The Role of Emotions and Perceived Ad Effectiveness: Evidence From the Truth FinishIt Campaign

Abstract

Purpose:

Examine association between emotional valence and intensity prompted by anti-tobacco advertising messages and perceived ad effectiveness among youth/young adults.

Design:

Online forced-exposure survey data from a nationally weighted, cross-sectional sample of youth/young adults, collected periodically over a 4-year period.

Setting:

National.

Participants:

Thirty-seven cross-sectional surveys conducted online from June 2015 to January 2018; total N = 9534. All participants, aged 15 to 21, were in the intervention; no control group.

Intervention:

Individuals participating in premarket testing of truth ads were forced exposed to one of 37 anti-tobacco ads.

Measures:

Emotional response, emotional intensity, and perceived ad effectiveness. Emotional response has been previously studied and measured. Including the discrete measure of “concerned” in positive emotions is unique to our study. It patterned with the other positive emotions when each ad was examined by each emotion. Intensity as measured in this study through the 5-point scale (“how much does this ad make you feel”) is unique in the anti-tobacco ad literature. Although several past studies ranked the degree of emotion elicited by ads, they have not incorporated the intensity of emotion as reported by the participant themselves. The scale was used to determine whether perceived ad effectiveness is similar to those used in previous studies.

Analysis:

Linear regressions were estimated to assess type of emotional sentiment and level of intensity in relation to perceived effectiveness of the message.

Results:

All 9534 participants were exposed; no control group. The βs indicate how strongly the emotion variable influences the study outcome of perceived ad effectiveness. Positive emotions (β = .76) were more highly associated with perceived ad effectiveness (β = .06). Higher intensity with positive emotional sentiment and high-intensity negative produced the highest scores for perceived ad effectiveness (β = .30).

Conclusion:

Eliciting a positive, high-impact emotional response from viewers can help improve perceived effectiveness, and in turn, overall ad effectiveness.

Keywords

tobacco prevention mass media campaigns adolescents young adults

Purpose

Despite recent declines in smoking, tobacco-related disease remains the leading cause of preventable mortality in the United States, causing over 480 000 deaths per year.¹ To combat tobacco’s deadly toll, anti-smoking mass media messages have been deployed in an effort to help youth and young adults reject tobacco. These campaigns have proven to be an effective and extremely efficient strategy to prevent tobacco use among young people. Since 2014, 3 large, national mass media campaigns have aired (truth, The Real Cost, and TIPS From Former Smokers) to reduce tobacco use initiation among youth and prompt cessation among current smokers.^2
-4

A growing body of literature has identified characteristics that make anti-smoking media campaigns most effective.⁵ Messages containing negative health consequences, notably those with emotional and graphic content, have been found to be effective, particularly among adults.⁶ Among youth, ad effectiveness relies on a number of factors. Ads that include “personal testimonials, surprising narratives, intense images, sound, and editing,” or visceral, negative themes are most likely to increase recall, engagement, and discussion.^7,8 Prompting an emotional response in both adults and youth can lead to greater persuasion of the intended message and increase an ad’s perceived effectiveness.^5,7,9

Emotional theory posits that behavior is influenced by motivation that is linked to a specific emotion.¹⁰ Viewing an advertisement may provoke an emotion through the use of imagery or messaging, which could be positive or negative.¹¹ Fear emotions can encourage people to move away from or change behavior to avoid a threat, such as quitting smoking to prevent death and disease.¹² Conversely, emotional theory research on adult response to alternative tobacco product advertisements showed that a hopeful message can motivate a problem-solving action, such as switching to a reduced harm tobacco product.¹⁰

