An Examination of Television Consumption by Racial and Ethnic Audiences in the U.S

An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. Implications for Multicultural Media Planning And Media Measurement

J. P. James Academic and practitioner studies have found that television consumption is highest among Salem State University American audiences of diverse races and ethnicities. The validity of ethnic-audience ratings [email protected] measurement in the past has been questionable, predicating diminished multicultural- Tyrha M. Lindsey- audience valuation. One result has been less spending in the ethnic broadcast-media Warren Baylor University landscape, which is hampered by media fragmentation. Using an analysis of Horowitz [email protected] Research data, a nationally representative dataset that measures multicultural cable viewership, the authors examined the relationship between television viewership and multiculturalism, mediated by programmatic and media-fragmentation influences and covaried by demographics influences.

INTRODUCTION Over the years, there has been controversy over “I encourage ethnic media companies, nation- the fact that television-audience measurement ally respected researchers and demographers to has not kept pace with the increasing diversity of get together and simply break the code of the the United States. Before 1999, the Nielsen Com- Target Rating Point. Properly done, this new sta- pany excluded less culturally assimilated ethnic tistical model can inform strategic media buying minorities from syndicated survey data (D’Rozario decisions, in a fashion that encourages the idea & Yang, 2015). In 2006, Nielsen was criticized for that good business thinking and multicultural surveying too few people of color with its People marketing are not mutually exclusive.” Meters to provide a statistically significant sam- —Jo Muse, The Shaman Chronicles Book One: ple size of ethnic consumers to develop accurate The 7 Senses ofMulticultural Marketing multicultural television ratings (Napoli, 2005). (2001, pp. 87–88) These audience ratings determine how a marketer’s

• Program genres and media fragmentation (through the use of technology) partially mediate the relationship of television viewership and multicultural consumers. • Specific program genres predict higher levels of media fragmentation, as demonstrated by the authors’ “multistep mediation” model, which takes into account viewers’ uses and gratifications (e.g. music programming, live sports) and the technology used for daily viewing. • The media research industry is encouraged to develop a television measurement standard that incorporates the multiculturalism of the audience, the impact of the genre, the use of technology, Submitted 30 March, 2015; and the influence of demographics. revised 31 March, 2017; • Context planning can leverage diverse audiences for multicultural media scheduling. accepted 30 June, 2017

DOI: 10.2501/JAR-2018-027 June 2018 JOURNAL OF ADVERTISING RESEARCH 1 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

advertising budget is spent to reach ethnic programming and have suggested using lines: between Whites (who had greater consumers through the planning and buy- a full-attention cost-per-thousand meas- access) and Blacks and other minority ing of media (Rincon, 2012). ure (Bearden, Headen, Klompmaker, and groups (who had less; DiMaggio, Hargit- Asian-Americans, Blacks, and Hispanics Teel, 1981). tai, Celeste, and Shafer, 2004). Ethnic con- combined represent almost $3 trillion of With the ubiquity of digital video sumers are coming to parity in Internet discretionary income and constitute nearly recorders (DVRs) and other, similar tech- access by consuming more digital media 32 percent of the American population nology, an adjusted cost-per-thousand than White audiences through mobile (Humphreys, 2015). Advertisers, however, has been suggested to place a premium technology (Nielsen Company, 2014). expend less than 10 percent of their aggre- on viewers who watch television advertis- As a result, more advertisers are placing gated marketing and advertising budgets ing versus those who skip through com- resources in digital media instead of tel- addressing these ethnic groups (Coffey, mercials (Wilbur, 2008). This phenomenon evision to reach multicultural segments. 2013; Kantar, 2013; La Ferle and Lee, 2005). of media fragmentation has changed for- For example, 60 percent of advertisers In comparison with Blacks and Hispanics, ever the practice of planning, buying, and spend more on new media efforts, such as advertisers have paid far less attention to selling media. In tandem with the media- mobile outreach, specifically to reach eth- Asian-Americans and their commensurate measurement issues, it creates profound nic audiences (Lopez, Gonzalez-Barrera, television platforms; many marketers see challenges for television properties, which and Patten, 2013; Association of National the segment as invisible (Coffey, 2013). now typically deliver less than one-third of Advertisers, 2012). Reliable television-ratings measurement the audiences they did in the past (Nelson- Language choice further magnifies the also is warranted for greater media invest- Field and Riebe, 2011; Rubinson, 2008). The fragmentation issue. Linguistic prefer- ment in the Asian-American segment. The 21st-century marketplace is a customer- ences for television programming among lack of a more representative sample size centric dynamic whereby the consumer Hispanics are split almost evenly among: to estimate audience ratings is holding has control of marketing communications down spending on Asian-American televi- instead of the one-way monologue of the • English only (33.4 percent); sion properties (Coffey, 2013). 20th century (Schultz and Schultz, 1998). • mostly English and some Spanish (31.9 Americans spend 32 minutes a day percent); Media Metrics, Fragmentation, on time-shifted television (70 percent of • mostly Spanish and some English (21.1 And Ethnic Audiences homes have a DVR), an hour per day using percent); Scholars have been working to break the the Internet on a laptop, and more than • Spanish only (12.8 percent) (Ad Age code of overall media measurement for one hour every day using a smartphone Datacenter, 2014). many years (e.g., Cannon, 1984; Cannon, (Nielsen Company, 2014). Thirty-one Leckenby, and Abernethy, 2002; Leckenby percent of video content is not consumed More than half of Asian-Americans are and Boyd, 1984; Leckenby and Ju, 1989; on television; consumers are interacting more comfortable speaking their native Leckenby and Kim, 1994; Leckenby and more with television content on the third tongue, with 24 percent preferring their Kishi, 1984; Leckenby and Rice, 1985). The or fourth screen (Albert and Jacobs, 2008; indigenous language for entertainment and target rating point (TRP) and the related Bondad-Brown, Rice, and Pearce, 2012). In reading (Kaufman-Scarborough, 2000). gross rating point (GRP), reach, and fre- 2018, advertiser spending on digital adver- quency are media metrics the accuracy of tising will overtake television and reach Ethnic-Media Consumption which has been debated in the extant lit- $103 billion, to represent 36 percent of all And Audience Valuation erature. A stream of literature states that advertising spending (Thielman, 2014). Advertisers historically have discrimi- gross rating points are biased in favor of Media fragmentation may be more pro- nated against television platforms and schedules that have maximum reach at nounced among ethnic audiences. It can other media that target African-American minimum levels of frequency, which can be argued that the proliferation of the or Hispanic audiences. Minority television introduce media-vehicle bias into media- Internet through mobile technology has executives contend that advertisers and planning choices (Dickson, 1991; Farris and become a confounding issue in the ethnic- their media agencies discriminate against Parry, 1991). Other scholars have argued media landscape. This is a legacy of the ethnic-media properties by demand- that television rates are inflated because digital divide, whereby online access was ing discounts or generally spending less many viewers are not fully attentive to the divided strongly along racial and ethnic on advertising in these platforms (Ofori,

