Abstract
Purpose:
The aim of the study was to assess awareness of and receptivity to the U.S. Food and Drug Administration’s Fresh Empire tobacco public education campaign designed to reach Hip Hop–identified youth, who are at higher smoking risk.
Methods:
The evaluation uses a randomized treatment–control design with 15 campaign-targeted treatment and 15 control markets. We conducted surveys among 12- to 17-year-olds before campaign launch and at approximately 6-month intervals. Analyses explore treatment–control differences in Fresh Empire brand and video advertisement awareness at individual survey rounds and over time and perceived effectiveness of advertisements.
Results:
Awareness of the Fresh Empire brand was higher among youth in treatment than control markets following campaign launch (ps < .01). Awareness of the Fresh Empire brand increased more in treatment than control over time (adjusted odds ratio = 3.26, 95% confidence interval = 1.90–5.58). At follow-ups 1 and 3, youth in treatment (vs. control) were more likely to report high and less likely to report low awareness of video advertisements (ps < .05). There were no treatment–control differences in video awareness at follow-up 2 (not significant). Fresh Empire video advertisements had perceived effectiveness scores ranging from 3.66 to 4.11 (1–5 scale) across three survey rounds.
Conclusions:
Among the campaign audience of Hip Hop–identified youth, awareness of the Fresh Empire campaign was higher in treatment than control markets at individual survey rounds, and increases in campaign awareness were greater in treatment than control markets over time. Campaign advertisements also elicited positive audience reactions. Findings suggest that heavily digital campaigns may take longer to achieve Centers for Disease Control and Prevention–recommended 75% quarterly awareness.
Keywords: Youth, Public health, Tobacco, Mass media, Health promotion
Media campaigns aimed at youth have been effective at reducing prevalence, preventing initiation, and halting progression to established smoking [1–4]. Although youth cigarette smoking among the general population has been declining for the past 20 years [5], smoking prevalence varies among youth subpopulations (e.g., race/ethnicity and identification with various social groups) [6,7].
Over the past 5 years, the U.S. Food and Drug Administration (FDA) developed several tobacco public education campaigns designed to reach youth and young adults at higher risk for tobacco use. Fresh Empire launched in 2015 and is aimed at preventing and reducing tobacco use among 12- to 17-year-old youth who identify with the Hip Hop peer crowd (defined as a macrolevel subculture based on identification with Hip Hop culture and values) [14], have experimented with smoking or are susceptible to future cigarette smoking, and identify as non-Hispanic and black or Hispanic, Asian/Pacific Islander, or multiracial. Hip Hop identification was chosen because it is associated with elevated tobacco use and other risk behaviors [8–13]. Fresh Empire is designed to associate hip hop culture with tobacco-free lifestyles and change perceived norms and beliefs that contribute to tobacco use [13].
Fresh Empire is rooted in contractor Rescue Agency’s Social Branding framework, which uses peer crowds to segment audiences and designs social brands to associate healthy behaviors with desirable lifestyles [14–16]. Peer crowds allow campaign developers to reach audiences through messaging that appeals to the crowd’s sense of identity, values, norms, and beliefs [14]. Fresh Empire’s message strategy [17] encourages youth to reach goals of being successful, attractive, and in control by living tobacco free [13] and aims to counteract typical hip hop imagery of tobacco use [18]. Campaign messages are delivered on campaign media channels and by trusted influencers (e.g., individuals aligned with the peer crowd who have a large, dedicated social media following) and media partners to engage youth.
Historically, large-scale tobacco public education campaigns rely heavily on broadcast (e.g., television) [19]. In addition to using traditional media (e.g., television and print) and hosting local events, Fresh Empire uses a heavily digital and social media marketing strategy, as these tactics allow for tailoring of media to Hip Hop youth using myriad characteristics (e.g., age, location, race/ethnicity, and interests).
Awareness of and receptivity to tobacco public education campaigns are early indicators of campaign success [19–25]. Traditional campaigns such as truth and The Real Cost have achieved more than 75% advertisement awareness when campaigns are on-air [23,24]. According to the Centers for Disease Control and Prevention (CDC), traditional campaigns should reach 75%–85% of their audience for each quarter in a year and are expected to produce awareness at 3–6 months, attitude change at 6–12 months, and behavior change at 12–18 months after launch [19]. CDC states that digital media are promising tools for reaching audiences, but there is not enough evidence to make recommendations regarding the efficacy or optimal means of delivering digital media [19]. Therefore, less is known about timing expectations for reaching expected levels of awareness and change in campaign-targeted outcomes for heavily digital campaigns, such as Fresh Empire, particularly given the approach of reaching Hip Hop (rather than all) youth. Exposure alone is insufficient to ensure processing of messages. Health communication models suggest that for a message to be persuasive, viewers must attend to, understand, and process it [26]. Evaluators use perceived effectiveness as a measure to understand perceived quality and appeal; this measure of receptivity is predictive of attitudes and behaviors [20–22,25].
