Stern 1997). While African Americans and Asians have been found to be well represented overall in advertising in accordance to their proportion to the US population, Latinos continue to be underrepresented, and South Asian and Middle Eastern groups have been reported to be nearly nonexistent (Taylor, Landreth, and Bang 2005). The proportionality criterion (Wilkes and Valencia 1989) has been widely used to predict the frequency of group representation in comparison to the group’s representation in the general population. In the context of this research, it refers to the frequency in which minority groups and women appear in ads in comparison to their proportion in the US population. While the overall representation of African Americans and Asian Americans has been found to exceed proportionality in recent research, minority groups have been found to be substantially underrepresented in terms of their pro- portion of appearances in major roles relative to Caucasian models (Peter- son 2007; Taylor, Landreth, and Bang 2005). Though there has been some increase over time in the proportion of minority models in major roles over time (e.g., Bailey 2006; Peterson 2007; Stevenson and Swayne 2011), it is widely reported that there is still more backgrounding of minority models. For our first research question, we employ the proportionality criterion to examine whether the percentage of ads in which a minority group is portrayed exceeds their overall representation in the US population. Recent research also suggests that gender equality in advertising has improved and that while women and minorities are being represented more, they appear on a less than equal basis (Rubie-Davies, Liu and Lee 2012). Thus: RQ 1a: Are (a) females and (b) minorities portrayed in Super Bowl advertising in proportion to their overall proportion of the population? RQ 1b: Are (a) females and (b) these groups portrayed inmajor rolesin proportion to their overall proportion of the population? Product Category Some previous research has focused on the relative frequency of por- trayals of minority groups in technical vs. nontechnical product categories 1542 THE JOURNAL OF CONSUMER AFFAIRS (e.g. Bailey 2006). Most studies show that Caucasians are more evenly represented between technical and nontechnical products, in comparison to minority groups. Asian Americans and Latinos continue to be paired with stereotypical products (e.g. technology and electronic products, and food, respectively; Mastro and Stern 2003; Taylor, Landreth, and Bang 2005; Tukachinsky, Mastro, and Yarchi 2015). Expectancy theory sug- gests that the pairing of a product category with a minority group may set societal expectations and result in negative wellbeing experienced by the minority group. Concerning females, research has shown that women are paired disproportionately more with beauty and personal care prod- ucts (e.g. Furnham and Lay 2017; Matthes, Prieler, and Adam 2016). Thus: RQ 2: Are (a) females and (b) minorities represented proportionately between tech- nical and nontechnical product categories? Settings Prior research demonstrates that Latinos and African Americans appear in more social settings than Asian Americans, who commonly appear in business settings and are depicted less frequently in family or social settings (e.g. Rubie-Davies, Liu, and Lee 2013; Taylor, Landreth, and Bang 2005). Drawing on expectancy and cultivation theory, depiction in these stereotyped settings can result in negative implications for how these groups are viewed in accordance to how they value work vs. familial and social relationship priorities. Similarly, we look to investigate whether women are portrayed across different settings, as stereotypes can operate in the same manner (Ganahl, Prinsen, and Netzley 2003; Uray and Burnaz 2003). Especially, for women, it is worthwhile to examine the relative proportion in family and social settings vs. in business settings because of traditional roles. Thus: RQ 3: Are (a) females and (b) minorities represented proportionately across settings in comparison to males and Caucasians, respectively? Celebrities, Appeal Type, Corporate Social Responsibility and Music In addition to the above advertising content variables that have been analyzed in prior studies of female and minority portrayals, we are inter- ested in examining the frequency of the appearance of women and minori- ties in ads using appeals that have historically been commonly used in Super Bowl ads including celebrity endorsements, humor, sexual appeals WINTER 2019 VOLUME 53, NUMBER 4 1543 and rational vs. emotional appeals (Yelkur et al. 2013). Two additional issues worth examining is minority representation in ads that include a corporate social responsibility (CSR) message and ads that use music. Thus: RQ 4: Are (a) females and (b) minorities represented proportionately regarding com- monly used advertising content features in the Super Bowl (celebrity endorsement, humorous appeals, sexual appeals, CSR messages, and use of music)? Trend Analysis In addition to examining mean levels of representation, we also are interested in analyzing any trends that occur in the variables outlined above. For example, it is interesting to examine whether female characters appear relatively more frequently in workplace settings vs. the home in more recent years. This analysis should reveal whether the portrayal of females and minorities has changed in recent years and whether stereotypes have become more or less common. As a result, we pose the following broad research question: RQ 5: Are there significant trends in the portrayal of (a) females and (b) minorities in terms of overall representation; representation in: major roles, product categories, set- ting, relationships; and representation in ads using celebrity endorsement, humorous appeals, sexual appeals, CSR messages, and music? Figure 1 provides a summary of the variables analyzed. METHOD AND DATA The Super Bowl as a Major Advertising Event The Super Bowl is an appropriate context for this study because it is a dominant media event that consistently commands broadcast attention. The 2018 Super Bowl was viewed by an average of 103.4M viewers or 68% of US homes, with streaming platforms lifting overall viewership to 106 million (Crupi 2018). Female viewership accounts for 46% of the overall audience (Nielsen 2018). Viewers pay close attention to the ads, as evidenced by Nielsen (2018) reporting that a mere 41% of viewers felt the game was the most important part of the show. It is also the case that some ads go viral and reach even larger audiences (Picchi 2018). Marketers face high stakes, paying $5.2 million for a 30-second ad as of 2019 (CNBC 2019). In this vein, examining portrayals of minorities and women in the Super Bowl is important because such portrayals represent advertisers’ best efforts in front of the largest audience. 1544 THE JOURNAL OF CONSUMER AFFAIRS FIGURE 1 Conceptual Framework Type of Appeal Sexual Humorous Emotional vs. rational Setting Outdoor/nature Business Home (inside/outside) Social (outside home) Multiple Other CSR Message Role of Music Role of Celebrity Type of Product (technical vs. non-technical) Gender Only female Both gender Only male Ethnic Group Only minority Mixed (minority and Caucasian) Only Caucasian Time Trend – – – – – – – – – – – – – – – Model Set-Up The statistical tests pertaining to the distribution of portrayals of women and minorities relative to the overall population can be achieved by Chi-square tests. To test the other research questions about the proportional presentation of women and minorities in relation to ad content features such as type of product or appearance of celebrities, we apply a multinomial logit model: Principal Character(gender∕ethnicity)= 0+ 1·Year Trend + 1·Type of Product+ T 2·Setting+ 3·Celebrity + 4·Sexual Appeal+ 5·Humorous Appeal + 6·Emotional vs.Rational Appeal+ 7·CSR+ 8·Music+ The logic of this model is the following. The dependent variableType of Principal Characteris either the (a) gender or the (b) ethnic group of the principal character(s) in the ad. For each outcome, there are three possible values because the ad can feature only male models, only female models, or female and male models. In the same way, the ad can feature only Caucasian, only minority, or a mix of Caucasian and minority principal characters. For the gender of the model, we use the most frequent value WINTER 2019 VOLUME 53, NUMBER 4 1545 “male-only” ad as the baseline condition. For the ethnic group model, we select the most frequent “Caucasian-only” ad as the baseline condition. If any of the other groups appear more