2 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

1999). This is evident, given that African- Studies have shown, however, that any product Americans represent approximately 13 or service that is marketed properly to a minority percent of the population, yet fewer than 3 percent of advertising dollars are spent audience can be advantageous, because racial on Black media (Humphreys, 2015; Nielsen Company, 2014; Kantar Media, 2014). His- and ethnic groups tend to be more brand loyal. panics are more than 17 percent of the population but represent only 6 percent of the national advertising spend (Humphreys, media properties and content (Adweek, is in broadcast (LaFerle and Lee, 2005), 2015; Kantor Media 2014; Rincon, 2012). 2001). This suggests that some advertisers which suggests that English-language The greatest disparity between the racial and their media agencies may place less television is the best method to reach population and media spending is with the value on or forgo ethnic segmentation. multicultural consumers. The aforemen- Asian-American segment. This segment is Studies have shown, however, that any tioned is a media-planning tactic that just over 5 percent of the population, with product or service that is marketed prop- exemplifies a greater marketing strategy less than 1 percent (i.e., 0.07 percent) of erly to a minority audience can be advan- presently debated among advertisers advertising spending in Asian-American tageous, because racial and ethnic groups relative to overall ethnic marketing: the media (Coffey, 2014; Humphreys, 2015; tend to be more brand loyal (Desphande, total market approach. The total market Kantar Media, 2014). Hoyer, and Donthu, 1986). The current approach is defined as viewing ethnic Minority discounts are the unwritten total advertising spend for U.S. multi- and general-market audiences as a single policy of some advertisers, who purchase cultural marketing is approximately $10 segment and reflecting diversity through media on an ethnically targeted plat- billion (Coffey, 2014; Nielsen Company, culturally nuanced tactics (e.g., Franklin, form at a rate less than what is paid to a 2014; Kantar Media, 2014; Rincon, 2012). 2014, p. 259). A study by the Association general-market-formatted platform with Although substantial, this figure is just over of Hispanic Advertising Agencies (2013, a comparable audience size (Ofori, 1999). 7 percent of the total U.S. spend of $140 bil- p. 18) showed that 37 percent of media The incongruity is based on the cost per lion (Kantar Media, 2014), which may be agencies had implemented a total market point. Advertisers targeting their brands the result—in part—of current methods of approach relative to media-planning and to minority consumers often pay minor- multicultural-audience valuation. buying activities. ity- formatted media platforms on the basis The burgeoning growth of the ethnic- This article specifically examines the of the Hispanic cost per point or the Black minority population predicates that mar- relationship between ethnic consumers cost per point—the cost of reaching 1 per- keters reevaluate how they advertise to and television viewership, furthering a cent of the Hispanic or Black population multicultural consumers as they strive for body of previous work (Albarran and in the metro market (Ofori, 1999). When efficiencies and accountability with their Umphrey, 1993, 1994; Albert and Jacobs, a minority discount is used, media prop- advertising dollars (Henderson and Wil- 2008; Lin, 1999). Overall, Americans watch erties are not evaluated at parity. Unless liams, 2013). The current method of evalu- more than five hours of television every race or ethnicity is stipulated in the demo- ating racial and ethnic television audiences day. African-Americans watch the most graphics of a media plan, this may dis- is obsolete and not entirely accurate. For television; Whites, Hispanics, and Asian- qualify minority groups and related media instance, income is not always a true indi- Americans watch progressively less, in properties from being included in the cator of discretionary income. Ethnicity respective order (Nielsen Company, 2014). media buy—unless the advertiser specifi- and income are correlated negatively—the The rationale for an assessment of televi- cally requests their inclusion (Ofori, 1999). exception is the Asian-American audience, sion viewership by American multicultural It historically has not been standard for whom income is correlated with English audiences is warranted by the television practice for media agencies to propose proficiency (Coffey, 2013; Napoli, 2013). consumption of the ethnic population. ethnic media buys as routine parts of Because of cultural nuance, ethni- The advertising and media-research general-market media plans. Ethnic-media cally relevant media are suggested to be industry’s ubiquitous proclamations of planning oftentimes is done without the an advantageous tactic to reach ethnic- the overconsumption of television among context of the general- or mass-market minority consumers. The greatest amount the predominant American ethnic groups plan; instead, the primary focus is on ethnic of time spent with English-language media necessitates an evaluation that goes beyond

June 2018 JOURNAL OF ADVERTISING RESEARCH 3 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

The researchers posited that multicultural consumers and pleasurable feelings among consumers and actually may fulfill consum- are triggered by the uses and gratifications of ers’ salient emotional desires and needs (Holbrook and Hirschman, 1982). The television programming more than are Whites. audience’s ability to become so involved in the storytelling in specific genres of television programs is explained best by the debated reach and frequency numbers. relationship to their television viewing. narrative transportation, which is defined A thorough evaluation should consider An earlier study stated that Blacks and as the process of telling a story whereby the mediating influence of the program- Hispanics had a higher consumption of the receiver of the narrative becomes matic storylines and narratives that mul- television than other races and ethnici- involved fully and transported into a fic- ticultural populations might have in their ties (La Ferle and Lee, 2005). Uses and tional world. relationship with television viewership. gratifications theory (UGT) historically The characteristics of the consumer In one study, an advertiser’s valuation of has best explained overall media con- viewing the story are crucial to the sto- an audience’s distinct cultural traits and sumption (Albarran and Umphrey, 1994; ryteller and the overall narrative experi- preferences, defined as characteristics and Albert and Jacobs, 2008; Bondad-Brown ence. The demographics of the receiver conventions tied to one's ethnic group, was et al., 2012; Lin, 1999; Rubin, 1983). UGT (e.g., race and ethnicity, dominant-society the most common predictor of investment posits that specific broadcast and digital acculturation, age, gender, socioeconom- in Spanish-language audiences across all programming can fulfill one’s psychologi- ics, and language preference) are a sali- media platforms—more than language cal needs (Albert and Jacobs, 2008). The ent component of narrative transportation preference and social class (Coffey, 2014). programmatic UGT factors for television from the perspective of the receiver of the With the rapid growth of digital tech- are entertainment, surveillance, escape story (Green, Brock, and Kaufman, 2004; nology, what influence does media frag- and companionship, problem solving, Van Laer et al., 2014). Demographics, as mentation—through the audience’s use and personal identity (Albert and Jacobs, well as prior knowledge of the story, can of technology—have on multicultural tel- 2008; Lin, 1999). lead to outcomes of an altered pleasurable evision viewing? It is important to study UGT is more relevant today, given the state for the viewer when he or she is expe- ethnic television fragmentation so that rise of “computer-mediated communica- riencing narrative transportation (Green, advertisers and the media industry can tion” (Ruggiero, 2000, p. 3). The aspect of 2004; Leary and Buttermore, 2003). Early give greater context to the simplistic dec- interactivity and the consumer’s ability to literature on race and ethnicity suggested laration that minorities watch more televi- participate authentically with television that once demographic variables are con- sion. With this examination, ethnic media programs (e.g., reality-based television trolled, consumer differences by race and spending and the proportion of culturally shows such as NBC’s The Voice or ABC’s ethnicity became nonsignificant (Cui, 1996; nuanced creative content may begin to Dancing with the Stars) increase audience Feldman and Star, 1968; Sojka and Tansu- become more representative of the multi- control, convenience, and a choice set that haj, 1995). As a result, demographics may cultural population and its buying power can fulfill the audience’s self-indulgent influence—or covary—the levels of televi- in the marketplace. The current approach entertainment needs (Ruggiero, 2000). sion viewership (McCarty & Shrum, 1993), to minority-audience valuation and per- With this increased interactivity in the so it is necessary to consider the effects ception of multicultural marketing is doing media landscape, the applicability of UGT of demographics with UGT variables a disservice to media companies, advertis- has more significance in describing the (Sheth, 1977). ers, media agencies, and the American eth- psychological needs of today’s consumer nic consumer. relative to television viewership. Diffusion of Innovations Theory All audiences, in particular ethnic view- UGT and narrative transportation also are THEORETICAL FRAMEWORK ers, also are watching specific genres of related closely to the medium in which Uses and Gratifications Theory (UGT) television programs for their entertain- ethnic consumers view the storytelling. In The theoretical support for this research ment value and receiving pleasure from today’s society, consumers multitask con- is rooted in further understanding the storytelling. Entertainment products, stantly in their everyday life and interact the behaviors of ethnic audiences in such as movies, arouse latent whimsical with multiple modalities, such as visual