Past research examining the role of emotion in anti-tobacco advertisements targeted at youth and young adults has focused largely on the role of negative emotions, finding that those evoking strong negative emotions, such as sadness, anger, and fear, are the most successful.^5,13,14 Biener et al concluded that “anti-tobacco ads perceived as most effective by youth were those that evoked strong negative emotions.”⁵ There is also evidence that effective prevention and cessation advertisements are those with stronger messages, which provoke greater emotional response.¹⁴ Chung et al studied the effect of high and low fear antismoking messages among college students and found that intensity of emotion played a role in the efficacy of the message, such that the stronger the fear, the lower intention to smoke.¹⁴

Less studied is the role of positive emotions, such as hope, in tobacco prevention messages, particularly among those targeted at young audiences. Bigsby et al studied how positive emotions related to cessation among adults and found that anti-tobacco advertisements with a high rating on hope and pride were associated with greater intentions to quit smoking.¹⁵ Prior research has also suggested that anti-tobacco messages with informative or positive messages can produce feelings of hope, which may lead to a change in behavior among adults.^15,16

Although the literature is mixed on the type of emotions most likely to increase perceived ad effectiveness, both positive and negative emotions appear to play a role in how anti-tobacco advertisements are perceived, as does the strength of the sentiment. This study is aimed at understanding the role emotions have on perceived ad effectiveness among a nationally representative sample of young people who were exposed to a subset of national anti-tobacco advertisements. We hypothesize that the relationship between positive emotions and perceived ad effectiveness will vary by the level of intensity of negative emotions. To determine this, emotional sentiment, both positive and negative, was assessed, as was the intensity of emotional arousal.

Methods

In 2014, the national anti-tobacco “truth” campaign launched a new campaign, FinishIt, which focused on a slightly older target audience of youth and young adults, aged 15 to 21, and encouraged young people to join a social movement to end smoking. This study examines responses to 37 distinct advertising executions that were tested over the past 4 years.

Design and Sample

This study uses the data from a total of 37 cross-sectional surveys conducted online via the Survey Sampling International panel from June 2015 through January 2018. Individuals were between the ages of 15 and 21 and participated in 1 premarket testing of a truth ad during this time frame (ie, testing of concepts and advertisements prior to airing). A total of 200 to 300 unique respondents per ad made up the total sample of 9534 participants. All study protocols were reviewed and approved for human participant research by the institutional review board of Chesapeake (protocol no. 00009656).

Intervention

The goal of these cross-sectional premarket surveys was to examine the effectiveness of ads to (1) convey factual information, (2) ensure message comprehension, (3) positively shift anti-tobacco attitudes and/or intentions, (4) prompt perceived ad effectiveness, and (5) build brand equity. Study participants completed a baseline assessment, which included their attitudes toward smoking, smoking intentions, and smoking behaviors, after which they were randomized to view a near final version of one truth advertisement (video). Throughout the survey, the participant viewed the same truth ad 3 times. After each forced exposure to the ad, they were asked a series of questions related to message communication and emotion. The survey took approximately 25 minutes to complete.

Measures

Demographic variables included respondents’ age in years (15-17, 18-21), gender (male, female), race/ethnicity (white, non-Hispanic; black, non-Hispanic; Hispanic; Other, non-Hispanic), and educational attainment (less than high school, high school graduate/General Education Development [GED], some college/associate’s degree/bachelor’s degree +).

After viewing the advertisement, respondents were asked to identify the emotions they experienced as well as the intensity. Survey items asked “how much does this ad make you feel” for each of the following emotions: sad, angry, irritated, and afraid, which were categorized as negative emotions, and happy, powerful, inspired, hopeful, confident, motivated, and concerned, which were categorized as positive emotions. Concerned was included with positive emotions because it patterned with the other positive emotions when each ad was examined by each emotion. Negative emotions were categorized to include sad, angry, irritated, and afraid. Intensity of discrete emotions as measured in this study through the 5-point scale (“how much does this ad make you feel”) is unique in the anti-tobacco advertisement literature. Although several past studies have ranked the degree of emotion elicited by ads, they have not incorporated the intensity of emotion as reported by the participant themselves.^5,17