Fresh Empire launched in four cities in the Southeastern U.S. and expanded to 36 U.S. media markets in October 2015. Fresh Empire uses a randomized treatment control evaluation design comparing campaign-targeted treatment markets (15 of 36 markets receiving the campaign) to control markets with minimal campaign exposure. The campaign is delivered to the 36 media markets via multiple media channels, including digital and social media (using internet protocol addresses and geolocation targeting data from social media platforms), radio, print, out of home (e.g., billboards), and local events. The campaign also includes a national broadcast buy with campaign media delivered in all U.S. markets to maximize cost-efficiency, meaning there is some control market exposure to broadcast advertisements. This study focuses on comparing Fresh Empire awareness between treatment and control markets from baseline (precampaign launch) to approximately 2 years after launch and describes perceived effectiveness of Fresh Empire video advertisements to assess receptivity.
Methods
We selected markets for the Fresh Empire evaluation using a stratified (regional) random selection process (Supplementary Materials). To evaluate the campaign, we use a repeated cross-sectional pre-post data collection design (baseline and five follow-up surveys fielded at approximately 6-month intervals). At each wave, we invite youth who participated in any previous survey to complete follow-up surveys in subsequent data collection rounds and recruit new participants via social media (new cross-sectional sample) to account for attrition. Eligible cross-sectional participants are 12- to 17-year-olds who identify with the Hip Hop peer crowd (defined using a photo selection exercise described in detail under Measures section) residing in one of the 30 evaluation markets. Longitudinal participants meet the same eligibility criteria but are followed to age 18 years in follow-up surveys. For the present study, we used data from four rounds of data collection: baseline (July to November 2015), follow-up 1 (April to June 2016), follow-up 2 (January to June 2017), and follow-up 3 (September 2017 to January 2018).
At baseline and follow-up 1, we used an address-based sampling frame to identify households in evaluation markets likely to have eligible youth. At follow-up 1, when we returned to those households for follow-up interviews, we also screened siblings of those participants for eligibility. Households were mailed screening surveys. Eligible youth identified by mail-based screeners were subsequently visited by field interviewers who obtained parental permission/youth assent and had youth complete the survey on a laptop. The sample was supplemented with 13- to 17-year-old youth recruited via targeted social media advertisements on Facebook and Twitter to complete a Web screener. Field interviewers contacted the parent/guardian of eligible youth recruited via social media for parental permission; youth were sent an email or text with a survey link that they completed after assenting (refer the study by Lee et al. [10] for detailed baseline methods).
For each subsequent survey, we mail a letter to youth who completed a baseline household survey and email or text youth recruited via social media inviting them to complete follow-up surveys. We do not contact youth who are no longer eligible (e.g., >18 years old) or request no future contact. We supplemented the baseline and all follow-up samples with cross-sectional participants recruited via social media advertisements on Facebook and Instagram. Field interviewers secure required parental permission in-person or over the phone. Participants provide assent (if younger than 18 years) or consent (if age 18 years) for surveys and receive $25 for each completed survey. At follow-up 3, participants who participated online (rather than in-person) in the first 2 weeks of data collection received a $5 bonus. The study was approved by RTI International’s Institutional Review Board.
Measures
Outcome variables.
Brand awareness was measured by displaying the campaign logo and asking participants whether they had seen or heard of Fresh Empire in the past 3 months (baseline), 6 months (follow-up 1), or the number of months since the last survey (follow-ups 2 and 3) (“yes,” “no,” and “not sure”). “Yes” responses were aware; “no” and “not sure” were not aware. Brand awareness was assessed before viewing advertisements.
Video advertisement awareness was a combined measure of awareness of advertisements shown in each survey following baseline. Participants watched 30-second Fresh Empire advertisements that aired in months leading up to and during the time each survey was fielded (four advertisements at follow-up 1, five advertisements at follow-up 2, and two advertisements at follow-up 3). After viewing each advertisement, participants rated how frequently (“rarely,” “never,” “sometimes,” “often,” and “very often”) they had seen the advertisement in the past 6 months (follow-up 1) or the number of months since the last survey (follow-ups 2 and 3). Video advertisement awareness was defined based on the highest exposure reported across all advertisements viewed at each survey round, with low defined as “never” or “rarely” maximum frequency of exposure across advertisements, medium defined as “sometimes” maximum frequency of exposure, and high defined as “often” or “very often” maximum frequency of exposure.
Perceived effectiveness for each advertisement shown was assessed using a 6-item scale similar to those from previous research [24,27,28]. Items were assessed on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree) (e.g., “This video is powerful” and “This video is informative”). For each video advertisement, items were combined as scales with high reliability (Table 1).
Table 1.