or less frequently together with any of the content features in the ads compared to the baseline condition, the respective regression coefficient is significant. For example, if the regression coefficient 4ofSexual Appealwas significantly positive in the gender model’s equation for the “female-only” ads, then female principal characters appear more frequently in ads with sexual appeals compared to the baseline condition “only male models.” The multivariate model set-up is advantageous compared to bivariate Chi-square tests because we can control for and study various content features and their relative effects simultaneously. Ye a r Tr e n dmeasures whether the proportion of gender/ethnic group representation in the ads changes over time.Type of Productmeasures whether the ad features technical products or not. The vectorSetting contains dummy variables for the different settings such as business, social, and outdoor. The variableCelebritymeasures the prominence of celebrities in the ad. The variableSexual Appealmeasures whether a sexual appeal is present or not.Humorous Appealmeasures whether the ad contains humor or not.Emotionalvs.Rational Appealmeasures whether the ad has a stronger emotional or stronger rational appeal.CSRmeasures the presence of a CSR message in the ad.Musicmeasures the prominence of music in the ad. We estimate the models with heteroscedasticity-robust standard errors. To test whether any of the features changes its relationship with the appearance of gender/ethnic groups over time, we estimate each model also with the interactionYe a r Tr e n d×Feature. Since the number of inter- action terms would be excessive and create multicollinearity problems, we estimate the interaction terms one by one. Data Coding Instrument and Procedure Following general procedures for conducting content analysis outlined by Krippendorf (2013), the authors created a codebook of operational definitions for the measured variables (see Appendix 1 for the data coding instrument and Appendix 2 for the codebook). 1 Paid student research assistants were employed as coders, and for purposes of reliability, each coder coded all ads in the sample (Kolbe and 1. The authors attempted to code mixed race models, Native American models, and LGBT models; however, the reliability of the mixed-race group was insufficient and the frequency of Native American and LGBT models too low (less than 1%) to run statistical tests. 1546 THE JOURNAL OF CONSUMER AFFAIRS Burnett 1991). The coders were trained extensively by the research team using 30 prime time television ads that were not Super Bowl ads and at least 3 years old. The coders independently coded every national Super Bowl ad that ran between 2008 and 2017. Only ads for movies and video games were excluded from the analysis as these ads tend to follow a standard “movie trailer” format that follows a storyline rather than employing other ad appeals. Reliability for all of the measured variables was assessed using Cohen’s kappa, which takes the number of coding categories into account and corrects for chance agreement between coders (Hughes and Garrett 1990). It is a more accurate quantitative reliability measure than a raw percentage agreement score. Kappa values for the coding categories ranged between 0.53 and 0.90 with an average value of 0.71. Although there is no general guideline to judge Cohen’s kappa, the average agreement between both coders can be considered as substantial (Landis and Koch 1977). Disagreements on all coded variables were resolved through discussion between the coders and one of the researchers. The model variables are the following: Gender of Principal Characterused a nominal scale and measures the presence of only female, only male, and mixed (female and male) principal characters in the ad. Ethnic Group of Principal Characterused a nominal scale and measures the presence of the principal character representing only ethnic minority groups (African, Latino, Asian, South Asian, Middle Eastern; Mixed ethnicity), only Caucasian, or a mix of minority and Caucasian. Perceived Importance of Minority Character. For each ethnic group, we additionally measured on an ordinal scale whether the character not only appears as a primary character but also plays a major, minor, or background role. Ye a r Tr e n dis a metric variable and measures the year of the Super Bowl ad. Type of Productis binary and measures whether the featured product is technical (1) or not (0). Settingis a nominal variable, which measures whether the setting is outdoor/nature, business, home (in- or outside), social (outside the home), multiple or other. This variable enters the model as a set of five dummy variables. The most frequent setting serves as the baseline condition. Celebrityis ordinal and measures whether celebrity characters are not present (0) or play a minor (1) respectively a major (2) role in the ad. Sexual Appealis binary and measures whether the ad contains sexual appeal (1) or not (0). WINTER 2019 VOLUME 53, NUMBER 4 1547 Humorous Appealis binary and measures whether the ad contains humor (1) or not (0). Emotionalvs.Rational Appealis ordinal and measures whether the appeal is more emotional (1), balances emotional and rational aspects (0) or is more rational (−1). CSRis ordinal and measures whether a CSR message is not present (0) or plays a minor (1) respectively a major (2) role in the ad. Music is ordinal and measures whether music is not present (0) or plays a minor (1) or a major (2) role in the ad. EMPIRICAL FINDINGS General Appearance of Females and Minorities Over the 10-year period from 2008 to 2017, 523 unique Super Bowl ads were coded, 406 of which featured human models as principal characters. Table 1 shows the perceived importance of minority groups. Table 2 shows the ad number distribution of the principal characters by gender and ethnic groups. Table 3 shows all raw frequencies for gender and ethnic groups by year and ad content features. In terms of overall representation of groups, Table 1 shows that more than half of the ads (54.1%) in the sample included an African Ameri- can model, while 21.4% included an Asian American model, and 18.5% included a Latino model. Middle Eastern models (4.8%) and South Asian models (6.9%) were also present in a significant number of ads. All numbers for African American, Latino, and Asian minorities exceed the proportionality criterion. Out of the total population according to 2018 census figures, Hispanic/Latino accounted for 16.3%, African Americans 12.6%, Asian, Hawaiian and Pacific Islander 4.9%, American Indian and Alaska Native 0.7%, and another racial group 2.0%. While the South Asian and Middle Eastern groups are not reported separately by the Cen- sus Bureau, it is clear that the numbers for these groups exceed propor- tionality. Table 1 also shows the extent to which each group’s representa- tion exceeds proportionality. While the numbers are+41.5% for African Americans and+16.5% for Asian Americans, the number is just+2.2% for Latinos. Thus, when strictly applying the proportionality criterion, all groups including Latinos, South Asians, and Middle Eastern exceed proportionality. While it is clear that Super Bowl ads depict a great deal of diversity and score well on the overall presence of minority groups, it is important to examine the nature of portrayals in the ads. As can be seen from Table 1, Caucasians appear as principal characters in 86.2% of ads, African 1548 THE JOURNAL OF CONSUMER AFFAIRS TA B L E 1 Perceived Importance of Minority Groups in Super Bowl Ads from 2008 to 2017 Ethnicity of Minority Group African Asian Latino Middle Eastern South Asian Mixed Population proportion (2018 US census) 13.4% 5.8% 18.1% n/a n/a 2.7% Total number of ads with human models as principal characters406 Principal character Ad count 47 8 7 1 2 15 Ad percentage 11.6% 2.0% 1.7% 0.2% 0.5% 3.7% Difference to population−1.8%−3.8%−16.4% n/a n/a 1.0% Total number of ads with human models in any role518 Any appearance Ad count 280 111 96 25 36 89 Ad percentage 54.1% 21.4% 18.5% 4.8% 6.9% 17.2% Difference to population 40.7% 15.6% 0.4% n/a n/a 14.5% At least one major role Ad count 83 17 20 4 7 26 Ad percentage 16.0% 3.3% 3.9% 0.8% 1.4% 5.0% At least one minor role Ad count 77 34 33 9 20 22 Ad percentage 14.9% 6.6% 6.4% 1.7% 3.9% 4.2% At least one background role Ad count 120 60 43 12 9 41 Ad percentage 23.2% 11.6% 8.3% 2.3% 1.7% 7.9% WINTER 2019 VOLUME 53, NUMBER 4 1549 TA B L E 2 Frequencies of Ads by Ethnic Group and Gender From 2008 to 2017 Gender of Principal Character Only Female Both Only Male Total Ethnic group of principal character Only minority 