To assess emotional intensity, respondents could select (1) “not at all,” (2) “a little,” (3) “some,” (4) “a lot,” or (5) “a great deal” for each emotion. If an individual scored a 4 or 5 (“a lot” or “a great deal”) on any of the items, they were categorized as “high intensity” for that particular emotion. Likewise, individuals who indicated <3 (“some,” “a little,” “not at all”) were categorized as “low intensity” on that particular emotion. For those respondents who indicated no emotional response by reporting 1 (“not at all”) for all items, responses were recoded to missing because the models were meant to examine the relationship between emotional response and perceived ad. Conceptually, having no emotional reaction could not be classified as either positive or negative and proves difficult to interpret.

Perceived ad effectiveness was assessed using the following: (4) items about the ad: (1) “it captured my attention,” (2) “it was informative/educational,” (3) “it was convincing,” and (4) “it gave me good reasons not to smoke.” Response choices were on a 5-point Likert scale, ranging from (1) “strongly disagree,” (2) “disagree,” (3) “neither agree or disagree,” (4) “agree,” to (5) “strongly agree.” The variables were summed and averaged to create a mean scale. The resulting mean scale was used as a continuous outcome in the models. This scale is similar to those used in previous studies.^18

-21

Statistical Analysis

All analyses were conducted using SAS Enterprise Guide 7.1.²² Bivariate analyses were used to understand the distribution of emotions with respect to perceived ad effectiveness. Ordinary least square linear regression was estimated to examine the relationships between emotional responses and perceived ad effectiveness. An additional model was estimated to examine whether the relationship between positive emotions and perceived ad effectiveness varied by intensity of negative emotion. All models are controlled for age, gender, race, and education.

Results

Males and females were equally represented (51% vs 49%). Participants were older (57% 18-21-year-olds), 53% white, 16% blacks/African American, and 19% Hispanic. Most of the participants (41%) had not yet completed high school, 24% had completed high school or a GED, 32% had completed some college/associates degree, and 3% had completed a bachelor’s degree or higher (see Table 1).

Table 1.

Total Sample Demographics (N = 9534).

Variable	n (%)
Age
15-17	4099 (42.99)
18-21	5435 (57.01)
Gender
Male	4827 (50.63)
Female	4707 (49.37)
Race/ethnicity^a
White, non-Hispanic	5099 (53.48)
Black, non-Hispanic	1498 (15.71)
Hispanic	1838 (19.28)
Other, non-Hispanic	1041 (10.92)
Education
Less than high school	3921 (41.13)
High school degree/GED	2260 (23.7)
Some college/associate’s degree	3028 (31.76)
Bachelor’s degree +	325 (3.41)

Abbreviation: GED, General Education Development.

^aPercentages do not sum to 100% due to missing responses on this variable.

Models were developed in 2 phases. Variables for low- and high-intensity positive emotion and low- and high-intensity negative emotion were included in the main-effects model controlling for age, race, gender, and education (Table 2). Results indicated that individuals who reported higher intensity positive emotions toward ads, compared to those who reported lower intensity positive emotions, had higher mean scores on perceived ad effectiveness (β = .76; P = <.001). Likewise, individuals reporting higher intensity negative emotions toward ads, compared to those reporting lower intensity negative emotions, had slightly higher mean scores on perceived ad effectiveness (β = .06; P < .01).

Table 2.

Main-Effect OLS Regression Models Examining the Relationship Between Emotions and Perceived Ad Effectiveness (N = 6680).^a

Variable	Unstandardized β (SE)	Standardized β
Positive emotions
High intensity	.76 (.02)^b	.43
Low intensity	Ref	Ref
Negative emotions
High intensity	.06 (.02)^c	.04
Low intensity	Ref	Ref
Age
15-17	.13 (.03)^b	.08
18-21	Ref	Ref
Gender
Female	.14 (.02)^b	.02
Male	Ref	Ref
Race/ethnicity
Black, non-Hispanic	.06 (.03)^d	.03
Hispanic	−.03 (.02)	−.02
Other, non-Hispanic	−.06 (.03)^d	−0.02
White, non-Hispanic	Ref	Ref
Education
Less than high school	.04 (.05)	.03
High school degree/GED	.03 (0.05)	.02
Some college/associate’s degree	.03 (.02)	.02
Bachelor’s degree +	Ref	Ref

Abbreviations: GED, General Education Development; SE, standard error.