Perceived effectiveness (PE) of Fresh Empire video advertisements follow-up 1, follow-up 2, and follow-up 3
| PE scale Cronbach’s α | Mean, SD | |
|---|---|---|
|
| ||
| Follow-up 1 video advertisements | ||
| Little Brother | .93 | 4.11, .76 |
| Break Free | .93 | 4.03, .83 |
| We Watch Out | .93 | 4.03, .77 |
| I Got This | .93 | 3.93, .82 |
| Follow-up 2 video advertisements | ||
| Break Free | .94 | 3.97, .81 |
| Sage the Gemini | .92 | 3.82, .80 |
| Lex Lane | .94 | 3.77, .87 |
| J. Reyez | .93 | 3.76, .85 |
| Big Krit | .93 | 3.75, .81 |
| Follow-up 3 video advertisements | ||
| Keep it Moving | .90 | 3.75, .78 |
| Forward | .91 | 3.66, .80 |
SD = standard deviation.
Predictor variables
The first predictor was evaluation market (youth in treatment vs. control markets). The second predictor, time, was defined using survey round (baseline or follow-up 1, 2, or 3).
Covariates and demographics
Demographics measured were age, gender, and race/ethnicity. Tobacco-related covariates were household tobacco use and smoking susceptibility and status [24,29,30]. Hip Hop peer crowd identification (referred to as Hip Hop score) was measured using a photo selection exercise (refer to the studies by Walker et al. [13] and Jordan et al. [8] for detailed description of Rescue Agency’s I-Base Survey measure and example images) that results in a score ranging from −12 to 12. Any positive I-Base score (1–12) indicates Hip Hop identification. For this study, respondents with index scores of 4 or higher were identified as sufficiently influenced by the Hip Hop peer crowd and screened into the survey. The threshold of 4 was determined based on the distribution of scores from formative testing. Hip Hop scores of respondents eligible for the study range from 4 to 12, representing a range of identification with the peer crowd. For analysis, we dichotomized the measure into 4–6 versus 7–12, which was determined based on the distribution of the data. We controlled for misattributed campaign awareness by measuring awareness of other tobacco public education campaign advertisements (truth, Tips From Former Smokers, The Real Cost), defined as “low,” “medium,” “high” awareness using video advertisement awareness definition mentioned previously, and awareness of a fake brand called Digital Youth Against Tobacco. Television, YouTube, social media, and Web site use were included as covariates. Original recruitment source (address-based sampling vs. social media) was included as a covariate.
In addition to individual-level covariates, state and market-level covariates were included in multivariable analyses to control for location-based variation. Variables included state-level tobacco variables [31] and market-level TV household and smoking variables [31,32]. Refer to Supplementary Materials for detail on state- and market-level covariates and additional definition of individual-level covariates.
Analysis
We used descriptive statistics to describe sample characteristics and brand and video advertisement awareness. We used t-tests to examine treatment–control differences in sample characteristics and brand and video advertisement awareness. We employed a logistic difference-in-difference multivariable regression with covariates and a time × treatment interaction term to test for increases in brand awareness by market from baseline to follow-up 3 to determine whether increases in brand awareness were greater in treatment than control over time. We controlled for repeated measures with participant-level clustering. Means and standard deviations of perceived effectiveness are presented for the full sample because all participants viewed and rated advertisements. We completed analyses in 2018 with unweighted panel data using Stata 15.
Results
The baseline and follow-ups 1, 2, and three samples included 2,194, 2,403, 2,050, and 2,090 youth. Youth were distributed approximately equally between treatment and control markets. Participant retention was high, with 72%, 86%, and 69% of follow-ups 1, 2, and 3 samples having participated in a previous round. Black, non-Hispanic youth made up the largest proportion of each sample, followed by Hispanic youth. More than half of each sample reported scores on the lower end of the Hip Hop score range (i.e., 4–6).
The largest proportion of each sample identified as never smokers, not susceptible to smoking, followed by experimenters then never smokers, susceptible to smoking. Current or former smokers made up the smallest proportion of all samples, ranging from 2% at baseline to 5% at follow-up 3. Shifting patterns of smoking over time were expected, given the aging longitudinal sample, as youth susceptibility to and experimentation with smoking increases with age [33]. Few differences emerged in sample characteristics of youth in treatment versus control (Table 2).
Table 2.