11 20 1 32 Both 12 36 48 Only Caucasian 5 150 171 326 Total 16 182 208 406 Americans appear as a principal character in 10.2% of ads, followed by Asian Americans at 1.6%, Latinos (1.4%), South Asian (0.4%) and Middle Eastern models (0.2%). All of these totals for minorities fall short of the proportionality criterion, indicative of underrepresentation in these capacities. In terms of appearance by prominence in the ad, Table 1 shows that the most common depiction is in background roles. For the South Asian group, minor roles account for 36% of portrayals of this group, while major roles are just 19.4% and background roles 25.0%. For all other minority groups, background roles are the most common level of depiction. Gender As can be seen from Table 2, males were featured as principal characters in a higher percentage of the ads than females (86.0% vs. 14.0%). The analysis shows that “male-only” ads (51.2% of the total number of ads) are most frequent. Ads of 44.8% feature both male and female models as principal characters. Ads with only female principal characters are rarer (3.9%). Hence, female and male models are not equally presented in Super Bowl ads (p<.01). The analysis of the trend over time (see Table 3 and Panel A in Figure 2) shows however that the percentage of ads which feature female and male models together in major roles has increased from 2008–2012 (36.9%) to 2013–2017 (50.1%). This effect is significant because the coefficient of theYe a r Tr e n dvariables is significantly larger than 0 (.147 withp<.01; see column “Mixed gender vs. Male-only” in Table 4). There is also some indication that the percentage of “female-only” ads has increased as well because their percentage doubled from 2008–2012 (2.6%) to 2013–2017 (5.3%). Controlling for the other ad content features, theYe a r Te n dvariable produces a positive coefficient but this effect is not significantly larger than 0 (.90 withp>.10; see column “Female-only vs. Male-only” in Table 4). 1550 THE JOURNAL OF CONSUMER AFFAIRS TA B L E 3 Frequencies for Gender and Ethnic Group Differences by Ad Feature Gender Ethnic Group Ad Content Feature ValueFemale- Only MixedMale- OnlyMinority- Only MixedCaucasian- Only Year 2008 4 8 27 5 10 24 2009 6 24 4 26 2010 21 21 3 4 35 2011 19 24 1 4 38 2012 1 19 18 1 3 34 2013 1 26 16 5 6 32 2014 3 23 22 4 3 41 2015 4 16 15 7 2 26 2016 1 14 21 3 6 27 2017 2 30 20 3 6 43 Technical product No 12 105 126 19 29 195 Yes 4 77 82 13 19 131 Setting Outdoor/nature 3 44 71 7 14 97 Business 1 34 31 5 7 54 Home (in/outside) 2 44 39 9 12 64 Social (outside) 1 15 17 4 3 26 Multiple 5 25 28 4 3 51 Other 4 20 22 3 9 34 Role of celebrities No role 7 105 125 14 20 203 Minor role 16 6 1 2 19 Major role 9 61 77 17 26 104 Sexual appeal No 13 165 205 29 47 307 Ye s 3 1 7 3 3 1 1 9 Humorous appeal No 12 78 88 13 25 140 Yes 4 104 120 19 23 186 Emotional vs. rational appealMore rational 2 7 1 8 Mixed 4 30 42 7 10 59 More emotional 12 150 159 25 37 259 Role of CSR No CSR message 16 169 200 32 46 307 Minor CSR message 6 3 1 8 Major CSR message 7 5 1 11 Role of music No role 1 20 25 2 6 38 Minor role 4 136 143 20 30 233 Major role 11 26 40 2 6 38 Summary While the overall level of females and minorities in either minor or major roles in Super Bowl advertising has improved and often exceeds proportionality, these groups are underrepresented regarding the ads’prin- cipal characters. The typical ad features only male Caucasians as a princi- pal character. Although portrayals of women in Super Bowl ads as major characters are falling short of the proportionality criterion, the time trend WINTER 2019 VOLUME 53, NUMBER 4 1551 FIGURE 2 Gender and Ethnic Group of Principal Characters Over Time Panel A: Gender of principal characters over time Panel B: Ethnic group of principal characters over time 10.3% 0.0% 0.0% 0.0% 2.6% 2.3% 6.3% 11.4% 2.8% 3.8%20.5% 20.0% 50.0% 44.2% 50.0% 60.5% 47.9% 45.7% 38.9% 57.7%69.2% 80.0% 50.0% 55.8% 47.4% 37.2% 45.8% 42.9% 58.3% 38.5% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% 20082009 2010 2011 2012 2013 2014 2015 2016 2017 Pe rc. of ads Year Ge nder of Principal Actor only female both only male 12.8% 0.0% 7.1% 2.3% 2.6% 11.6% 8.3% 20.0% 8.3% 5.8%25.6% 13.3% 9.5% 9.3% 7.9% 14.0% 6.3% 5.7% 16.7% 11.5%61.5% 86.7% 83.3% 88.4% 89.5% 74.4% 85.4% 74.3% 75.0% 82.7% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% 20082009 2010 2011 2012 2013 2014 2015 2016 2017 Pe rce ntage of ads by year Year Ethnic Gro up of Principal Actor only minor ity both only caucasian is pointing toward convergence with the equality criterion at least for ads featuring both female and male models in major roles. Further, the results suggest that the portrayal of minority groups (Table 1) in minor roles is much more common (even overrepresentation) compared to major or prin- cipal roles (underrepresentation) in Super Bowl ads. Unlike for women, a significant time trend of principal models from a minority group does not exist. 1552 THE JOURNAL OF CONSUMER AFFAIRS TA B L E 4 Results for Gender and Ethnic Group Differences Gender of Principal Character a Ethnic Group of Principal Character a Dependent Variable Mixed vs. Male-Only Female-Only vs. Male-Only Mixed vs. Caucasian-Only Minority vs. Caucasian-Only Year trend .147*** (.039) .090 (.103)−.087 (.060) .047 (.065) Technical products .255 (.226)−.882 (.713)−.070 (.343) .077 (.375) Setting Business .875** (.348) .389 (1.223)−.086 (.560) .224 (.632) Home (in/outside) .591*(.329) .609 (.877) .584 (.470) .605 (.643) Outdoor/nature .481 (.429) .287 (1.360)−.473 (.708) .542 (.722) Social (outside) .363 (.359) 1.627** (.818)−1.114 (.681)−.369 (.672) Multiple .367 (.413) 1.085 (.875) .343 (.526)−.192 (.734) Role of celebrities−.122 (.124) .070 (.279) .570*** (.178) .426** (.206) Appeal Sexual 2.212*** (.667) 2.281** (1.082)−2.031*(1.040)−.021 (.727) Humor .079 (.240)−.952 (.651)−.552 (.371) .151 (.437) Emotional .579** (.251)−.691 (.444) .055 (.306)−.006 (.381) Role of CSR .436 (.289)−12.082*** (.794)−.281 (.519)−11.885*** (.488) Role of Music−.114 (.212) 2.056*** (.716) .206 (.347) .764** (.360) Intercept−295.934*** (77.626)−187.052 (206.197) 172.315 (120.859)−97.626 (131.646) Model fit Log-likelihood−299.570−235.843 2(p-value) 1,007.472 (.000) 824.685 (.000) Pseudo-R 2 .111 .076 N406 406 Note:Standard errors in parentheses.aMultinomial logit regression with heteroscedasticity-robust standard errors. *p<.10. **p<.05. ***p<.01. WINTER 2019 VOLUME 53, NUMBER 4 1553 The Relation Between Ad Characteristics and the Appearance of Females and Minorities Table 4 shows the relationship between the various ad content features and the appearance of different gender and ethnic groups in Super Bowl ads. Table 5 reports the time trend for each of these content features. The following discussion separates the gender from the ethnic group variable. Gender Group The number of ads with females only in major roles is rather low (N female =16; see Table 1), which requires a more careful interpreta- tion of significant differences with male-only ads. The few differences that do exist, however, provide a rather consistent picture. Ads featur- ing females only in major roles are rather diverse in terms of setting (coeff. Multiple =1.627,p<.05; see Table 3), are more likely to use sex- ual appeals (coeff. Sexual =2.281,p<.05; see Table 3), and music plays an important role (coeff. Music =2.056,p<.01; see Table 3; the time trend is negative though, see Table 5). CSR messages play a significantly less important role when ads feature females only (coeff. CSR =−12.082, p<.01; see Table 3 and Figure 3). Over time, the amount of ads featuring females only in home settings is decreasing (coeff. Year Trend×Home =−.435, p<.05; see Table 5). We next focus on the differences between ads featuring males only (N male-only =208) and ads featuring both female and male models as principal characters (N female & male =182) (see column “Mixed gender vs. Male-only” in Table 4). Any difference would also be indicative for the diverging portrayal of females in Super Bowl ads but give a slightly different interpretation compared to the female-only ads because in these ads females interact with male models. The results do not show significant differences between the por- trayal of mixed gender groups relative to males only concerning technical products (Table 4: coeff.