^a Adj. R ² = 0.22. F ₍₁₀₎ = 185.20, P < .001. The total effects of education are not significant.

^b P < .001.

^c P < .01.

^d P < .05.

To understand how negative and positive emotions work together to effect ads based on reasons not to smoke, a second model examining the effect of the interaction between positive and negative emotions was assessed. This model also controlled for age, race, gender, and education (Table 3). Figure 1 depicts the significant interaction between positive and negative emotions on the outcome perceived ad effectiveness (overall effect: F ₍₁₎ = 41.12, P < .001). Figure 1 looks at how the relationship between intensity of positive emotions varies by level of intensity of negative emotions. At high-intensity positive emotions, perceived ad effectiveness was the highest for those who also had high-intensity negative emotions. Perceived ad effectiveness was similar for those at high-intensity positive emotions who also had low-intensity negative emotions. Conversely, at low-intensity positive emotions, perceived ad effectiveness was the lowest for those who have high-intensity negative emotions. Perceived ad effectiveness was slightly higher for those at low-intensity positive emotions who also had low-intensity negative emotions.

Table 3.

OLS Regression Models Examining the Interaction Effect Between Emotions and Relationship With Perceived Ad Effectiveness (N = 6680).^a

Variable	Unstandardized β (SE)	Standardized β
Positive emotions
High intensity	.68 (.02)^b	.39
Low intensity	Ref	Ref
Negative emotions
High intensity	−.19 (.04)^c	−.11
Low intensity	Ref	Ref
Positive emotions × negative emotions
High-intensity positive, high-intensity negative	0.30 (0.04)^b	.18
High-intensity positive, low-intensity negative	Ref	Ref
Low-intensity positive, high-intensity negative	Ref	Ref
Low-intensity positive, low-intensity negative	Ref	Ref
Age
15-17	.13 (.08)^b	.08
18-21	Ref	Ref
Gender
Female	.13 (.02)^b	.08
Male	Ref	Ref
Race/ethnicity
Black, non-Hispanic	.06 (.03)^d	.02
Hispanic	−.03 (.02)	−.03
Other, non-Hispanic	−.06 (.03)^d	−.02
White, non-Hispanic	Ref	Ref
Education
Less than high school	.04 (.05)	.02
High school degree/GED	.02 (.05)	.01
Some college/associate’s degree	.03 (.05)	.02
Bachelor’s degree +	Ref	Ref

Abbreviations: GED, General Education Development; SE, standard error.

^aAdj. R ² = 0.22. F ₍₁₁₎ = 173.11, P < .001. The total effects of education are not significant.

^b P < .001.

^c P < .01.

^d P < .05.

Figure 1.

Two-way interaction of positive high intensity and negative high intensity of perceived ad effectiveness.

To further explore the significant interaction effects, we estimated post hoc tests that included looking at (a) the simple effects (based on least square means) of intensity of positive emotion across the levels of intensity of negative emotion (and vice versa) and (b) comparisons of the simple effects.²³ In the first step, we estimated the simple effects (least square means) and found that that there are significant simple effects of positive emotions within the levels of negative emotions, as well as significant simple effects of negative emotions within the levels of positive emotions. For example, high-intensity positive emotions and low-intensity positive emotions differ on perceived ad effectiveness for both low-intensity negative emotions (estimate = 0.68) and high-intensity negative emotions (estimate = 0.99). Likewise, low-intensity negative emotions and high-intensity negative emotions differ on perceived ad effectiveness for both low-intensity positive emotions and high-intensity positive emotions. In the second step, we wanted to test whether the simple effect of intensity of positive emotions at low-intensity negative emotions is the same as the effect of intensity of positive emotions at high-intensity negative affect (and vice versa). In both cases, the estimate was the same (estimate = 0.30), indicating that all simple effects were different from one another.