4 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

and auditory stimuli through smartphones study titled “State of Cable and Digital Asian-Americans were interviewed in and tablets. The consumption of entertain- Media: Multicultural Edition” (Horowitz English. (For additional information on ment narratives through new technol- Research, 2013). Based in New Rochelle, the methodology, please visit the follow- ogy—smartphones, tablets, laptops—is New York, Horowitz Research is a coming website: http://www.horowitzresearch. explained best by the diffusion of innova- pany known for practitioner multicultural com/syndicated-research/2018-studies/ tions (Rogers, 1976). The rate of adoption is cable-television consumption studies and state-of-pay-tv-2018/.) influenced by multiple perceptions of rela- has conducted this particular research tive advantage, compatibility, and ease of study since 1999. In lieu of Nielsen ratings Data Screening and Descriptives use over an existing technology, whereby data, which to date never have been sub- The variables in the dataset were coded a community chooses to modify this inno- jected to academic analysis, the Horowitz for parametric testing. The dataset then vation to fit its culture (Trahan, 2013). For dataset is a robust and representative sam- was screened for univariate outliers and instance, 90 percent of Blacks and 86 per- ple of the U.S. population. normal distribution of the appropriate cent of Hispanics own a cell phone, com- Because this was a multicultural study, variables. Within the 2,086-cases dataset, pared with 84 percent of Whites (Lopez Asian-Americans, Blacks, and Hispanics some of the variables had missing values. et al., 2013)—a carryover from the digital were oversampled. With data collection Each variable of interest had 2,086 cases divide. The high rate of adoption of mobile managed by Horowitz Research, the sur- unless otherwise noted. Each variable is technology among people of color qualifies vey was conducted between January and described below in subsequent paragraphs them as early adopters (Rogers, 1976). February 2013 among a national sample and tables. Together, UGT and diffusion of innova- of 2,086 heads of television households The mean age of the sample was 46.2 tions, when controlling for demographics, 18 years of age or older. Potential partici- years (SD = 16.38). The racial and ethnic may work in concert to better explain—or pants were contacted through 1,267 tele makeup was 42 percent Hispanic, 30 per- mediate—the relationship multicultural phone interviews and 819 online surveys cent Black, 17 percent White, and 11 per- consumers have with television viewer- and included multichannel (e.g., cable, cent Asian. Education composition (n = ship. Specifically, the current researchers satellite, and telecommunications televi- 2,008) was as follows: posited that: sion subscribers) and nonmultichannel subscribers. • 9.3 percent attended some high school RP: Multicultural consumers are trig- A stratified random sample of consum- or less; gered by the uses and gratifica- ers using cable systems was the primary • 24.3 percent graduated high school; tions of television programming, sampling method. The researchers chose • 15.2 percent attended trade or vocational more than whites. Uses and grat- at random 100 qualified sampling points school; ifications then entices the mul- (cities with a population of 50,000 or more) • 3.2 percent attended some college; ticultural consumer to interact from among all cities with a population of • 21.5 percent completed college; with the programming via tech- 50,000 or more in the continental United • 26.5 percent had a postgraduate degree. nology (diffusion of innovations). States (N = 516), with probability of selec- As a result, the use of technology tion in proportion to population. Quotas Operationalization of Variables (media fragmentation) will have of 50 percent men and 50 percent women Independent Variable (Multicultural- a negative effect on daily televi- were set within each sampling point. ism). The independent variable was a sion viewership. The exploration For ethnic composition, quotas were set dichotomous, dummy variable represent- of this relationship can have by race and ethnicity. A total of 883 inter- ing multicultural heritage. It was coded implications for multicultural views were completed among Hispanic as African-American, Asian-American, or media planning and ethnic- homes, 689 among Black and African- Hispanic =1 and White = 0 (N = 2,086; 84 audience investment. American homes, and 206 among Asian- percent multicultural). American homes. Data were weighted METHODology to match U.S. Census figures, includ- Dependent Variable (Self-Reported Sample ing ethnicity. Hispanic respondents Daily Television Consumption). Partici- The authors were given access to Horow- were surveyed in Spanish or English, pants were asked, “Approximately how itz Research’s raw data from its 2013 depending on their language preference. many hours, on average, do you watch TV

June 2018 JOURNAL OF ADVERTISING RESEARCH 5 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

Table 1 Dichotomous Diffusion of Innovations Variables Survey Question n % Answered “Yes” How many Smart TVs—TVs built with a built-in Internet connection that allows you to access various 1,918 19.9 websites, applications, and services—if any, do you have at home? Do you personally have a PS3 (PlayStation 3)? 2,061 22.8 Do you personally have an Xbox 360? 2,062 24.2 Do you personally have a Wii or Wii U? 2,057 30.0 Do you personally have a Blu-ray DVD player? 2,054 33.3 Do you personally have a paid DVD subscription to Netflix? 2,065 21.5 Do you personally have an online stream service to Netflix, Hulu, or Amazon Prime? 2,067 27.5 Do you personally have an iPod or portable MP3 player that allows you to watch videos? 2,074 33.7 Do you personally own an iPhone, Android, or Windows phone? 2,075 46.5 Do you personally own a desktop computer or laptop? 2,075 77.6