=.255,p>.10; see also Figure 3: Perc. mixed gender×technical =47.2% vs. Perc. male-only×technical =50.3%). Yet, the time trend is clearly positive, i.e., mixed gender-ads are now more prone to feature technical products than in the past (coeff. Year Trend×Techn. Products =.249,p<.01; see Table 5). Concerning the setting, there are few differences. Male-only ads (major role) feature an outdoor/nature setting (60.2% of ads; see Figure 3) more than mixed-gender ads (37.3%). Looking at business and home settings this relationship reverses, and mixed-gender ads are the most frequent type. These differences are significant (coeff. Business =.875, 1554 THE JOURNAL OF CONSUMER AFFAIRS TA B L E 5 Results for Time Trends in Gender and Ethnic Group Differences Gender of Principal Character a Ethnicity of Principal Character a Dependent Variable Mixed vs. Male-Only Female-Only vs. Male-Only Mixed vs. Caucasian-Only Minority-Only vs. Caucasian-Only Year Trend×Technical products .249*** (.082) .281 (.173)−.088 (.125) .452*** (.124) Year Trend×Setting Year Trend×Outdoor/nature .137 (.085) .158 (.283) .012 (.133) .168 (.172) Year Trend×Business .081 (.114) .289 (.211)−.371 (.235)−.056 (.186) Year Trend×Home (in/outside)−.118 (.086)−.435** (.191) .029 (.134)−.153 (.130) Year Trend×Social (outside)−.039 (.151) .360* (.200)−.124 (.283) .015 (.236) Year Trend×Multiple−.104 (.140) .270 (.273) .209 (.242) .196 (.170) Year Trend×Other−.073 (.120)−.245 (.264) .152 (.150)−.014 (.274) Year Trend×Role of celebrities−.095** (.040)−.151 (.112)−.013 (.061) .068 (.068) Year Trend×Appeal Year Trend×Sexual .018 (.176)−.128 (.292)−.278* (.159)−.301 (.194) Year Trend×Humor−.114 (.077) .274 (.177) .009 (.121) .145 (.139) Year Trend×Emotional vs. Rational−.001 (.078)−.181 (.146)−.005 (.130)−.126 (.097) Year Trend×Role of CSR .115 (.109) .002 (.181) .189 (.157) .075 (.178) Year Trend×Role of music .069 (.064)−.467*** (.172) .082 (.110) .021 (.116) Note:Standard errors in parentheses.aMultinomial logit regression with heteroscedasticity-robust standard errors. *p<.10. **p<.05. ***p<.01. WINTER 2019 VOLUME 53, NUMBER 4 1555 FIGURE 3 Relation Between Gender of Principal Characters and Ad Features Setting l a e p p A l a u x e S t c u d o r P f o e p y T r o m u H s e i t i r b e l e C 2.5% 1.5% 2.4% 3.0% 8.6% 8.7%37.3% 51.5% 51.8% 45.5% 43.1% 43.5%60.2% 47.0% 45.9% 51.5% 48.3% 47.8% 0.0% 20.0% 40.0% 60.0% 80. 0% Outdoor/na ture BusinessHome (in/ outside)Socia l (outside )multiple other Perc. of ads Setting Ge nder of Principal Actor only female both only male 4.9% 2.5%43.2% 47.2%51.9% 50.3% 0.0% 20.0% 40.0% 60.0% Non-technicalTechnical Perc. of ads Type of Product Ge nder of Principal Actor only female both only male 3.4% 13.0%43.1% 73.9% 53.5% 13.0% 0.0% 20.0% 40.0% 60.0% 80. 0% not sexualsexual Perc. of ads Appeal Ge nder of Principal Actor only female both only male 3.0% 0.0% 6.1%44.3% 72.7% 41.5%52.7% 27.3% 52.4% 0.0% 20.0% 40.0% 60.0% 80. 0% no roleminor rolemajor role Perc. of ads Appearance of Celebrities Ge nder of Principal Actor only female both only male 6.7% 1.8%43.8% 45.6%49.4% 52.6% 0.0% 20.0% 40.0% 60.0% no humor humor Perc. of ads Appeal Ge nder of Principal Actor only female both only male p<.05; coeff. Home =.591,p<.10; see Table 4). No significant time trends are apparent for any of the setting related dummy variables (Table 5). Mixed-gender ads and male-only ads are not more or less likely to fea- ture celebrities (coeff. Celebrity =−.122,p>.10; see Table 4). Interestingly, 1556 THE JOURNAL OF CONSUMER AFFAIRS FIGURE 3 Continued e g a s s e m R S C l a n o i t a R . s v l a n o i t o m E c i s u M 4.2% 0.0% 0.0%43.9% 66.7% 58.3% 51.9% 33.3% 41.7% 0.0% 20.0% 40.0% 60.0% 80. 0% nominormajor Perc. of ads CS R message Ge nder of Principal Actor only female both only male 2.2% 1.4% 14.3%43.5% 48.1% 33.8%54.3% 50.5% 51.9% 0.0% 20.0% 40.0% 60.0% no roleminor rolemajor role Perc. of ads Music Ge nder of Principal Actor only female both only male the time trend is negative (coeff. Year Trend×Celebrity =−.095,p<.05; see Table 5) which implies that in the past mixed-gender ads more frequently featured celebrities than male-only ads, further suggestive of underrepre- sentation of female celebrities. A more detailed analysis reveals that 47 ads (11.3%) feature female celebrities as principal characters whereas 123 ads (30.3%) feature male celebrities as a principal character. The time trend for male celebrities is more positive (r=.50) than for female celebrities (r=.31), indicative of the gender gap increasing for celebrities over the sample period. For sexual appeals, the differences are highly pronounced (Figure 3). Ads depicting both female and male models in major roles more frequently use sexual appeals (73.9% of all that use a sex appeal) than male-only ads (13.0%). These differences are significant (coeff. Sexual =2.212,p<.01; see Table 4) and stable over time (coeff. Year Trend×Sexual =.018,p>.10; see Table 5). For humorous appeals, significant group differences and time trends are not apparent (see Tables 4 and 5). Interestingly, ads featuring WINTER 2019 VOLUME 53, NUMBER 4 1557 both genders are more likely to have emotional rather than rational appeals compared to male-only ads (coeff. Emot. vs. Rat. =.579,p<.05; see Table 4). Concerning the two remaining content features, CSR messages and music, significant group differences and time trends cannot be observed (see Tables 3 and 4). Ethnic Groups The number of differences between ads featuring only minorities (N minority-only =32) and mixed ethnic groups (N mixed =48) compared to ads featuring only Caucasian models in major roles (N Caucasian-only =326) is lower compared to gender differences (Tables 3 and 4 as well as Figure 4). We find that minorities are rather equally represented in ads compared to Caucasians with regard to setting, humor, and emotional vs. rational messages. Some differences and time trends are however noteworthy. Although there are no differences concerning technical products across the whole sample period (coeff. Technical =.047,p>.10; see Table 4), the time trend for “minority-only” ads is positive, and those ads more likely feature technical products in more recent years than in earlier years (coeff. Year Trend×Technical =.452,p<.01; see Table 5). Ads with minorities as principal characters are more likely to feature celebri- ties (coeff. Celebrity =.570,p<.01 compared to mixed ethnic group ads; coeff. Celebrity =.426,p<.05 for minority-only ads; see Table 4). Further, music plays a more prominent role in ads featuring only minorities (coeff. Music =.764,p<.05; see Table 4) while CSR messages are nonex- istent in those ads (Figure 4). Finally, sexual appeals are less common in ads with primary characters from minority and Caucasian groups (coeff. Sexual =−2.031,p<.10; see Table 4), and the time trend is negative (coeff. Year Trend×Sexual =−.278,p<.10; see Table 5). Summary Females and minorities are represented in Super Bowl ads differently from males and Caucasians in several ways. Female principal characters are more likely to appear in ads with business and home settings, sexual appeals, and emotional messages. The pattern is a bit different for ads featuring only women since these ads generally do not feature CSR messages and are more likely to use music as a major element, and, like mixed gender ads, sexual appeals are more prominent. Further, the time trend shows “female-only” ads in home settings are less frequent in recent years. Minority groups are more likely to appear in ads which feature celebrities (who frequently belong to minority groups themselves), music, 1558 THE JOURNAL OF CONSUMER AFFAIRS FIGURE 4 Relation Between Ethnic Group of Principal Characters and Ad Features Setting l a e p p A l a u x e S t c u d o r P f o e p y T r o m u H s e i t i r b e l e C 7.6% 13.0% 12.3% 4.3%80.2% 82.6% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% not sexualsexual Perc. of ads Appeal Et h n i c Gr o up of Pr i nc ipal Actor only minority both only caucasia n 5.9% 4.5% 11.6% 8.4% 9.1% 17.7%85.7% 86.4% 70.7% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% no roleminor rolemajor role Perc. of ads Appearance of Celebrities Et h n i c Gr o up of Pr i nc ipal Actor only minority both only caucasia n 7.3% 8.3%14.0% 10.1%78.7% 81.6% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% no humor humor Perc. of ads Appeal Et h n i c Gr o up of Pr i nc ipal Actor only minority both only caucasia n and, in more recent years, technical products, but less likely to appear in ads featuring CSR messages. Finally, minority groups are now less likely than before to be integrated with Caucasians in ads which feature sexual appeals. This latter result should be interpreted with caution, however, as the latter end of the 10-year period analyzed featured very few sexual appeals. WINTER 2019 VOLUME 53, NUMBER 4 1559 FIGURE 4 Continued e g a s s e m R S C l a n o i t a R . s v l a n o i t o m E c i s u M 8.3% 0.0% 0.0%11.9% 11.1% 8.3%79.7% 88.9% 91.7% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% nominormajor Perc. of ads CS R message Et h n i c Gr o up of Pr i nc ipal Actor only minority both only caucasia n 4.3% 7.1% 13.0%13.0% 10.6% 15.6%82.6% 82.3% 71.4% 0.0% 20.0% 40.0% 60.0% 80. 