Discussion

The current study finds that both positive and negative emotions are associated with perceived ad effectiveness. In this study, it is the combination of high-intensity positive and high-intensity negative emotions that best predicts perceived ad effectiveness. However, just eliciting negative emotions is not sufficient to elicit perceived ad effectiveness. Findings indicate that those ads that elicit high-intensity negative emotions with low-intensity positive emotions have the lowest mean scores for perceived ad effectiveness. If only one emotion is elicited, high-intensity positive was found to elicit more favorable responses than high-intensity negative emotions in relation to perceived ad effectiveness.

Previous studies find varying results with respect to the role of emotional response in relation to ad-related outcomes.^{5

-16} The current study findings are inconsistent with previous studies, which argue for either positive or negative emotions as the most effective prevention strategy among youth. This study determines that it is the combination of both positive and negative emotions that best predicts ad effectiveness in this age group. This could reflect a variety of factors, including the differences in the generational audiences over time, the expanded definitions of emotions in this study, and/or the changing media landscape. Data for this study were collected from 15- to 21-year-olds from 2015 through 2018. This current cohort of youth is very different as a generation than the rebellious Generation X, the target audience of the original truth campaign in 2000. The current target audience of the campaign are known as Millennials and Generation Z. They are less rebellious than the previous generation and have an interest in and enthusiasm for social change and passion for social justice. What motivated a rebellious generation may not be the most effective tactics for Millennials and Generation Z. Given the generational differences and the addition of motivating and inspiring to the possible range of positive emotions within the study, it is not surprising that the combination of high-intensity positive and high-intensity negative emotions provides the ideal mix to change smoking intentions. Although more negative emotional appeals worked without positive emotions for previous generations, findings indicate that this approach may be less than ideal for today’s younger generations who want to be a part of something bigger than themselves.

Employing the concept of joining an anti-tobacco social movement was used to harness the current interest and enthusiasm young adults have for being involved in social causes. Thus, the emotional factors of “motivated” and “inspired” were added into the positive emotions definition. To date, studies have limited their examination of emotions to include happiness, hope, anger, fear, and sadness.^{5,10,13
-15,24} This study broadens the emotional portfolio of responses to anti-tobacco media messages, which can help provide a more nuanced understanding of factors associated with perceived ad effectiveness. The concept of no emotional intensity (answering a 1 on the 5-point scale) cannot be classified as either negative or positive emotion and therefore needs further exploration. The authors chose to drop these cases because it is so difficult to interpret and to determine where emotions fall on a continuum of high to low. A model was run, which included the “no emotions” category in order to test the outcome with the 2820 participants, indicating “no emotion” on the positive or negative or both intensity scales and findings did not change. It would be useful to better understand the concept of the “no emotion category.” This would require additional testing, which is outside the scope of these data.

Given the significant changes in the media landscape, which now includes numerous media channels competing for the attention of youth and young adults, anti-tobacco media messages need to be able to break through a very cluttered media environment to influence behavior. The results of this study suggest that anti-tobacco ads should continue to use relevant topics and an inspirational tone to prompt positive and negative emotions, and in turn, help shift smoking-related attitudes and behaviors.

Measuring not only the type but also the self-reported intensity of discrete emotions was central to the analysis of this study, allowing for insights into the level of emotion necessary for changes in perceived ad effectiveness. To our knowledge, this is the first analysis to assess intensity of discrete emotions with respect to anti-tobacco perceived ad effectiveness. Although several past studies have ranked the intensity of emotion as measured by television gross rating points, they have not incorporated the intensity of emotion as reported by the participant themselves (how much an emotion is felt).^5,17 This measurement enabled the unique finding that while positive emotions do support changes in attitude, it is the self-reported intensity that matters.