Sample characteristics by survey round
| Variables | Baseline sample (N = 2,194), n (%) | Follow-up 1 sample (N = 2,403), n (%) | Follow-up 2 sample (N = 2,050), n (%) | Follow-up 3 sample (N = 2,090), n (%) |
|---|---|---|---|---|
|
| ||||
| Participated in previous wave | 1,721 (71.6) | 1,752 (85.5) | 1,447. (69.2) | |
| Recruitment source | ||||
| Address-based sample | 2,008 (91.5) | 1,711 (71.2) | 1,540 (75.1)** | 1,204 (57.6) |
| Social media | 186 (8.5) | 692 (28.8) | 510 (24.9) | 886 (42.4) |
| Treatment market | 1,167 (53.2) | 1,221 (50.8) | 1,039 (50.7) | 1,065 (51.0) |
| Age (M, SD) | 15.02, 1.54 | 15.77, 1.46 | 16.15, 1.34 | 16.35, 1.13 |
| Female | 1,340 (61.5) | 1,428 (59.8) | 1,289 (63.3) | 1,307 (63.2) |
| Race/ethnicity | ||||
| White, non-Hispanic | 179 (8.2) | 341 (14.2) | 240 (11.7)** | 353 (16.9) |
| Black, non-Hispanic | 1,202 (54.8) | 1,167 (48.6) | 1,024 (50.0) | 904 (43.3) |
| Hispanic | 551 (25.1) | 616 (25.6)** | 534 (25.4) | 564 (27.0) |
| Other/multiracial, non-Hispanic | 262 (11.9) | 279 (11.6)* | 252 (12.3) | 269 (12.9) |
| Household tobacco use | 886 (42.1) | 982 (42.2) | 834 (42.0) | 847 (41.8) |
| Smoking susceptibility and smoking status | ||||
| Never smokers, not susceptible | 1,326 (60.7) | 1,279 (53.8) | 1,070 (52.7) | 1,070 (51.8) |
| Never smokers, susceptible | 385 (17.6) | 419 (17.6) | 276 (13.6) | 223 (10.8) |
| Experimenters | 427 (19.6)* | 591 (24.9) | 614 (30.2) | 675 (32.7) |
| Current or former smokers | 45 (2.1) | 89 (3.7) | 72 (3.5) | 96 (4.7) |
| Hip Hop score | ||||
| 4–6 | 1,252 (57.1) | 1,269 (52.8)* | 1,135 (55.4) | 1,161 (55.6) |
| 7–12 | 942 (42.9) | 1,134 (47.2)* | 915 (44.6) | 929 (44.5) |
| TV use | ||||
| Less than once a day | 127 (5.8) | 826 (34.7) | 656 (32.2) | 818 (39.3) |
| More than once a day | 2,048 (94.2) | 1,555 (65.3) | 1,383 (67.8) | 1,263 (60.7) |
| YouTube use | ||||
| Less than once a day | 359 (16.4) | 1,064 (44.7) | 771 (37.8) | 783 (37.6) |
| More than once a day | 1,827 (83.6) | 1,315 (55.3) | 1,269 (62.2) | 1,298 (62.4) |
| Social media use | ||||
| Less than once a day | 192 (8.8) | 460 (19.3) | 264 (12.9) | 279 (13.4) |
| More than once a day | 1,994 (91.2) | 1,918 (80.7) | 1,775 (87.1) | 1,811 (86.7) |
| Web site use | ||||
| Less than once a day | 236 (10.8) | 927 (39.0) | 713 (35.1) | 741 (35.6) |
| More than once a day | 1,945 (89.2) | 1,448 (61.0) | 1,319 (64.9) | 1,339 (64.4) |
Percentages may not add to 100 because of rounding.
M = mean; SD = standard deviation.
p < .05
p < .01 signify differences between Treatment and Control markets.
As expected, at baseline, awareness of the Fresh Empire brand did not differ between treatment (6%) and control (6%; not significant). At follow-up 1, there was higher brand awareness among youth in treatment (41%) than control (28%), p = .001. At follow-up 2, brand awareness was high across markets and higher among youth in treatment (79%) than control (60%), p = .001. Brand awareness was high across markets at follow-up 3 with higher brand awareness among youth in treatment (85%) than control (68%), p < .001 (Table 3).
Table 3.
Fresh Empire brand and video adverisement awareness in treatment versus control at baseline, follow-up 1, follow-up 2, and follow-up 3
| Baseline sample |
Follow-up 1 sample |
Follow-up 2 sample |
Follow-up 3 sample |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Treatment (N = 1,167) |
Control (N = 1,027) |
p value | Treatment (N = 1,221) |
Control (N = 1,182) |
p value | Treatment (N = 1,039) |
Control (N = 1,011) |
p value | Treatment (N = 1,065) |
Control (N = 1,025) |
p value | |
| n (%) | n (%) | n (%) | n (%) | n (%) | n (%) | n (%) | n (%) | |||||
|
| ||||||||||||
| Brand awareness | 72 (6.2) | 62 (6.1) | .847 | 485 (40.9) | 313 (28.0) | .001 | 799 (79.2) | 564 (60.4) | .001 | 842 (85.0) | 566 (68.3) | .000 |
| Video advertisement awareness | ||||||||||||
| Low | — | — | — | 384 (31.6) | 503 (42.0) | .043 | 317 (30.9) | 432 (38.5) | .215 | 390(36.1) | 606 (52.9) | .000 |
| Medium | — | — | — | 211 (16.7) | 235 (18.3) | .350 | 212 (18.5) | 205 (19.6) | .555 | 278 (26.6) | 231 (23.9) | .169 |
| High | — | — | — | 623 (51.7) | 442 (39.6) | .039 | 508 (50.6) | 373 (41.9) | .166 | 395 (37.3) | 188 (23.2) | .002 |
At follow-up 1, youth in treatment (vs. control) were more likely to report high (52% vs. 40%; p < .05) and less likely to report low video advertisement awareness (32% vs. 42%; p < .05); medium awareness did not differ in treatment (vs. control; 17% vs. 18%, not significant). At follow-up 2, there were no differences between youth in treatment (vs. control) for high (51% vs. 42%; p = .17), medium (19% vs. 20%; p = .56), or low awareness (31% vs. 39%; p = .22). At follow-up 3, youth in treatment (vs. control) were more likely to report high awareness (37% vs. 23%; p = .002). Youth in treatment (vs. control) were also less likely to report low awareness (36% vs. 53%; p < .001), but medium awareness did not differ in treatment (vs. control; 27% vs. 24%; p = .17; Table 3).