0% 100.0% no roleminor rolemajor role Perc. of ads Music Et h n i c Gr o up of Pr i nc ipal Actor only minority both only caucasia n DISCUSSION There is little doubt that, as a collective, Super Bowl advertisers are making efforts to include minority groups in their ads, more than that had been the case in past decades. All minority groups are represented in numbers higher than proportionality to the overall population. With the exception of Latinos, most groups exceed proportionality by a wide margin. Compared to some other previous studies (e.g., King 2012; Mastro and Stern 2003; Taylor, Grau and Bang 2005), it also appears to be the case that less stereotyping by product category is taking place. At the same time, it is clear that Latinos remain the least represented of the major minority groups. It may be that the presence of Spanish language media leads some advertisers to concentrate on targeting Latinos through those media outlets only. However, this issue warrants additional research. At a societal level, the apparent improvement in all groups level of representation is positive, as all groups are visible, and in many cases in substantial numbers. 1560 THE JOURNAL OF CONSUMER AFFAIRS While the overall representation of minority groups is high, the data on principal character frequency and role prominence paints a very different picture. The fact that none of the minority groups are portrayed as princi- pal characters at a level that meets proportionality is problematic. While the situation is different for African Americans, for whom 11.6% of ads feature them as a principal character, other minorities only rarely appear as principal characters. Perhaps even more important, our data suggests that most minority portrayals are in minor and background roles. What is potentially problematic about the relative lack of prominence of minority groups in advertising is that, according to cultivation theory, background- ing can cultivate the idea that minorities do not play as prominent of a role in society. Moreover, according to expectancy theory, failure to show minorities as frequently as talented professionals or heads of the household and business may reduce expectancies for minority group members to play those roles. Our results also show striking evidence that females are underrepre- sented in Super Bowl ads both as principal characters, and as celebrity endorsers. Over the 10-year period analyzed, males appeared as principal characters more than twice as often (79.1% of ads) as females (38.2% of ads). In terms of celebrity endorsers (either major or minor role), male celebrities appeared 123 times, 261% more often than the appearances of female celebrities. While NFL football may have been traditionally viewed as an event skewed male in terms of viewership, this is no longer true of the Super Bowl. As a result, viewership does not explain the difference in representation among groups, nor does the mix of product types advertised. In terms of representation of product categories, minority groups are represented now more evenly compared to findings in past studies. It is striking that ads which feature minorities are for nontechnical products (60.0%; Asian Americans 75.0%), much more often than technical prod- ucts. This is a far cry from the 90% technical product representation for Asian Americans found by Taylor and Lee (1994). Additionally, women are not overrepresented in ads for nontechnical products. In fact, technical product advertisers have even made an increasing effort to include women, perhaps because they are often in the target market. Women are overrep- resented in ads using music as a major element and ads with emotional appeals. This finding may simply be a byproduct of these types of ads being more effective with female audiences. With respect to settings, it appears that minority groups are better dispersed. If only minorities appear in ads as the principal character, they appear, on average, in the same settings as Caucasians. The same holds for ads in which minorities and Caucasians appear together in major roles. WINTER 2019 VOLUME 53, NUMBER 4 1561 While some overrepresentation of Asian Americans in business settings is present, the differences are not at levels reported in past studies. Yet, one finding is noteworthy: It seems that ads that feature minorities and Caucasians simultaneously are less likely to feature sexual appeals, and this gap has become even become stronger across the sample period. Role congruity theory (Eagly and Karau 2002) suggests that alignment between a group’s characteristics and society’s requirements of the group’s social role results in congruity. A group’s characteristics are valued by society when they are perceived to facilitate success in these social roles. On the contrary, prejudice can result when there is incongruity for the group’s characteristics failing to meet society’s expectations of the group’s social role (Eagly and Karau 2002). While our analysis of Super Bowl advertising over the 10-year period from 2008 to 2017 suggests an absence of many blatant stereotypes of the past and substantial progress in the representation of females, the lack of appearance of female and minority models in more prominent roles, with the exception of minority celebrities, is a remaining issue. The relatively lower frequency of portrayals of minorities as a principal character could cultivate problematic perception based on absence from these roles and reduce expectancies to take on leadership roles in society. That minorities are seldom shown in ads using a CSR appeal may also build expectancies that caring about these issues is not expected and cultivate a perception that only Caucasians are interested in CSR. In terms of our theoretical perspectives, the results show a better overall representation of women and minorities and a trend toward more balanced portrayals in terms of product categories and settings than shown in past studies. However, the relative lack of portrayals women and most minorities may build expectancies that it is difficult or unusual to rise to leadership or other prominent positions in our society. Moreover, the lack of prominence of female and minority characters in major roles can cultivate perceptions that this is normal at a societal level. At a managerial level, marketers can effectively reach their targeted audience when advertising portrayals represent groups in a positive manner by targeting the group’s primary beliefs (La Ferle and Lee 2005). In this light, ads can be an efficient communication tool to categorize and transmit a personally relevant message to the targeted consumer (Windels 2016) and can enable the targeted consumer to arrive at a purchasing decision quickly (Courtney and Whipple 1983; La Ferle and Lee 2005; Rossiter and Chan 1998). Similarly, advertising portrayals that accurately reflect cultures or subcultures can enlighten consumers who have not been exposed to these cultures themselves (La Ferle and Lee 2005). Thus, it is 1562 THE JOURNAL OF CONSUMER AFFAIRS important for advertising to portray individuals in a manner that promotes consumer wellbeing. Recent research (e.g., Åkestam 2018; Lord, Putrevu, and Collins 2019; Pickett and Brison 2019) finds that ethnicity and gender have a profound influence on reactions to ads and reactions to stereotypes portrayals (De Meulenaer et al. 2018; Mayer, Kumar, and Yoon 2019). Considering viewership is of importance in the context of Super Bowl advertising. Women are one of the most important target groups for future viewership growth. In 2017, 49% of the 108 million viewers were female. More importantly, female viewers also focused more on the ads compared to men (Salkowitz 2018). Thus, not only do women represent half of the current viewership, but they are also more prone to pay attention to, recognize, and recall advertising during the Super Bowl. Thus, providing advertising content that targets and/or resonates with this group could boost the potential of the Super Bowl for advertisers considerably. Similarly, minority groups are growing as a proportion of the population and delivering portrayals these individuals can relate to is important. In terms of policy implications, it would behoove the US advertising industry to be aware of the role prominence issue and any stereotypes