Note: Responses were coded as 0 = no, 1= yes.

content, including news, sports, TV shows, analysis (Linting and Van Der Kooij, 2012) The authors averaged these variables movies, music videos, etc., on a TV set or to determine the nonlinear relationships together to create a continuous (.00–1.00) any other devices in a typical day? And by among the respective UGT and diffusion- variable of uses and gratifications (n = that, I mean both day and night” (n = 1,829; of-innovations variables. They used a 1,992; M = .60, SD = .24). M = 3.91, SD = 1.92). method prescribed in previous research Similarly, the authors used categori- (Linting and Van der Kooij, 2012) to find cal principal-components analysis to Mediating Variable (Uses and Gratifica- single dimensions for uses and gratifica- determine the relationship among the tions). In keeping with previous research tions and for diffusion of innovations for diffusion-of-innovations variables, which (Lin, 1999; Rubin, 1983), the authors coded further analysis. determined participants’ ownership of programming genres within the dataset as The 10 uses and gratifications variables various media devices. These nine vari- dichotomous variables in terms of whether had a Cronbach’s alpha of .70, with all ables had a Cronbach’s alpha of .78, respondents watched a specific genre component items (in parentheses) loading with component loadings (in paren- at least once or twice a week. Research above .40 (as prescribed by Linting and theses) above .40 (Linting and Van der assistants who were unfamiliar with the Van der Kooij, 2012): Kooij, 2012): project and research question grouped the genre variables by UGT factors (See • music programming (.56); • online streaming service to Netflix, the Appendix). • talk shows (.54); Hulu, or Amazon Prime (.68); • live sports events (.54); • iPod or portable MP3 player (.64); Mediating Variable (Diffusion of Innova- • sports news (.54); • smartphone (iPhone, Android, or Win- tions). The dichotomous variables from the • reality shows (.53); dows phone; .63); Horowitz dataset were grouped together • documentaries (.51); • DVD subscription to Netflix (.61); to represent technology consumption and • movies (.48); • Wii or Wii U (.60); usage (See Table 1). • comedy programs (.48); • Blu-ray DVD player (.58); Given the nominal characteristics of • news (.47); • PS3 (Playstation 3; .57); the independent variables, the authors • cooking and home-improvement • Xbox (.55); used nonlinear principal-components shows (.47). • desktop or laptop computer (.53).

6 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

The authors also averaged these items together to create a continuous (.00–1.00) X Y Multicultural c Daily TV Viewership measure of diffusion of innovations (n = 1,999; M = .35, SD = .27).

Covariates. Five demographic covariates were selected from the Horowitz M a M dataset. In addition to age and education 1 3 2 Uses and Gratifications Diffusion of Innovations (See “Data Screening and Descriptions”, earlier) the demographics were as follows:

a1 b2

• gender (N = 2,086; 57 percent female); a2 b1 • language preference (0 = English, 1 = Spanish; N = 2,086; 22 percent Spanish); • natural-born citizenship (0 = yes, 1 = X Y Multicultural Daily TV Viewership no; n = 2,060; 27 percent not born in the c' United States). Figure 1 Multistep Mediation Model of the Relationship Income was not considered as a covariate because there were many missing cases Between Multiculturalism and Television Viewing for this variable—only 1,267 usable obser- vations—which would have reduced fur- self-reported hours of television watched sequence of mediators—namely, uses ther the power of this study. Social-class daily (See Figure 1). The mediators, in and gratifications (M1) and diffusion of identification is related most strongly to sequential order, were the uses and grati- innovations (M2)—partially mediated the educational credentials, followed by occu- fications (M1) and the diffusion of innova- effect of multiculturalism on the daily pation and income (Coleman, 1983). Given tions (M2) variables operationalized above hours of television consumed. the completeness of the education vari- (See “Covariates”, earlier). The independ- The psychological needs of ethnic televi- able, it was the best proxy of social class in ent variable (X) was the dichotomous sion consumption were met by uses and this dataset. dummy variable representing multicul- gratifications (Path a1 in Figure 1; B = 0.06, tural heritage. p < .00), which resulted in a significant RESULTS To rule out confounding effects, the effect on the use of technology (diffusion

Multistep Mediation authors included the following demo- of innovations; Path a3 in Figure 1; B = 0.18, Mediation analysis posits how, or by graphic variables in the analysis as covari- p < .00). Last, the use of technology (dif- what means, an independent variable (X) ates: gender, language preference (Spanish fusion of innovations) contributed to the affects a dependent variable (Y) through versus English), native-born citizenship, decrease in daily television viewership one or more potential intervening varia- and level of education. They saved the (Path b2 in Figure 1; B = −0.64, p < .01). The bles, or mediators (Mx: Preacher & Hayes, addition of the age covariate for a second, 95 percent confidence interval (CI) for the 2008). The goal of multistep mediation separate multistep mediation analysis. indirect effect was obtained with 10,000 is to investigate the direct and indirect Given the missing cases, the final sample bootstrap resamples and supports the par- effects of X on Y while modeling a pro- size for this analysis was 1,617. tial multistep mediation (BX → M1 → M2 cess where X causes M1, causing M2 con- The path estimates confirmed a par- → Y = −0.007, 95% CI [−0.01, −0.002]). cluding with Y as the final consequence tial multistep mediation process. The An additional multistep mediation (Hayes, 2017). total-effect model confirmed a signifi- analysis was performed in which age The authors submitted the Horowitz cant relationship between multicultural- was added to the model as an additional data to a multistep mediation analysis ism and the hours of television watched covariate, accompanying the four existing using PROCESS, Model 6 (Hayes, 2017). daily (Path c in Figure 1; B = 0.52, p < .00). covariates (gender, language preference, The dependent variable (Y) was the The results confirmed that the proposed native-born citizenship, and education)

June 2018 JOURNAL OF ADVERTISING RESEARCH 7 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

Table 2 → Y = 0.10, 95% CI [0.05, 0.17]). (Tables 2 Multistep Mediation Results with Four Covariates and 3 detail the results of both multistep Model Path Estimates mediation analyses.)