The regression used to assess differences in changes in brand awareness between treatment and control over time revealed a time × treatment interaction wherein brand awareness increased more among youth in treatment compared with youth in control between baseline and follow-up 3 (p < .001). Several covariates were associated with brand awareness. Compared with non-Hispanic black youth, non-Hispanic white (p < .001), Hispanic (p = .004), and non-Hispanic other/multiracial youth (p = .010) were less likely to be brand aware. Youth with higher levels of awareness of other tobacco public education campaigns were more likely to be aware of the Fresh Empire brand. Youth with medium truth advertisement awareness (p = .022), medium (p = .031) and high Tips advertisement awareness (p < .001), and medium (p = .024) and high The Real Cost advertisement awareness (p = .010) were more likely to be Fresh Empire brand aware than youth with low awareness of these campaigns’ advertisements. Youth who reported awareness of a fake brand, Digital Youth Against Tobacco (p < .001), were also more likely to be aware of the Fresh Empire brand, suggesting overreporting of Fresh Empire brand awareness (full results in Table 4). Note that awareness levels of the fake brand were similar between treatment and control at all waves, with the exception of higher awareness in control (23%) than treatment (18%) at follow-up 2 (p < .05).
Table 4.
Differences in changes in Fresh Empire brand awareness over time
| Variables | OR (95% CI) | p value |
|---|---|---|
|
| ||
| Treatment × time | 3.26 (1.90–5.58) | <.001 |
| Age | 1.01 (.94–1.09) | .735 |
| Female | 1.09 (.89–1.33) | .399 |
| Race/ethnicity | ||
| White, non-Hispanic | .42 (.28–.64) | <.001 |
| Black, non-Hispanic | REF (–) | – |
| Hispanic | .64 (.47–.87) | .004 |
| Other/multiracial, non-Hispanic | .60 (.50–.91) | .010 |
| Household tobacco use | 1.07 (.89–1.30) | .473 |
| Smoking susceptibility and smoking status | ||
| Never smokers, not susceptible | REF (–) | – |
| Never smokers, susceptible | .93 (.68–1.29) | .679 |
| Experimenters | 1.04 (.83–1.31) | .723 |
| Current or former smokers | .61 (.36–1.03) | .062 |
| Hip Hop score: 7 to 12 | 1.01 (.84–1.22) | .911 |
| Recruitment source: social media | .43 (.31–.60) | <.001 |
| Truth advertisement awareness | ||
| Medium | 1.37 (1.05–1.80) | .022 |
| High | 1.25 (.98–1.58) | .069 |
| Tips advertisement awareness | ||
| Medium | 1.34 (1.03–1.74) | .031 |
| High | 1.85 (1.36–2.52) | <.001 |
| The Real Cost advertisement awareness | ||
| Medium | 1.38 (1.04–1.81) | .024 |
| High | 1.41 (1.09–1.84) | .010 |
| Aware of digital youth against tobacco brand | 2.15 (1.59–2.89) | <.001 |
| TV use: more than once a day | 1.17 (.93–1.48) | .174 |
| YouTube use: more than once a day | 1.01 (.82–1.25) | .911 |
| Social media use: more than once a day | 1.26 (.93–1.70) | .133 |
| Web site use: more than once a day | .90 (.72–1.11) | .324 |
| State cigarette tax (2012) | 1.00 (1.00–1.00) | .714 |
| Per capita tobacco control expenditures | 1.00 (.99–1.01) | .673 |
| Total tobacco control expenditures | 1.00 (1.00–1.00) | .336 |
| Percentage of black TV households | .99 (.97–1.00) | .061 |
| Percentage of Hispanic TV households | .99 (.97–1.01) | .267 |
| Total TV households | 1.00 (.99–1.00) | .001 |
| DMA smoking prevalence | .97 (.92–1.02) | .222 |
Reference category for Hip Hop score is 4–6. Reference category for recruitment source is address-based sample.