that still exist. Considering regulations recently passed by the UK Advertising Standards Authority that explicitly ban advertising that features stereotypes of women (Safronova 2019), individual advertisers and the industry as a whole would be wise to be sensitive to the findings of this study and others like it. While legislation as strong as the UK standards would appear highly unlikely considering the First Amendment and public attitudes in the United States, avoiding criticism while simultaneously targeting large consumer groups effectively is advisable. A topic worth further exploration is whether the relatively small number of top decision-makers in the advertising industry who are female and/or minorities (see Bendick and Egan 2009; Maclaran, Lorna Stevens, and Catterall 1997) has an impact on portrayals. While the situation has improved since the “Mad Men” era, it has been argued that obstacles remain and additional change is needed to aid in more effective portrayals of women and minorities (Boulton 2013; Maclaran, Lorna Stevens, and Catterall 1997). CONCLUSION This research suggests that Super Bowl advertising portrayals of minorities are quite inclusive in terms of overall minority representation. However, issues persist in terms of role prominence in the ads, where minorities are clearly underrepresented. While blatant stereotypes no WINTER 2019 VOLUME 53, NUMBER 4 1563 longer appear to be present, it appears that some degree of backgrounding of minority characters is present. Meanwhile, females are still underrepre- sented as principal characters, although the number of ads featuring both female and male models in major roles has increased over the sample period. Yet, females still appear more frequently in ads with sexual appeal but less frequently in home settings. We hope this study encourages more research in this important domain and extends our analysis to other contexts and settings. APPENDIX 1 Codebook for Advertising Content Variables 1 Advertising number _______ Coder number _______ Year _______ Commercial length _______ 2 Primary type of product/Service advertised 0=Nontechnical product 1=Technical product 3 Commercial characters celebrities 0=No presence 1=Any appearance (minor role) 2=Principal character (major role) 4 Commercial approach—Rational vs. Emotional appeal 0=More rational 1=Balanced rational and emotional 2=More emotional 5 Use of sexual appeal as a major element 0=No 1=Ye s 6 Presence of humor as a major element 0=No 1=Ye s 7 Presence of music 0=No 1=Minor element 2=Major element 8 Principal character(s) female 0=No 1=Ye s 9 Principal character(s) male 0=No 1=Ye s 10 Presence of male models as primary character: Ethnic group of the primary male model0=No male model 1=Caucasian 2=African American 3=Latino 4=Asian 5=Middle Eastern 6=Mixed 7=South Asian 1564 THE JOURNAL OF CONSUMER AFFAIRS 11 Presence of female models as primary character: Ethnic group of the primary female model0=No female model 1=Caucasian 2=African American 3=Latino 4=Asian 5=Middle Eastern 6=Mixed 7=South Asian 12 Perceived importance of minority characters 0=No role 1=Background role 2=Minor role 3=Major role (a) Perceived importance of Asian American models (b) Perceived importance of African American models (c) Perceived importance of Latino American models (d) Perceived importance of Middle Eastern models (e) Perceived importance of Mixed ethnicity models (f) Perceived importance of South Asian models_______ _______ _______ _______ _______ _______ 13 Setting 1=Business setting 2=Home setting (indoor or outdoor) 3=Outdoor/natural scenery 4=Social setting outside the home 5=Multiple settings 6=Other 14 Corporate social responsibility as message 0=Not mentioned 1=Mentioned 2=Main message APPENDIX 2 Operational Definitions of Coded Variables Primary Type of Product/Service Advertised—Indicate whether the product falls into the technical or nontechnical category based on the following list (0=nontechnical; 1=technical). Nontechnical products: Food and beverages, alcoholic beverages, tobacco products, OTC and prescription drugs, household, lawn and garden supplies, cosmetics and personal care products, diet and exercise product, clothing, shoes and apparel, furniture, entertainment supplies, sporting goods, pets, pet food and pet supplies, publications, movies, retailers, transportation services, other nontechnical products/services. Technical products: Automobile, automobile-related, personal trans- portation vehicles, electronic appliances and products, home comput- ers and, office equipment and supplies, banking and financial services, telecommunications services, other technical products/services. WINTER 2019 VOLUME 53, NUMBER 4 1565 Commercial Characters—Celebrities aPrincipal character celebrity: Code “1” if the character(s) delivering the main portion of the message on camera is well known either by name or face. Celebrities may be athletes, movie or entertainment stars or well-known corporate figures (but not simply the identified head of a corporation). If there is no celebrity in the ad or a celebrity appears but is not the principal character code “0”. bPresence of male celebrity as principal character: Code “1” if a male celebrity is the principal character as defined in (a) above. Otherwise code “0”. cPresence of female celebrity as principal character: Code “1” if a male celebrity is the principal character as defined in (a) above. Otherwise code “0”. dPresence of any celebrity: Code “1” if any celebrity well known by name or face from sports, entertainment, or industry appears in the ad (identified heads of corporations should not be counted). Code “0” if there is no celebrity in the ad. ePresence of any male celebrity: Code “1” if any male celebrity well known by name or face from sports, entertainment, or industry appears in the ad (identified heads of corporations should not be counted). Code “0” if there is no male celebrity in the ad. fPresence of any female celebrity: Code “1” if any female celebrity well known by name or face from sports, entertainment, or industry appears in the ad (identified heads of corporations should not be counted). Code “0” if there is no female celebrity in the ad. Commercial Approach—Rational vs. Emotional Appeal Rational/emotional appeal: A straightforward presentation of the prod- uct’s attributes and claims is a rational appeal. An emotional appeal does not appeal to reason but to feelings. Is the commercial primarily making a rational or an emotional appeal to the audience? (Codes: 1=more rational; 2:=balance of rational and emotional appeals; 3=more emotional). Use of a Sexual Appeal as a Major Element—Code “1” if the main focus of the commercial is on sexual cues. “0” otherwise”. Use of Humor as a Major Element—Code “1” if the commercial takes a humorous approach. “0” otherwise. Presence of Music—Code “0” if there is no music; “1” if music is present in the ad in any form; “2” if music is a major element. Principal Character(s)Male—The character(s) carrying the major on-camera role of delivering the commercial message is male. Incidental, 1566 THE JOURNAL OF CONSUMER AFFAIRS background on-camera appearance is not applicable. Code “1” if a princi- pal male character appears in the ad. Otherwise, code “0.” Principal Character(s)Fe m a l e—The character(s) carrying the major on-camera role of delivering the commercial message is female. Incidental, background on-camera appearance is not applicable. Code “1” if a princi- pal female character appears in the ad. Otherwise, code “0.” Principal Character(s)Child or Infant—The character(s) carrying the major on-camera role of delivering the commercial message is a child or infant. Incidental, background on-camera appearance is not applicable. Code “1” if a principal female character appears in the ad. Otherwise, code “0.” General Presence of Minority Models a Presence of any minority model—Code “1” if any minority model is shown in the ad. Otherwise code “0.” b Presence of male minority model—Code “1” if any minority male model is shown in the ad. Otherwise code “0.” c Presence of female minority model—Code “1” if any minority female model is shown in the ad. Otherwise code “0.” d Presence of African American model—Code “1” if any African American model is shown in the ad. Otherwise code “0.” e Presence of Latino American model—Code “1” if any Latino Amer- ican model is shown in the ad. Please include all information avail- able in the ad including name and context in making this determina- tion. Otherwise code “0.” f Presence of Asian American model—Code “1” if any Asian Ameri- can model is shown in the ad. “Asian American” refers to