This study is not without limitations. Methodology included forced exposure to the advertisements using cross-sectional surveys, which lack real-world context and may impact emotional response patterns. Advertisements viewed in isolation also lack the reinforcing effect of social media and the surrounding content in the television media setting. Nonetheless, this study provides important evidence that eliciting both positive and negative emotional responses for young people to reject tobacco can help improve ad effectiveness. Future research should examine emotional response in a longitudinal study to assess the long-term effect of emotional intensity on tobacco-related behaviors.

So What?

What is already known on this topic?

Prior research has identified characteristics that make anti-smoking media campaigns most effective. Among adults, emotionally and graphically charged negative health consequences, have been found to be effective, while personal testimonials, narratives, and visceral, negative themes are effective among youth. Overall, prompting an emotional response in both adults and youth leads to greater persuasion of the intended message and an increase in an ad’s perceived effectiveness.

What does this article add?

This study adds to the literature by providing evidence of the combination and strength of emotional response more likely to increase perceived ad effectiveness, particularly in the context of recent changes in tobacco use behavior and an evolving media environment. The increasing declines in the prevalence of combustible tobacco use, coupled with a diversified tobacco product portfolio, which includes little cigars and electronic nicotine delivery devices or e-cigarettes, has changed how Americans think and feel about tobacco use. Moreover, media message delivery has significantly increased, becoming more efficient and cost-effective due to the penetration of digital devices. Given these important shifts in message delivery platforms and shifts in tobacco use, this study examines the influence of emotional valence and intensity of emotion related to anti-tobacco media messages on perceived ad effectiveness among contemporary youth and young adults.

What are the implications for health promotion practice or research?

Eliciting emotions through advertising is central to the success of any campaign. This study found that both high-intensity positive and negative emotions were significantly related to perceived ad effectiveness among youth and young adults. However, combining high-intensity negative emotions with high-intensity positive emotions was more likely to elicit perceived ad effectiveness. Given the potential of mass media campaigns to help young people reject tobacco and reduce the burden of tobacco-related disease and death, hard-hitting, emotionally evocative messages should be used maximize the impact of anti-tobacco prevention campaigns, particularly among today’s younger generation.

Footnotes

Authors’ Note

J.R. conceptualized the research question, developed the measures, led the data collection, formulated the analysis plan, edited all sections, and wrote the measures, discussion, and conclusions. M.G. reviewed the literature, provided a factual review, drafted and edited the introduction, and edited the manuscript. D.V. provided topical expertise and edited the manuscript. J.B. wrote the introduction. M.P. conducted preliminary data analysis and reviewed the manuscript. J.G. provided technical expertise surrounding the research conceptualization and data analysis. L.P. conducted data analysis, drafted the methods and results, and created the tables and figure. E.H. advised on the data analysis, conceptualized the research questions, and edited the manuscript.

Acknowledgments

The authors would like to acknowledge Muftau Shinaba for his assistance with the background literature for the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded internally by Truth Initiative.

ORCID iD

Molly P. Green, MPH

References

US Department of Health Human Services. The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. Princeton, NJ: CiteseerX; 2014.

Wakefield

Flay

Nichter

Giovino

. Effects of anti-smoking advertising on youth smoking: a review. J Health Commun. 2003;8(3):229–247.

Durkin

Brennan

Wakefield

. Mass media campaigns to promote smoking cessation among adults: an integrative review. Tob Control. 2012;21(2):127–138.

Farrelly

Niederdeppe

Yarsevich

. Youth tobacco prevention mass media campaigns: past, present, and future directions. Tob Control. 2003;12(suppl 1):i35–i47.