Reference category for truth, Tips, and The Real Cost advertisement awareness is low. Reference category for TV, YouTube, Social Media, and web site use is less than once a day.
CI = confidence interval; DMA = designated market area; OR = odds ratio.
Fresh Empire video advertisements elicit positive reactions, with perceived effectiveness scores ranging between 3.66 and 4.11 across survey rounds (Table 1).
Discussion
Fresh Empire, one of the first FDA tobacco public education campaigns for a hard-to-reach population, is effectively reaching its audience of Hip Hop–identified youth, sustaining awareness over time, and eliciting positive reactions. Fresh Empire brand awareness is high and has steadily increased since launch. Brand awareness has been consistently higher in treatment than control markets at each survey round since launch (follow-ups 1, 2, and 3). Youth awareness of the Fresh Empire brand also increased more in treatment than control between baseline and follow-up 3, approximately 2 years after launch. Similarly, we observed higher levels of Fresh Empire video advertisement awareness among youth in treatment than control at follow-ups 1 and 3, but no differences at follow-up 2.
Fresh Empire video advertisements were positively received by youth at each round, with the highest perceived effectiveness ratings for advertisements assessed at follow-up 1. Scores are similar to those observed for The Real Cost campaign advertisements, which have been shown to be effective at preventing youth smoking initiation [3,24].
Although our study detected treatment–control differences in awareness, youth in control markets were exposed to the campaign at moderate to high levels, which may partially explain the lack of follow-up 2 treatment–control differences in video advertisement awareness. A national broadcast media buy was included as part of the campaign media strategy. We recognized that this would result in some control exposure but accepted this trade-off because national broadcast buys are more cost-efficient than local buys. Although the expectation was that this delivery would be evenly distributed across markets, the broadcast delivery fell more heavily than expected in several control markets, resulting in less treatment–control contrast in campaign awareness than anticipated. Moderate to high exposure across markets underscores broad campaign reach and highlights reasons to explore campaign effects using multiple exposure metrics, such as through self-reported awareness or media delivery (e.g., target rating points). In future analyses, we will use measures of media delivery and exposure for digital, social, and broadcast media as another measure of exposure. Target rating points reflect potential campaign exposure among the campaign audience in the market where youth reside and, because they are external to the individual, are not subject to selective attention bias, which is a limitation of self-reported awareness [34].
Campaign exposure is a necessary first step before achieving changes in tobacco-related attitudes and beliefs targeted by Fresh Empire. In future analyses, we will use multiple exposure metrics to examine campaign effects on key attitudes and beliefs (e.g., perceived smoking risks, perceived norms, and attitudes about tobacco-free lifestyles). As we examine Fresh Empire’s impact, it is important to control for exposure to other tobacco-related media campaigns, including campaigns such as truth and The Real Cost. Future research could examine whether exposure to multiple campaigns has a synergistic effect on outcomes.
Past studies of tobacco public education campaigns show that they can reach and engage a variety of different types of youth, although salience of messages may vary across groups [35]. For instance, the truth campaign’s strongest effects on attitudes and intentions were for black youth [36]. Fresh Empire represents the first large-scale application of a peer-crowd strategy, suggesting that a federal government campaign can be successful at reaching and resonating with a hard-to-reach peer crowd of youth at higher smoking risk.
The Fresh Empire evaluation also provides key insights for establishing benchmarks for digital campaign evaluation. CDC guidance for mass media campaigns was developed based on broad reach, primarily broadcast campaigns using different methods to assess campaign awareness (i.e., confirmed awareness assessed via telephone interviews) [19] and is not well-suited for measuring success of heavily digital campaigns focused on peer crowds at greater risk for smoking evaluated using online surveys, such as Fresh Empire. Despite its differences from traditional mass reach campaigns, Fresh Empire awareness was at 79% in treatment markets approximately 1.5 years after campaign launch and 85% approximately 2 years after launch. These awareness levels suggest that heavily digital campaigns may take longer to achieve the 75% awareness level recommended by the CDC for the first year of campaigns [19], which makes sense, given an increasingly fragmented media market and changing use patterns.