people of East Asian descent. Otherwise code “0.” g Presence of Middle Eastern model—Code “1” if any Middle Eastern model is shown in the ad. Otherwise code “0.” h Presence of South Asian model—South Asia is defined as India, Pakistan, Bangladesh, Sri Lanka, Nepal and Bhutan (which consti- tute the Indian subcontinent) along with Afghanistan and Maldives. Code “1” if any South Asian model is shown in the ad. Otherwise code “0.” Race of the Primary Male Model—Indicate the race of the primary male model in the ad Code: 1=Caucasian 2=Black 3=Latino 4=Asian WINTER 2019 VOLUME 53, NUMBER 4 1567 5=Middle Eastern 6=South Asian 0=No male model Race of the Primary Female Model—Indicate the race of the primary female model in the ad Code: 1=Caucasian 2=Black 3=Latino 4=Asian 5=Middle Eastern 6=South Asian 0=No female model Perceived Importance of Minority Characters—For each item in this section, please apply the following definitions: Major role: A model who is very important to the commercial theme or layout, shown in the foreground and/or shown holding the product and/or appears to be speaking. Minor role: A character who is of average importance to the advertising theme or layout, does not appear to speak or handle the product. (Generally, these characters are not spotlighted but they are not difficult to find in the ad while casually looking at it.) Background role: A character who is difficult to find in an ad, and not important to the commercial theme or layout. a Perceived importance of Black model—Choose the most prominent Black model in the advertisement. Then, code “1” if this model plays a major role, “2” for a minor role, and “3” for a background role. Code “0” if no Black model appears in the ad. b Perceived importance of Latino model—Choose the most prominent Latino model in the advertisement. Then, code “1” if this model plays a major role, “2” for a minor role, and “3” for a background role. Code “0” if no Latino model appears in the ad. c Perceived importance of Asian (East) model—Choose the most prominent Asian model in the advertisement. Then, code “1” if this model plays a major role, “2” for a minor role, and “3” for a background role. Code “0” if no Asian model appears in the ad. d Perceived importance of Middle Eastern model—Choose the most prominent Middle Eastern model in the advertisement. Then, code “1” if this model plays a major role, “2” for a minor role, and “3” for a background role. Code “0” if no Asian model appears in the ad. 1568 THE JOURNAL OF CONSUMER AFFAIRS e Perceived importance of South Asian model—Choose the most prominent Other Minority model in the advertisement. Then, code “1” if this model plays a major role, “2” for a minor role, and “3” for a background role. Code “0” if no Other Minority model appears in the ad. Setting—Choose the setting of the advertisement based on the following list (choose only one): (1)Business setting: Factories, sales or office rooms and retail settings in which consumers are depicted inside stores. (2)Home,indoor or outdoor: Recognizable as a residence, room, or rooms, garage, yard, home or apartment driveway or parking space. (3)Outdoors/natural scenery: Includes forests, rivers, ocean, fields or sky as well as streets, public roads, sidewalks, or pathways. Does not include outdoor settings at individuals’ homes or outdoor social settings. (4)Social setting outside home: Includes public places, auditoriums, restaurants, movie theaters, where people meet and congregate for social purposes. (5)Multiple settings: Commercials in which there is no single domi- nant setting. Choose this option only if there is not primary setting depicted in the ad (i.e., if 75% of the ad was in a place of business and 25% was outdoors), the commercial should be coded as having a business setting. (6)Other: Any type of setting not listed above. If this option is chosen, please indicate the setting in the space provided. Corporate Social Responsibility—Code “0” if there is no mention of corporate social responsibility, “1” if the primary message in the ad focused on some aspect of corporate social responsibility and “2” if the CSR aspect is the main message in the ad. REFERENCES Åkestam, Nina. 2018. Caring for Her: The Influence of Presumed Influence on Female Consumers’ Attitudes Towards Advertising Featuring Gender-Stereotyped Portrayals.International Journal of Advertising, 37 (6): 871–892. Ashmore, Richard D. and Frances K. Del Boca. 1981. Conceptual Approaches to Stereotypes and Stereotyping. InCognitive Processes in Stereotyping and Intergroup Behavior, edited by D. Hamilton, (1–36). Hillsdale, NJ: Erlbaum. Bailey, Ainsworth Anthony. 2006. A Year in the Life of the African-American Male in Advertising: A Content Analysis.Journal of Advertising, 35 (1): 83–104. Bandura, Albert. 2002. Social Cognitive Theory in Cultural Context.Applied Psychology,51(2): 269–290. WINTER 2019 VOLUME 53, NUMBER 4 1569 Bang, Hae-Kyong and Bonnie B. Reece. 2003. Minorities in Children’s Television Advertis- ing Commercials: New, Improved, and Stereotyped.Journal of Consumer Affairs,37(1): 42–67. Belkaoui, Ahmed and Janice M. Belkaoui. 1976. A Comparative Analysis of the Roles Portrayed by Women in Print Advertisements: 1958, 1970, and 1972.Journal of Marketing Research,13(2): 168–172. Bendick, Marc Jr. and Mary Lou Egan. 2009.Research Perspectives on Race and Employment in the Advertising Industry. Washington, DC: Bendick and Egan Economic Consultants, Inc. Boulton, Christopher. 2013. The Ghosts of Mad Men: Race and Gender Inequality Inside Amer- ican Advertising Agencies. InThe Routledge Companion to Advertising and Promotional Culture, edited by Matthew P. McAllister and Emily West, (252–266). New York, NY: Routledge. Branchik, Blaine and Judy David. 2018. From Servants to Spokesmen: Black Male Advertising Models and Changing US Culture Post World War II.Journal of Historical Research in Marketing,10(4): 451–473. Briggs, Ellen. 2008. Services’ Influence on Minority Portrayals in Magazine Advertising.Journal of Services Marketing, 24 (3): 80–93. CNBC. 2019. This is How Much it Cost to Air a Commercial During the 2019 Super Bowl; February 4. https://www.cnbc.com/2019/01/30/how- much- it- costs- to- air- a- commercial- during- super- bowl- liii.html. Colby, Sandra L. and Jennifer M. Ortman. 2017. Projections of the Size and Composition of the US Population: 2014 to 2060: Population Estimates and Projections. Current Population Reports; March 2015. Courtney, Alice E. and Thomas W. Whipple. 1983.Sex Stereotyping in Advertising. New York: Free Press. Crupi, Anthony. 2018. Despite the Thriller in Minny, Super Bowl Ratings Take a Tumble. AdAge. https://adage.com/article/special- report- super- bowl/s/312265/. Davis, J.F. 2018. Selling Whiteness?—A Critical Review of the Literature on Marketing and Racism. Journal of Marketing Management, 34 (1–2): 134–177. De Meulenaer, Sarah, Nathalie Dens, Patrick De Pelsmacker, and Martin Eisend. 2018. How Con- sumers’ Values Influence Responses to Male and Female Gender Role Stereotyping in Advertising. International Journal of Advertising, 37 (6): 893–913. Eagly, Alice H. and Steven J. Karau. 2002. Role Congruity Theory of Prejudice Toward Female Leaders.Psychological Review, 109 (3): 573–598. Eisend, Martin. 2010. A Meta-Analysis of Gender Roles in Advertising.Journal of the Academy of Marketing Science, 38 (4): 418–440. Esses, Victoria M., Geoffrey Haddock, and Mark P. Zanna. 1993. Values, Stereotypes, and Emotions as Determinants of Intergroup Attitudes. InAffect, Cognition and Stereotyping, edited by Diane M. Mackie and David L. Hamilton (137–166). San Diego, CA: Academic Press. Fowler, Kendra and Veronica Thomas. 2015. A Content Analysis of Male Roles in Television Advertising: Do Traditional Roles Still Hold?Journal of Marketing Communications,21(5): 356–371. Furnham, Adrian and Alixe Lay. 2017. The Universality of the Portrayal of Gender in Television Advertisements: A Review of the Studies This Century.Psychology of Popular Media Culture, 8 (2): 109–124. Ganahl, Dennis J., Thomas J. Prinsen, and Sara Baker Netzley. 2003. A Content Analysis of Prime Time Commercials: A Contextual Framework of Gender Representation.Sex Roles, 49 (9–10): 545–551. Gerbner, George, Larry Gross, Michael Morgan, and Nancy Signorielli. 1980. The “Mainstreaming” of America: Violence Profile No. 11.Journal of Communication, 30 (3): 10–29. Graham, Morris, A. 1983. Acculturative Stress Among Polynesian, Asian and American Students on the Brigham Young University—Hawaii Campus.International Journal of Intercultural Relations, 7 (1): 79–103. 1570 THE JOURNAL OF CONSUMER AFFAIRS Grau, Stacy Landreth and Yorgos C. Zotos. 