Model Path Estimates Practical Outcomes Path Coefficient (B) SE t p Overall, television viewership is a fickle human behavior; thus, it is difficult to pre- a1 0.06 0.02 3.41 <.000 dict. One study (Rubin, 1983), for example, a2 0.03 0.02 1.98 .04 had linear regression models of television- a3 0.18 0.03 7.05 <.000 viewership motivations with small effect b1 1.88 0.20 9.54 <.000 sizes. Other hidden variables that might explain television viewership among mul- b2 −0.64 0.19 −3.45 <.001 ticultural consumers were not discovered c 0.52 0.13 4.02 <.000 in this dataset. c′ 0.44 0.13 3.52 <.000 Given this limitation, the results of this analysis provide support for and answer Indirect Effect the research proposal of uses and gratifica- Path Effect (B) LL 95% CI UL 95% CI SE tions and diffusion of innovations—controlling for demographic influence—mediating X → M1 → Y 0.10 0.05 0.17 .03 the relationship of multiculturalism and tel- X → M → Y −0.02 –0.05 −0.003 .01 2 evision viewership. This mediation model

X → M1 → M2 s → Y −0.007 −0.01 −0.0002 .003 better explains the behavior of multicultural

2 2 television consumption. Multicultural con- Note: Path a1: F(5, 1611) = 5.11, p < .001, R = .02. Paths a2 and a3: F(6, 1610) = 29.92, p < .000, R = .10. Paths b1, b2, and c′: F(7, 1609) = 17.72, p < .000, R2 = .07. Path c: F(5, 1611) = 5.62, p < .000, R2 = .02. Covariates were gender, language sumers watch television for the uses and preference (Spanish versus English), native-born citizenship (born in the United States) and level of education. LL = lower gratifications they receive from watching level; CI = confidence interval; UL = upper level. particular genres of television programming. This predicates a use of technology Table 3 and results in less television being con- Multistep Mediation Results with Age as the Fifth Covariate sumed. Multicultural consumers watch television for the uses and gratifications they Path Effect (B) LL 95% CI UL 95% CI SE receive from watching particular genres of

X → M1 → Y 0.10 0.05 0.17 0.03 television programming. This predicates a use of technology and results in less televi- X → M2 → Y −0.005 −0.03 0.004 0.007 sion being consumed. X → M → M → Y −0.005 −0.01 −0.001 0.003 1 2 An example is the ABC network’s show Note: Covariates were gender, language preference (Spanish versus English), native-born citizenship (born in the United “Scandal”. A prime-time drama with a States), level of education, and age. Evidence for single-step mediation with the inclusion of the age covariate: (Path b , 2 strong African-American following, the B = −0.47, p = .06), F(8, 1608) = 16.18, p < .000, R2 = .08. LL = lower level; CI = confidence interval; UL = upper level. show is a trending topic on Twitter at times correlating to its time slot (Cabosky, 2016). Partial multistep mediation exists (Path c, from the initial multistep mediation. through diffusion of innovations (BX → M2 B = 0.52, versus Path c′, B = 0.44); however, Although the indirect effect of the mult → Y = −0.005, 95% CI [−0.02, 0.004]) was no the mediation of uses and gratifications istep mediation remained significant, with longer significant, because the 95 percent (Path b1, B = 1.88) explains a majority of the 95 percent CI not crossing zero (BX CI crossed zero. The addition of age as the the indirect effect of multicultural televi-

→ M1 → M2 → Y = −0.005, 95% CI [−0.01, final covariate ruled out multistep media- sion viewership.

−0.001), Path b2 was now marginally signifi- tion. Single-step mediation was significant Last, the influence of age appeared to cant (Path b2 = −0.37, p = .06). The mediation through uses and gratifications (BX → M1 have a significant demographic influence

8 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

on ethnic television viewership. When For multicultural media planning, ethnic ethnic viewership of a program or of the age was added to the multistep media- media is siloed, and, many times, minority- population. The result is creative spillover tion model as a fifth covariate, diffusion targeted platforms only are considered for between general and multicultural markets of innovations became marginally signifi- multicultural audiences. It makes sense and between digital and broadcast modal- cant (Path b2: Age, B = −0.37, p = .06), from intuitively that American racial and eth- ities. This is where a context planner, being highly significant without age (Path nic groups do not consume ethnic media versed in multicultural nuance, can inject b2: B = −0.64, p < .01). Age—in tandem with solely unless they are isolated linguisti- a multicultural consumer strategy into the the other covariates and mediating vari- cally (Coffey, 2008). media-planning and -buying process that ables—had a spurious influence on the The results of this research demonstrate sets the course for greater media engage- relationship of diffusion of innovations that program genre best explains the rela- ment with ethnic audiences. to daily multicultural television viewer- tionship racial and ethnic groups have ship. A key takeaway from this analysis with television, which makes the prac- Limitations is that a total market approach—instead tice of minority discounts tantamount to Because the Horowitz data were collected of a siloed, ethnic-media approach—to discrimination. In television, a target rat- for a practitioner-based study, there are multicultural media planning is impor- ing point—or any commensurate media limitations. The dataset was not designed tant, because programmatic genres, gauge—should be a standard unit of for academic analysis. As previously technological media fragmentation, and measurement, whether one is examining a mentioned, there were hidden variables the influence of demographics have general market or an ethnic-media broad- that showed up as noise and missing been demonstrated to explain the influ- cast property. Genre platforms should be cases for variables (e.g., income) that bet- ence of television viewership among evaluated instead of networks—in partic- ter might explain or covary television ethnic consumers. ular for ethnic consumers—hence the role viewership and multiculturalism. Despite of and the need for context planning. A this limitation, significant relationships DISCUSSION total market approach to media planning, explaining ethnic television viewership Implications for Practice whereby multicultural and general-market were discovered. All successful media planners ensure that segments are viewed as one segment with The responses in this dataset were also their media plans have the best media cultural nuance addressed in the tactics, is all self-reported survey data. A previous mix in the most efficient manner. Given appropriate to improve targeting, message study (Nisbett and Wilson, 1977) refer- the challenges of measurement of ethnic rotation and scheduling, media buying, enced the challenge of self-reported data. audiences, the results of this study demon- and return on investment metrics for the Participants in Nielsen Media Research’s strate that content across all broadcast and ethnic population. samples were aware that they were part cable networks—regardless of ethnic spe- The study suggests that client-side of a study, however. This can lead to cialty—should not be overlooked because multicultural marketers and media agen- response bias in viewing habits (Aswin, of erroneous measures or perceptions. cies impose a spillover approach as a Mittal, and Vasudevan, 2016; Givon, These findings suggest the importance of method of implementing the total-market Davidman, Lodzki, and Sherman, 2013), the context-planning discipline in a media approach toward media planning and buy- which is common for news and prime- agency as it relates to media planning for ing. For instance, on minority-formatted time programming. Although both noise ethnic groups. The role of the context plan- networks, the culturally relevant creative and response bias are imperfections, sub- ner is to represent the media of interest from messaging—in the context of culturally sequent research can work to mitigate the consumer’s perspective. Context plan- nuanced programmatic storytelling—will these shortcomings. ners ensure that media plans are grounded be trafficked to air. As evidenced with this Related future research on the topic in consumer insight, are diversified relative research, ethnic groups are watching tele- of noise can delve into uncovering other to modality, and inspire creative media vision genres on general-market television mediators and covariates that explain the solutions, while maintaining the integrity networks. relationship ethnic consumers have with of both the brand and the communications- The authors suggest that media planners television. By incorporating experimental campaign idea (Hatcher, 2005). adopt a methodology of percentage of rota- design and qualitative methods, research- Media strategy focuses solely on the tion of culturally nuanced commercials, ers may investigate direct questions most efficient media-vehicle selection. which could be based on the percentage of related to viewership behavior that were