Limitations
Several limitations are important to acknowledge. First, measurements used to assess awareness involve showing campaign logo and advertisements to participants. Although consistent with procedures from previous evaluations [24,37], this may introduce overreporting and could have also caused panel bias among participants who completed surveys at previous waves (i.e., awareness influenced by participants’ assessment of awareness at previous survey waves) [38]. We did not include a measure to assess whether participants recognized the Fresh Empire brand as a tobacco public education brand. It is possible that brand-aware participants did not know what it represented. Although we observed steady brand awareness increases since launch, we were unable to assess whether video advertisement awareness increases with time because advertisement content is continually refreshed. Third, campaign exposure was moderate to high in control markets, which makes our ability to detect effects more challenging and increases reliance on additional exposure metrics (e.g., potential exposure to paid media) to assess effects. Fourth, because there is no known sampling frame for Hip Hop–identified youth to allow for a probability-based sample, findings may not generalize to all Hip Hop–identified youth. Finally, it is possible that reference to the Fresh Empire campaign in some follow-up recruitment materials and higher exposure to the campaign in treatment may have influenced some participants’ responses to survey items related to the campaign and tobacco or their decision to continue participating, although attrition differed between treatment and control at follow-up 2 only (p < .05).
Conclusion
Fresh Empire is one of the first FDA campaigns designed to reach peer crowds at greater risk for smoking using a heavily digital strategy that allows the campaign to reach Hip Hop–identified youth. Fresh Empire has achieved high awareness overall and higher awareness in campaign-targeted treatment than control markets. Fresh Empire advertisements also elicited positive reactions among this audience.
Supplementary Material
Supplementary Data
Supplementary data related to this article can be found at https://doi.org/10.1016/j.jadohealth.2019.09.005.
IMPLICATIONS AND CONTRIBUTION.
This study presents findings from Fresh Empire, one of the U.S. Food and Drug Administration’s first campaigns to prevent and reduce tobacco use among a hard-to-reach population: Hip Hop–identified youth who are primarily black, Hispanic, Asian/Pacific Islander, or multiracial. Findings demonstrate that media campaigns can effectively reach and appeal to this population.
Acknowledgments
The authors would like to thank the Office of Health Communication and Education at the U.S. Food and Drug Administration/Center for Tobacco Products and Rescue Agency for their invaluable work on the development of Fresh Empire.
Funding Sources
Funding for this work was provided by FDA Contract HHSF223201310001B and HHSF223201610032I.
Footnotes
Conflicts of interest: The authors have no conflicts of interest to disclose.
Disclaimer: This publication represents the views of the author(s) and does not represent U.S. Food and Drug Administration/Center for Tobacco Products position or policy.
References
- [1].Farrelly MC, Davis KC, Haviland ML, et al. Evidence of a dose-response relationship between “truth” antismoking ads and youth smoking prevalence. Am J Public Health 2005;95:425–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Farrelly MC, Nonnemaker J, Davis KC, et al. The influence of the national truth campaign on smoking initiation. Am J Prev Med 2009;36:379–84. [DOI] [PubMed] [Google Scholar]
- [3].Farrelly MC, Duke JC, Nonnemaker J, et al. Association between the real cost media campaign and smoking initiation among youths - United States, 2014–2016. MMWR Morb Mortal Wkly Rep 2017;66:47–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].National Cancer Institute US. The role of the media in promoting and reducing tobacco use: Executive summary. In: Davis RM, Gilpin EA, Loken B, et al. , eds. Tobacco control monograph no 19: The role of the media in promoting and reducing tobacco use, iii–22. 2008. [Google Scholar]
- [5].Johnston LD, Miech RA, O’Malley PM, et al. Monitoring the future national survey results on drug use: 1975–2017: Overview, key findings on adolescent drug use. Ann Arbor, MI: Institute for Social Research, The University of Michigan; 2018. [Google Scholar]
- [6].Kasza KA, Ambrose BK, Conway KP, et al. Tobacco-product use by adults and youths in the United States in 2013 and 2014. N Engl J Med 2017;376:342–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Odani S, Armour BS, Agaku IT. Racial/ethnic disparities in tobacco product use among middle and high school students - United States, 2014–2017. MMWR Morb Mortal Wkly Rep 2018;67:952–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Jordan JW, Stalgaitis CA, Charles J, et al. Peer crowd identification and adolescent health behaviors: Results from a statewide representative study. Health Educ Behav 2019;46:40–52. [DOI] [PubMed] [Google Scholar]