2016. Gender Stereotypes in Advertising: A Review of Current Research.International Journal of Advertising, 35 (5): 761–770. Greenberg, Bradley S., Dana Mastro, and Jeffrey E. Brand. 2002. Minorities and the Mass Media: Television into the 21st Century. InMedia Effects: Advances in Theory and Research, edited by J. Bryant and D. Zillmann (333–351). Mahwah, NJ: Lawrence Erlbaum Associates Publishers. Hatzithomas, Leonidas, Christina Boutsouki, and Paschalina Ziamou. 2016. A Longitudinal Analysis of the Changing Roles of Gender in Advertising: A Content Analysis of Super Bowl Commercials. International Journal of Advertising, 35 (5): 888–906. Hughes, Marie Adele and Dennis E. Garrett. 1990. Intercoder Reliability Estimation Approaches in Marketing: A Generalizability Theory Framework for Quantitative Data.Journal of Marketing Research, 27 (2): 185–195. Jhally, Sut and Jean Kilbourne. 2010.Killing USoftly 4: Advertising’s Image of Women. Northampton, MA: Media Education Foundation. Jussim, Lee. 1990. Social Reality and Social Problems: The Role of Expectancies.Journal of Social Issues, 46 (2): 9–34. King, B. 2012. Images of Gender and Race in Super Bowl Advertising.Media Report to Women,40 (1): 6–11. Kolbe, Richard H. and Melissa S. Burnett. 1991. Content-Analysis Research: An Examination of Applications with Directives for Improving Research Reliability and Objectivity.Journal of Consumer Research, 18 (2): 243–250. Krippendorf, Klaus. 2013.Content Analysis. An Introduction to its Methodology. 3rd edition. California, CA: Sage. La Ferle, Carrie and Wei-Na Lee. 2005. Can English Language Media Connect with Ethnic Audiences? Ethnic Minorities’ Media Use and Representation Perceptions.Journal of Advertising Research, 45 (1): 140–153. Landis, J.R. and G.G. Koch. 1977. The Measurement of Observer Agreement for Categorical Data. Biometrics, 33 (1): 159–174. Lord, Kenneth R., Sanjay Putrevu, and Alice F. Collins. 2019. Ethnic Influences on Attractiveness and Trustworthiness Perceptions of Celebrity Endorsers.International Journal of Advertising,38(3): 489–505. https://doi.org/10.1080/02650487.2018.1548196. Maclaran, Pauline, L. Lorna Stevens, and Miriam Catterall. 1997. The “Glasshouse Effect”: Women in Marketing Management.Marketing Intelligence & Planning, 15 (7): 309–317. Mager, John and James G. Helgeson. 2011. Fifty Years of Advertising Images: Some Chang- ing Perspectives on Role Portrayals Along with Enduring Consistencies.Sex Roles, 64 (3–4): 238–252. Mastro, Dana E. and Susannah R. Stern. 2003. Representations of Race in Television Commercials: A Content Analysis of Prime-Time Advertising.Journal of Broadcasting & Electronic Media,47 (4): 638–647. Mastro, D. and Riva Tukachinsky. 2011. Exemplar Versus Prototype-Based Processing of Media Content and the Influence on Racial/Ethnic Evaluations.Journal of Communication,61(5): 916–937. Mastro, Dana E., Elizabeth Behm-Morawitz, and Michelle Ortiz. 2007. The Cultivation of Social Perceptions of Latinos: A Mental Models Approach.Media Psychology, 9 (2): 347–365. Matthes, Jörg, Michael Prieler, and Karoline Adam. 2016. Gender-Role Portrayals in Television Advertising Across the Globe.Sex Roles, 75 (7–8): 314–327. Mayer, James M., Piyush Kumar, and Hye Jin Yoon. 2019. Does Sexual Humor Work on Mars, but Not on Venus? An Exploration of Consumer Acceptance of Sexually Humorous Advertising. International Journal of Advertising: 1–25. https://doi.org/10.1080/02650487.2019.1629226. Morgan, Michael and James Shanahan. 2010. The State of Cultivation.Journal of Broadcasting & Electronic Media, 54 (2): 337–355. Morrison, Margaret. 2014. Stereotypes Are a Necessary and Appropriate Strategy for Advertising. In Advertising and Society: An Introduction, edited by Carol Pardun. West Sussex, UK: Wiley. WINTER 2019 VOLUME 53, NUMBER 4 1571 Nielsen. 2018. Super Bowl LII Draws 103.4 Million TV Viewers, 170.7 Million Social Media Interactions. Nielsen online. https://www.nielsen.com/us/en/insights/news/2018/super- bowl- lii- draws- 103- 4- million- tv- viewers- 170- 7- million- social- media- interactions.html. Peterson, Robin T. 2007. Consumer Magazine Advertisement Portrayal of Models by Pace in the U.S.: An Assessment.Journal of Marketing Communications, 13 (3): 199–211. Phua, Joe. 2014. The Influence of Asian American Spokesmodels in Technology-Related Advertising: An Experiment.Journal of Consumer Affairs, 25 (4): 399–414. Picchi, Aimee. 2018. Super Bowl LII: Ratings for Football’s Biggest Game Lowest Since 2009.CBS News. https://www.cbsnews.com/news/super- bowl- lii- tv- ratings/. Pickett, Andrew C. and Natasha T. Brison. 2019. Lose Like a Man: Body Image and Celebrity Endorse- ment Effects of Weight Loss Product Purchase Intentions.International Journal of Advertising: 1–18. https://doi.org/10.1080/02650487.2019.1586208. Rossiter, John R. and Alvin M. Chan. 1998. Ethnicity in Business and Consumer Behavior.Journal of Business Research, 42 (2): 127–134. Rubie-Davies, Christine, Sabrina Liu, and Katie Lee. 2012. Watching Each Other: Portrayals of Gender and Ethnicity in Television Advertisements.The Journal of Social Psychology, 153(2): 175-195. Rubie-Davies, Christine M., Sabrina Liu, and Kai-Chi Katie Lee. 2013. Watching Each Other: Portrayals of Gender Equality and Ethnicity in Television Advertisements.Journal of Personality and Social Psychology, 153 (2): 175–195. Safronova, Valeriya. 2019. Gender Stereotypes Banned in British Advertising.New York Times, June 14. https://www.nytimes.com/2019/06/14/style/uk- gender- stereotype- ads- ban.html. (Accessed on July 5, 2019). Salkowitz, Rob. 2018. Data Shows Women Paid More Attention to Super Bowl Than Men. Forbes .com. https://www.forbes.com/sites/robsalkowitz/2018/02/05/data- shows- women- paid- more- attention- to- super- bowl- lii- than- men/#da2fd3e5c4a8. (Accessed on March 14, 2019). Sivulka, Juliann. 2009.Ad Women: How They Impact What We Need, Want, and Buy. Amherst, NY: Prometheus Books. Smith, Stacy L., Marc Choueiti, and Jessica Stern. 2011.Occupational Aspirations: What are G-Rated Films Teaching Children About the World of Work?Los Angeles: Geena Davis Institute for Gender and Media. Stevenson, Thomas H. and Linda E. Swayne. 2011. Is the Changing Status of African Americans in the B2B Buying Center Reflected in Trade Journal Advertising?Journal of Advertising,40(4): 101–122. Taylor, Charles R. and Ju Yung Lee. 1994. Not in Vogue: Portrayals of Asian Americans in Magazine Advertising.Journal of Public Policy & Marketing, 13 (2): 239–245. Taylor, Charles R. and Barbara B. Stern. 1997. Asian-Americans: Television Advertising and the “Model Minority” Stereotype.Journal of Advertising, 26 (2): 47–61. Taylor, Charles R., Stacy Landreth, and Hae-Kyong Bang. 2005. Asian Americans in Maga- zine Advertising: Portrayals of the Model Minority.Journal of Macromarketing,25(2): 163–174. Taylor, Charles R., Stacy Grau, and Hae-Kyong Bang. 2005. Asian Americans in Magazine Advertis- ing: Portrayals of the “Model Minority.”Journal of Macromarketing, 25 (2): 163–174. Tukachinsky, Riva, Dana Mastro, and Moran Yarchi. 2015. Documenting Portrayals of Race/Ethnicity on Primetime Television over a 20-Year Span and Their Association with National-Level Racial/Ethnic Attitudes.Journal of Social Issues, 71 (1): 17–38. U.S. Department of Labor. 2010. Women in the Labor Force in 2010. https://www.dol.gov/wb/ factsheets/Qf- laborforce- 10.htm. Uray, Nimet and Sebnem Burnaz. 2003. An Analysis of the Portrayal of Gender Roles in Turkish Television Advertisements.Sex Roles, 48 (1–2): 77–87. Vinacke, W.Edgar. 1957. Stereotypes as Social Concepts.Journal of Social Psychology,46(2): 229–243. 1572 THE JOURNAL OF CONSUMER AFFAIRS Wilkes, Robert E. and Humberto Valencia. 1989. Hispanics and Blacks in Television Commercials. Journal of Advertising, 18 (1): 19–25. Windels, Kasey. 2016. Stereotypical or Just Typical: How Do US Practitioners View the Role and Function of Gender Stereotypes in Advertisements?International Journal of Advertising,35(5): 864–887. Yelkur, Rama, Chuck Tomkovick, Ashley Hofer, and Daniel Rozumalski. 2013. Super Bowl Ad Likeability: Enduring and Emerging Predictors.Journal of Marketing Communications,19(1): 58–80. Zhang, Qin. 2010. Asian Americans Beyond the Model Minority Stereotype: The Nerdy and the Left Out.Journal of International and Intercultural Communications, 3 (1): 20–37. Copyright ofJournal ofConsumer Affairsisthe property ofWiley- Blackwell anditscontent may notbecopied oremailed tomultiple sitesorposted toalistserv without thecopyright holder's expresswrittenpermission. However,usersmayprint, download, oremail articles for individual use.