June 2018 JOURNAL OF ADVERTISING RESEARCH 9 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

missing from the Horowitz dataset. The group rating point” (James and Swartz, television plays a substantial role in Amer- authors anticipate that the results from this 2005). Commensurate research in this area ican society—especially among minority additional research will complement the could assist media-research companies to segments—as a medium to communicate analysis in this article and create a more develop a media-measurement standard entertainment, information, and news. complete picture in understanding ethnic that incorporates an influence of multicul- Minorities, in addition, overwhelmingly television viewership. turalism, uses and gratifications, diffusion patronize programming on networks that of innovations, and other hidden demo- may be hampered by minority discounts, Conclusions graphic factors and variables discovered in because it is their choice of programming The results of this study suggest a chal- subsequent research. This outcome would (Ofori, 1999). The practice of minority lenge for networks such as ABC, CBS, be similar to the suggestion of a previous discounts has implications for all Ameri- NBC, and Fox to attract the English- author (Wilbur, 2008) relative to media cans, however. With decreased incentives speaking Hispanic, Asian, and Black fragmentation caused by DVRs. to broadcast a diversity of ideas, includ- audiences through genre. This has been The phenomenon of greater ethnic tel- ing programming of interest to minori- evidenced with shows such as “Modern evision consumption and the subsequent ties, all Americans receive a narrower Family,” “Fresh off the Boat,” and “Black- ethnic-media fragmentation is not unique range of information and perspectives. ish”—general-market programming that to the United States. The rise of ethnically Media planners and advertisers must pro- resonates with and is viewed primarily by targeted media in the United Kingdom hibit practices that discriminate against ethnic consumers. Advertisers also must presents a similar challenge. A study by ethnic consumers in the media market- understand the population growth and the Institute of Practitioners in Advertis- place, while continuing to make strides in spending potential of the major American ing (2014) demonstrated that fewer than 20 deploying programming that is responsive ethnic groups. percent of ethnic minorities in the United to the needs of all segments of society. Because ethnic groups often are Kingdom solely watched general-market viewed as monolithic audiences in a television. Another 16 percent watched media-planning context, the findings only ethnic programming. This left another ACKNOWLEDGMENTS from this research should prompt adver- 64 percent of Britons who consumed both tisers to segment ethnic audiences by ethnic and mainstream television. “State of Cable and Digital Media Study, Mul- demographics, psychographics, and atti- Media fragmentation is also an issue ticultural Edition” © 2013 Horowitz Associ- tudes—just like they do for the overall, given that the Black and minority-ethnic ates, Inc., Market and Multicultural Research. general-market population. This, in turn, audiences consume more digital media These data are proprietary and cannot be used, will challenge media agencies to find the than the White British audience (Institute printed, or duplicated without express permis- very lucrative, appropriate, and brand- of Practitioners in Advertising, 2014). It sion from Horowitz Associates. loyal ethnic audience through the appro- can be argued that many of these minor- priate television programming and not ity groups might feel misrepresented and ABOUT THE AUTHORS relegate multicultural media planning disenfranchised by programming in the solely to ethnic-media networks. In the British mainstream media. The rise of new, J. P. James is an assistant professor of marketing at final analysis, all media companies need ethnically driven media channels and con- Salem State University. James conducts research on to deliver ethnic audiences through content, as it relates to television consumption multicultural marketing and marketplace diversity. He tent and programming to sustain their in the United Kingdom and other coun- has been published in the Journal of Public Policy & advertising revenues. tries, presents an opportunity for further Marketing, Journal of Management Policy and Practice, The answer to the research question study of ethnic television viewership in and Journal of Integrated Marketing Communications, posed in this article also has implications these nations. among other journals. for media-measurement models. There Whether in the United States or abroad, is a significant programmatic and tech- understanding ethnic audiences in terms Tyrha M. Lindsey-Warren is a clinical assistant professor nological fragmentation component that of their television consumption is of high of marketing at the Hankamer School of Business at influences multicultural television con- importance so that media agencies can Baylor University. Lindsey-Warren’s research streams sumption. A prior author (Muse, 2001, p. deliver these populations efficiently to are rooted in storytelling and its effects on millennials, 87) posited the idea of a “multicultural advertisers. As evidenced in this research, empowerment, health edutainment, movies, media, and

10 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

advertising. She also spent more than 15 years working Aswin, T. S., Mittal, K., and S. K. Vasudevan. Coleman, R. P. “The Continuing Significance of as a practitioner for the likes of Quincy Jones*David “Efficient Television Ratings System with Social Class to Marketing.” Journal of Consumer Salzman Entertainment, Alvin Ailey American Dance Commercial Feedback Applications.” In The Research 10, 3 (1983): 265–280. Theater, and UniWorld Group. Her work has been International Symposium on Intelligent Systems featured in the Journal of Cultural Marketing Strategy Technologies and Applications (pp. 433–447), S. Cui, G. “Integrate Ethnic Diversity into Market- and the book Entertainment Values (Palgrave UK). Mitra, S. Thampi S and E. S. El-Alfy, eds. Berlin: ing Education.” Marketing Education Review 6, 1 Springer International, 2016. (1996): 1–7.