Biener

Gilpin

Albers

. The impact of emotional tone, message, and broadcast parameters in youth anti-smoking advertisements. J Health Commun. 2004;9(3):259–274.

U.S. National Cancer Institute. A socioecological approach to addressing tobacco-related health disparities. In: National Institutes of Health, ed. Institute Tobacco Control Monograph. Vol 22. Bethesda, MD: Department of Health and Human Services, National Institutes of Health, National Cancer Institute; 2017.

Terry-Mcelrath

Wakefield

Ruel

, et al. The effect of antismoking advertisement executional characteristics on youth comprehension, appraisal, recall, and engagement. J Health Commun. 2005;10(2):127–143.

Allen

Duke

Davis

Kim

Nonnemaker

Farrelly

. Using mass media campaigns to reduce youth tobacco use: a review. Am J Health Promot. 2015;30(2):e71–e82.

Dunlop

Wakefield

Kashima

. Can you feel it? Negative emotion, risk, and narrative in health communication. Media Psychol. 2008;11(1):52–75.

10.

Popova

Sangalang

Neilands

Ling

. Do emotions spark interest in alternative tobacco products? Health Educ Behav. 2017;44(4):598–612.

11.

Lerner

Keltner

. Beyond valence: toward a model of emotion-specific influences on judgement and choice. Cognition Emotion. 2000;14(4):473–493.

12.

Kuang

Cho

. Reexamining fear appeal models from cognitive appraisal theory and functional emotion theory perspectives. Commun Monogr. 2016;83(1):120–144.

13.

Kim

Niederdeppe

. Emotional expressions in antismoking television advertisements: consequences of anger and sadness framing on pathways to persuasion. J Health Commun. 2014;19(6):692–709.

14.

Chung

Ahn

Kang

. Processing anti-smoking ads among college students: the role of emotional response and level of smoking. J Promot Manage. 2016;22(3):370–385.

15.

Bigsby

Cappella

Seitz

. Efficiently and effectively evaluating public service announcements: additional evidence for the utility of perceived effectiveness. Commun Monogr. 2013;80(1):1–23.

16.

Durkin

Bayly

Brennan

Biener

Wakefield

. Fear, sadness and hope: which emotions maximize impact of anti-tobacco mass media advertisements among lower and higher SES groups? J Health Commun. 2018;23(5):1–17.

17.

Pechmann

Zhao

Goldberg

Reibling

. What to convey in antismoking advertisements for adolescents: the use of protection motivation theory to identify effective message themes. J Market. 2003;67(2):1–18.

18.

Davis

Nonnemaker

Farrelly

Niederdeppe

. Exploring differences in smokers’ perceptions of the effectiveness of cessation media messages. Tob Control. 2010;20(1):26–33. doi:10.1136:tc.2009.035568.

19.

Duke

Alexander

Zhao

, et al. Youth’s awareness of and reactions to the real cost national tobacco public education campaign. PLoS One. 2015;10(12):e0144827.

20.

Davis

Duke

Shafer

Patel

Rodes

Beistle

. Perceived effectiveness of antismoking ads and association with quit attempts among smokers: evidence from the tips from former smokers campaign. Health Commun. 2017;32(8):931–938.

21.

Davis

Farrelly

Duke

Kelly

Willett

. Antismoking media campaign and smoking cessation outcomes, New York State, NY: 2003–2009. Prev Chronic Dis. 2012;9:E40.

22.

SAS Enterprise Guide 7.1 [computer program]. Cary, NC: Verizon. SAS Institute, Inc. 2017.

23.

Introduction to SAS. UCLA: Statistical Consulting Group https://stats.idre.ucla.edu/sas/seminars/analyzing-andvisualizing-interactions/. Accessed April 1, 2019.

24.

Sutfin

Szykman

Moore

. Adolescents’ responses to anti-tobacco advertising: exploring the role of adolescents’ smoking status and advertisement theme. J Health Commun. 2008;13(5):480–500.