- [9].Lee YO, Jordan JW, Djakaria M, et al. Using peer crowds to segment Black youth for smoking intervention. Health Promot Pract 2014;15:530–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Lee YO, Curry LE, Fiacco L, et al. Peer crowd segmentation for targeting public education campaigns: Hip Hop youth and tobacco use. Prev Med Rep 2019;14:1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Lisha NE, Jordan JW, Ling PM. Peer crowd affiliation as a segmentation tool for young adult tobacco use. Tob Control 2016;25:i83–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Sussman S, Pokhrel P, Ashmore RD, et al. Adolescent peer group identification and characteristics: A review of the literature. Addict Behav 2007;32:1602–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Walker MW, Navarro MA, Hoffman L, et al. The Hip Hop peer crowd: An opportunity for intervention to reduce tobacco use among at-risk youth. Addict Behav 2018;82:28–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Moran MB, Walker MW, Alexander TN, et al. Why peer crowds matter: Incorporating youth subcultures and values in health education campaigns. Am J Public Health 2017;107:389–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Fallin A, Neilands TB, Jordan JW, et al. Social branding to decrease lesbian, gay, bisexual, and transgender young adult smoking. Nicotine Tob Res 2015;17:983–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Ling PM, Lee YO, Hong J, et al. Social branding to decrease smoking among young adults in bars. Am J Public Health 2014;104:751–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].U.S. Department of Health and Human Services. Fresh empire. Available at: https://freshempire.betobaccofree.hhs.gov/. Accessed March 19, 2019. [Google Scholar]
- [18].DuRant RH, Rome ES, Rich M, et al. Tobacco and alcohol use behaviors portrayed in music videos: A content analysis. Am J Public Health 1997;87:1131–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Centers for Disease Control and Prevention and Health Promotion Office on Smoking on Health. Best practices for comprehensive tobacco control programs—2014. Atlanta, GA: Centers for Disease Control and Prevention; 2014. [Google Scholar]
- [20].Davis KC, Duke J, Shafer P, et al. Perceived effectiveness of antismoking ads and association with quit attempts among smokers: Evidence from the tips from former smokers campaign. Health Commun 2017;32:931–8. [DOI] [PubMed] [Google Scholar]
- [21].Dillard JP, Weber KM, Vail RG. The relationship between the perceived and actual effectiveness of persuasive messages: A meta-analysis with implications for formative campaign research. J Commun 2007;57:613–31. [Google Scholar]
- [22].Duke JC, Allen JA, Eggers ME, et al. Exploring differences in youth perceptions of the effectiveness of electronic cigarette television advertisements. Nicotine Tob Res 2016;18:1382–6. [DOI] [PubMed] [Google Scholar]
- [23].Farrelly MC, Healton CG, Davis KC, et al. Getting to the truth: Evaluating national tobacco countermarketing campaigns. Am J Public Health 2002;92:901–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Duke JC, Alexander TN, Zhao X, et al. Youth’s awareness of and reactions to the real cost national tobacco public education campaign. PLoS One 2015;10:e0144827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Noar SM, Barker J, Bell T, et al. Does perceived message effectiveness predict the actual effectiveness of tobacco education messages? A systematic review and meta-analysis. Health Commun 2018:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Petty RE, Cacioppo JT. The elaboration likelihood model of persuasion. Adv Exp Soc Psychol 1986;19:123–205. [Google Scholar]
- [27].Davis KC, Nonnemaker JM, Farrelly MC, et al. Exploring differences in smokers’ perceptions of the effectiveness of cessation media messages. Tob Control 2011;20:26–33. [DOI] [PubMed] [Google Scholar]
- [28].Davis K, Farrelly M, Duke J, et al. Antismoking media campaign and smoking cessation outcomes, New York State, 2003–2009. Prev Chronic Dis 2012;9:E40. [PMC free article] [PubMed] [Google Scholar]
- [29].Mowery PD, Farrelly MC, Haviland ML, et al. Progression to established smoking among US youths. Am J Public Health 2004;94:331–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Pierce JP, Choi WS, Gilpin EA, et al. Validation of susceptibility19 as a predictor of which adolescents take up smoking in the United States. Health Psychol 1996;15:355–61. [DOI] [PubMed] [Google Scholar]
- [31].U.S. Census Bureau, 2012 State Government Tax Dataset, Annual Survey of State Government Tax Collections, 2012. Available at: https://www.census.gov/data/datasets/2012/econ/stc/2012-annual.html.
- [32].Nielsen. 2012–2013 Nielsen TV household data. Nielsen TV household data. 2012–2013. [Google Scholar]
- [33].U.S. Department of Health and Human Services. The health consequences of smoking—50 years of progress: A report of the surgeon general. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2014. [Google Scholar]
- [34].Davis KC, Patel D, Shafer P, et al. Association between media doses of the Tips from Former Smokers campaign and cessation behaviors and intentions to quit among cigarette smokers, 2012–2015. Health Educ Behav 2018;45:52–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].U.S. National Cancer Institute. A socioecological approach to addressing tobacco-related health disparities. National Cancer Institute Tobacco Control Monograph 22. Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of Health; 2017. [Google Scholar]
- [36].Cowell AJ, Farrelly MC, Chou R, et al. Assessing the impact of the national ‘truth’ antismoking campaign on beliefs, attitudes, and intent to smoke by race/ethnicity. Ethn Health 2009;14:75–91. [DOI] [PubMed] [Google Scholar]
- [37].Duke JC, Davis KC, Alexander RL, et al. Impact of a U.S. antismoking national media campaign on beliefs, cognitions and quit intentions. Health Educ Res 2015;30:466–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Warren JR, Halpern-Manners A. Panel conditioning in longitudinal social science surveys. Sociol Methods Res 2012;41:491–534. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.