REFERENCES Deshpande, R., W. D. Hoyer, and N. Donthu. Bearden, W. O., R. S. Headen, J. E. Klompmaker, “The Intensity of Ethnic Affiliation: A Study of and J. E. Teel. “Attentive Audience Delivery of Ad Age Datacenter. (2014, July 27). “Hispanic the Sociology of Hispanic Consumption.” Jour- TV Advertising Schedules.” Journal of Marketing Fact Pack 2014.” Retrieved from the Advertising nal of Consumer Research 13, 2 (1986): 214–220. Research 18, 2 (1981): 187–191. Age website: http://adage.com/trend-reports/ report.php?id=94 Dickson, P. R. “GRP: A Case of a Mistaken Iden- Bondad-Brown, B.A., R. E. Rice, and K. E. tity.” Journal of Advertising Research 31, 1 (1991): Pearce. “Influences on TV Viewing and Online Adweek. (2001, February 26). “A Crippling Error: 55–59. User-Shared Video Use: Demographics, Genera- Leaving Ethnic Americans out of Your Plan.” tions, Contextual Age, Media Use, Motivations, Retrieved from http://www.adweek.com/ DiMaggio, P., E. Hargittai, C. Celeste, and S. and Audience Activity.” Journal of Broadcasting & news/advertising/crippling-error-leaving- Shafer. “From Unequal Access to Differenti- Electronic Media 56, 4 (2012): 471–493. ethnic-americans-out- your-plan-46891 ated Use: A Literature Review and Agenda for Research on Digital Inequality.” In Social Inequal- Cabosky, J. “Social Media Opinion Sharing: ity, K. Neckerman, ed. New York: Russell Sage Albarran, A. B., and D. Umphrey. “Profile: An Beyond Volume.” Journal of Consumer Marketing Foundation, 2004. Examination of Television Motivations and 33, 3 (2016): 172–181. Program Preferences by Hispanics, Blacks, and D’Rozario, D., and G. Yang. “A Comprehensive Whites.” Journal of Broadcasting & Electronic Cannon, H. M. “The Naive Approach to Demo- Typology of Ethnic Identities: Implications for Media 37, 1 (1993): 95–103. graphic Media Selection.” Journal of Advertising Marketing and Public Policy.” Journal of Public Research 24, 3 (1984): 21–25. Policy & Marketing 34, 2 (2015): 173–193. Albarran, A. B., and D. Umphrey. “Marketing Cable and Pay Cable Services: Impact of Ethnic- Cannon, H. M., J. D. Leckenby, and A. Aber- Farris, P. W., and M. E. Parry “Clarifying Some ity, Viewing Motivations, and Program Types.” nethy. “Beyond Effective Frequency: Evaluat- Ambiguities Regarding GRP and Average Fre- Journal of Media Economics 7, 3 (1994): 47–58. ing Media Schedules Using Frequency Value quency: A Comment on ‘GRP: A Case of Mis- taken Identity.’” Journal of Advertising Research Planning.” Journal of Advertising Research 42, 6 Albert, T. C., and R. D. Jacobs. “Television Atti- 31, 6 (1991): 75–77. tudes and TV Types of African- Americans, Lati- (2002): 33–47. nos, and Whites.” Journal of Advertising Research Feldman, L. P., and A. D. Star. “Racial Factors 48, 2 (2008): 235–246. Coffey, A. J. “The Case for Audience Isolation: in Shopping Behavior.” In A New Measure of Language and Culture as Predictors of Adver- Responsibility for Marketing, K. K. Cox and B. M. tiser Investment. International Journal on Media Association of Hispanic Advertising Agencies. Enis, eds. Chicago: American Marketing Asso- Management, 10, 2 (2008): 81–90. (2013). “AHAA Total Market Benchmark Study: ciation, 1968. Advertisers Preliminary Findings.” Retrieved Coffey, A. J. “Understanding the Invisibility from http://www.ahaa.org/Portals/0/Events/ Franklin, E. T. “Are You Reaching the Black- of the Asian-American Television Audience: AHAA%20U/Past%20Events/2014/Total%20 American Consumer? How the Rise of Multicul- Market%20 Agency.pdf Why Marketers Often Overlook an Audience turalism Ended Up Sending Mixed Messages.” of ”Model” Consumers. Journal of Advertising Journal of Advertising Research 54, 3 (2014): 259– Association of National Advertisers. (2012). Research, 53(1) (2013) 101–118. 262. https://doi.org/10.2501/JAR-54-3-259-262 “Marketers Leverage Newer Media Platforms to

Reach Growing Number of Multicultural Con- Coffey, A. J. “The Power of Cultural Factors Givon, D., G. Davidman, M. Lodzki, and V. Sher- sumers” (press release). Retrieved from http:// in Spanish-Language Advertising.” Journal of man. “Methods Circuits Apparatuses Systems www.ana.net/content/show/id/24292 Advertising Research 54, 3 (2014): 346–355. and Associated Computer Executable Code for

June 2018 JOURNAL OF ADVERTISING RESEARCH 11 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S.

Providing Viewer Analytics Relating to Broad- Retrieved from https://www.kantarmedia. Linting, M., and A. Van der Kooij. “Nonlinear cast and Otherwise Distributed Content” (U.S. com/us/newsroom/press-releases/kantar- Principal Components Analysis with CATPCA: Patent Application 13/892). Alexandria, VA: media-reports-us-advertising-expenditures- A Tutorial.” Journal of Personality Assessment 94, U.S. Patent and Trademark Office, 2013. increased-0-9-percent-in-2013 1 (2012): 12–25.

reen opez onzalez arrera at G , M. C. “Storytelling in Teaching.” APS Kaufman-Scarborough, C. “Asian-American L , M. H., A. G -B , and E. P - Observer 17, 4 (2004): 37–39. Consumers as a Unique Market Segment: Fact ten. Closing the Digital Divide: Latinos and Tech- nology Adoption. Washington, DC: Pew Hispanic or Fallacy?” Journal of Consumer Marketing 17, 3 Green, M. C., T. C. Brock, and G. F. Kaufman. Center, 2013. (2000): 249–262. “Understanding Media Enjoyment: The Role of Transportation into Narrative Worlds.” Commu- McCarty, J. A., and L. J. Shrum. “The Role of La Ferle, C., and W. N. Lee. “Can English Lan- nication Theory 14, 4 (2004): 311–327. Personal Values and Demographics in Predict- guage Media Connect with Ethnic Audiences? ing Television Viewing Behavior: Implications Ethnic Minorities’ Media Use and Representa- Hatcher, K. “Context Planning: The Door to for Theory and Application.” Journal of Advertis- tion Perceptions.” Journal of Advertising Research Media’s Future.” Journal of Integrated Marketing ing 22, 4 (1993): 77–101. 45, 1 (2005): 140–153. Communications 15 (2005): 37-43. Muse, J. The Shaman Chronicles Book One: The 7 Leary, M. R., and N. R. Buttermore. “The Evolu- Hayes, A. F. Introduction to Mediation, Moderation, Senses of Multicultural Marketing. Los Angeles: and Conditional Process Analysis: A Regression- tion of the Human Self: Tracing the Natural His- Muse Creative Holdings, 2001. Based Approach. New York: Guilford Press, 2017. tory of Self-Awareness.” Journal for the Theory of Social Behaviour 33, 4 (2003): 365–404. Napoli, P. M. “Audience Measurement and

Henderson, G. R., and J. D. Williams. “From Media Policy: Audience Economics, the Diver- Exclusion to Inclusion: An Introduction to the Leckenby, J. D., and M. M. Boyd. “How Media sity Principle, and the Local People Meter.” Com- Special Issue on Marketplace Diversity and Directors View Reach/Frequency Model Evalu- munication Law and Policy 10, 4 (2005): 349–382. Inclusion.” Journal of Public Policy & Marketing ation Standards.” Journal of Advertising Research 32 (2013): 1–5. 24, 5 (1984): 43–52. Napoli, P. M. Audience Economics: Media Insti- tutions and the Audience Marketplace. New York:

Holbrook, M. B., and E. C. Hirschman. “The Leckenby, J. D., and K. H. Ju. “Advances in Columbia University Press, 2013. Experiential Aspects of Consumption: Con- Media Decision Models.” Current Issues and sumer Fantasies, Feelings, and Fun.” Journal of Research in Advertising 12, 1–2 (1989): 311–357. Nelson-Field, K., and E. Riebe. “The Impact of Consumer Research 9, 1 (1982): 132–140. Media Fragmentation on Audience Targeting: An Empirical Generalization Approach.” Journal Leckenby, J. D., and H. Kim. “How Media Horowitz Research. State of Cable and Digital of Marketing Communications 17, 1 (2011): 51–67. Directors View Reach/Frequency Estimation: Media Study, Multicultural Edition. New Rochelle, Now and a Decade Ago.” Journal of Advertising NY: Author, 2013. Nielsen Company. Resilient, Receptive and Rel- Research 34, 5 (1994): 9–21. evant. New York: Author, 2013. Humphreys, J. M. The Multicultural Economy Leckenby, J. D., and S. Kishi. “The Dirichlet 2015. Athens, GA: Selig Center for Economic Nielsen Company. The U.S. Digital Consumer Multinomial Distribution as a Magazine Expo- Growth, University of Georgia, 2015. Report. New York: Author, 2014. sure Model.” Journal of Marketing Research 21, 1 (1984): 100–106. Institute of Practitioners in Advertising. The Nisbett, R. E., and T. D. Wilson. “Telling More New Britain. London: Author, 2014. than We Can Know: Verbal Reports on Mental Leckenby, J. D., and M. D. Rice. “A Beta Bino- Processes.” Psychological Review 84, 3 (1977): mial Network TV Exposure Model Using Lim- James, J. P., and J. E. Swartz. “The Multicultural 231–259. Paradigm.” Journal of Integrated Marketing Com- ited Data.” Journal of Advertising 14, 3 (1985):

munications 15 (2005): 44–52. 25–31. Ofori, K. A. (1999). “When Being No. 1 Is Not Enough: The Impact of Advertising Practices Kantar Media. (2014). “Kantar Media Reports Lin, C. A. “Online Service Adoption Likeli- on Minority-Owned & Minority-Formatted U.S. Advertising Increase 0.9 Percent in 2013, hood.” Journal of Advertising Research 39 (1999): Broadcast Stations.” Retrieved from the Office Fueled by Large Advertisers” (press release). 79–89. of Communications Business Opportunities,

12 JOURNAL OF ADVERTISING RESEARCH June 2018 An Examination of Television Consumption By Racial and Ethnic Audiences in the U.S. thearf.org

Federal Communications Commission: http:// Rubinson, J. “Editorial: Marketing in the Era of Surpass Total TV by 2018.” Retrieved from the civilrightsdocs.info/pdf/reports/Being-No-1-I Long-Tail Media.” Journal of Advertising Research Adweek website: http://www.adweek.com/ s-Not-Enough-Ad-Study-Ofori.pdf 48, 3 (2008): 301–302. news/television/digital- media- now- bigger- national-tv-advertising-will-surpass-total- Preacher, K. J., and A. F. Hayes “Asymptotic and Ruggiero, T. E. “Uses and Gratifications Theory tv-2018-158360 Resampling Strategies for Assessing and com- in the 21st Century.” Mass Communication & Soci- paring Indirect Effects in Multiple Mediator ety 3, 1 (2000): 3–37. Models. Behavior Research Methods, 40, 3 (2008): Trahan, M. P. “Beyond Technology: Revisiting

879–891. Schultz, D. E., and Schultz, H. F. “Transi- Diffusion of Innovation Theory Application to tioning Marketing Communication into the Other Disciplines.” In National Social Science incon Twenty-First Century.” Journal of Marketing Com- R , E. T. (2012, April 10). “Arbitron Case Proceedings, vol. 54: New Orleans Professional Shows You Can’t Trust Multicultural-Audience munications 4, 1 (1998): 9–26. Development Conference, 2013. Ratings.” Retrieved from the Advertising Age website: http://adage.com/article/the-big- Sheth, J. N. “Demographics in Consumer Van Laer, T., L.M. visconti, and M. wetzels tent/arbitron-case-confirms-flaws-minority- Behavior.” Journal of Business Research 5, 2 (1977): audience-ratings/234021/ 129–138. (2014), “The Extended Transportation-Imagery Model: A Meta-Analysis of the Antecedents and Rogers, E. M. “New Product Adoption and Dif- Sojka, J., and P. S. Tansuhaj. “Cross-Cultural Consequences of Consumers’ Narrative Trans- fusion.” Journal of Consumer Research 2, 4 (1976): Consumer Research: A Twenty-Year Review.” In portation,” Journal of Consumer Research 40 (Feb- 290–301. NA—Advances in Consumer Research, vol. 22, F. R. ruary) (2014): 797–817. Kardes and J. Sujan, eds. Provo, UT: Association

Rubin, A. M. “Television Uses and Gratifica- for Consumer Research, 1995. Wilbur, K. C. “How the Digital Video Recorder tions: The Interactions of Viewing Patterns and Motivations.” Journal of Broadcasting & Electronic Thielman, S. (2014, June 16). “Digital Media Is (DVR) Changes Traditional Television Advertis- Media 27, 1 (1983): 37–51. Now Bigger than National TV Advertising, Will ing.” Journal of Advertising 37, 1 (2008): 143–149.

APPENDIX Genre Variables Grouped by Uses and Gratifications Theory Factors Entertainment Surveillance Escape/Companionship Problem Solving Personal Identity

Live sports events News (n = 2,066) Movies (n = 2,064) Documentaries (n = Religious programming (n = 2,055) 86.4 percent (yes) 80.4 percent (yes) 2,062) (n = 2,070) 55.9 percent (yes) 64 percent (yes) 31.4 percent (yes) Sports news Talk shows Comedy programs or sitcoms (n = Programs for Programming (n = 2,066) (n = 2,065) 2,060) children (n = 2,070) from a foreign country 47.9 percent (yes) 48.9 percent (yes) 68.7 percent (yes) 40.1 percent (yes) (n = 2,053) 23.3 percent (yes) Dramas (n = 2,071) Soap operas/telenovelas (n = Cooking/home 63.9 percent (yes) 2,069) improvement (n = 2,058) 29.1 percent (yes) 54.4 percent (yes) Original series (e.g., HBO) Reality shows (n = 2,064) (n = 2,069) 36.4 percent (yes) 51.1 percent (yes) Music programming Travel and tourism (n = 2,059) (n = 2,060) 42.7 percent (yes) 41.5 percent (yes)

Note: Data were drawn from participants’ answers to the following question: “For each of the following types of television programs, please tell me if you watch it at least once or twice a week” (0 = no, 1 = yes).

June 2018 JOURNAL OF ADVERTISING RESEARCH 13