Just say no.

Dear Adolescent Health Coordinators (and interested others):

What follows is an assignment for a Liberal Studies class at the University of Minnesota. As many of you know, I’m trying to add some initials to my name and taking the opportunity to learn more that I can apply to our collective challenge: doing right by young people.

In this case, the class was about Fear. Over the course of a day, the professor gave us bite-size morsels from a variety of disciplines that study fear: everything from morality/religion to psychology to addiction treatment. I saw this as an opportunity to explore the literature about sending fear messages to adolescents in public health campaigns.

Background

Hanging out with you all, I picked up the conventional wisdom (based on the literature I assume, since you all are always about the literature) that scary, gross looking images of diseases were not an effective health communication for adolescents.  During the time we’ve been working together, the shaming, blaming “your life will be ruined by teen pregnancy” ad campaigns, sensational cable TV programming about life as a teen mom and the recent public health efforts targeting obesity by fat shaming kids and blaming their parents. And then there was that meth campaign. High production values, dramatic and scary as hell (rape, murder, suicide, prostitution) story-lines. Reports I read said it worked!

Purpose

My goal for this exploration was to 

1. See what sort of evidence supported the use or rejection of using fear-based messages in adolescent health campaigns, 
2. Find out why and how the meth campaign worked 
3. Find good stuff to support our work – for example, if fear messages don’t work -- what does? 
4. And as always, I’m looking for insights, tips, concepts that what in the might improve inform the use of media campaigns and communications to promote adolescent health. (Note: this means I stray a little bit from an exclusive focus on fear campaigns.)

Limitations

Most of you already know that I have very strong feelings on this subject and that my personal experience (as an advertising exec and then a public health groupie) colors my perception and understanding of this topic.  I make no claims that this is a comprehensive review: the citations are linked here. In considering what to read, I chose some meta-analyses, tracked down the names I saw most frequently in citations and Googled my way into a drop box that contained a researcher’s bullet point summary of his meth campaign results analysis written for the Governor of Montana. I was intrigued by some recent brain research (how adolescents deal with fear) and wandered into skin cancer and sunscreen use promotion. While typing up my references I was struck by the diversity of my reading – not unlike the class session itself – I read journals from the fields of psychology, health communication, health education, public policy and marketing, medicine, sex research, community health dermatology, marketing, and prevention science.

Without further hedging, the following are my thoughts on we all should know about fear and public health messages. Here's a glimpse of what is to follow, after the jump:

• Fear messages in public health campaigns aren’t effective.
• Measuring behavior change (the desired outcome of fear campaigns) is mostly bogus.
• Theoretical models of fear processing are cool!
• The unintended consequences are SCARY.
• Developmental Insights!
• The meth Campaign
• Glynis' reading of the adolescent brain research on fear
• Citations

Fear messages in public health campaigns aren’t effective.

No two articles I read agreed. Lots of conflicting reports and many studies plainly said: no go. In general, what evidence there was seemed weak.

Here’s an example. One of the big researchers in this area is Witte. Her meta-analysis assertively states 

“…strong fear appeals produce high levels of perceived severity and susceptibility and are more persuasive than low or weak fear appeals.”

A qualifier here: the studies weren’t specific to adolescents – but Witte also states “Generally, studies have found no effect on acceptance of fear appeal recommendations due to gender age, ethnicity or group membership. (602)” She claims effectiveness, promote little need to customize messages based on audiences and is confident that the use of fear can be effective.

Contrast Witte to Keller, who offered a meta-analysis on adolescent responses to fear messages or Pechmann, who looked at anti-smoking campaigns for adolescents. These researchers are far less certain, qualify everything and express concern over the effectiveness of the measure itself (more on this below). The majority of the other papers I read also contradict the assertion that audience differences don’t matter. The ATOD folks say fear messages don’t work: full stop.

That said, there are some helpful insights with evidence behind them. These are largely from Keller’s sexual health campaign meta-analysis.

Campaigns are only one part of a larger strategy:

• Projects that mobilize the community can diminish social disapproval and intimidation (re: AIDs).
• Media campaigns alone are not enough: they need to operate in conjunction with other activities at the individual, community, and policy levels.

Who creates them is essential:

• Always include members of the target audience as educators, coordinators, and program developers.

Messages matter:

• When messages are tailored (I still don’t understand in what world Witte thinks this isn’t necessary) to specific audience segments, campaigns can be effective at increasing access to health services and overcoming reluctance.
• Note:  There is a measure here! We might be able to track campaign impact on clinic visits. Surely there are many confounding elements, but it's a start!
• When your campaign is motivating and positive, uses clear and simple messages and appears in multiple media channels, you improve your likelihood of success.

Media mix:

• The more places/ways a young person gets exposed to your message, the more successful it will be. They see the TV ad, hear it from parents, talk about it at the clinic, etc.

Expectations depend on our efforts:

• Don’t expect a 6-month campaign to have a lasting impact.
• Must have a reasonable call to action  “Without a link to health services-such as a hotline number-it is unlikely that a media campaign will be successful.”

For the record, all of these points align with what I was taught as an ad exec.

Measuring behavior change (the desired outcome of fear campaigns) is mostly bogus.

All the studies rely on self-report of behaviors (as is our beloved YRBS data set). Keller speaks directly to the point:

"Substantial debate has occurred in recent years about whether health communication efforts have any effect on viewers at all and, if so, what the nature of these effects is. Due to the complexities and costs of conducting scientific evaluations of mass media campaigns, most large-scale health communication programs rely on self-report data to track their effects. Public health experts question the methods employed, pointing to at least two major flaws in how impact is evaluated: the lack of random sampling, and the lack of long-term studies."

In a nutshell, we can’t really assume that intention leads to actual behavior change, tracking behavior change would require a longer time period than anyone is will to track and no one is using randomized samples.

This is an essential point for public health, where we are accountable for our results and demand an ROI on our spending. When I worked in advertising, you could actually see the results of your advertising campaign or coupon drop for packaged goods products. Those lovely bar codes and scanners in grocery stores ensure that manufacturers (and retailers) know exactly what is being sold where. We’d know within a week if our coupon met the goal. We have no way to track the equivalent of “sales” when it comes to behavior change. Therefore you will never really know what impact you are having. Yeah, you can ask the young people what they think but that doesn’t necessarily track (Yes, Mr. Advertiser, I surely would like to buy that sweet new iPhone. Never mind that I don’t have the money).

Further, we’ll never be able to claim that the advertising on its own drove the entire behavior change. Keller talks about this, but the meth campaign research is even more illustrative. Yes! Meth usage (self reports) appeared to decrease during the campaign period. Although the Meth Foundation folks claim it, Erceg-Hurn points out that 1) usage was already trending down over the prior years and 2) shifts in policy regarding access to the over-the-counter ingredients for meth were implemented simultaneously – he posits that this had a far more dramatic impact than the ad campaign.  

I dwell on this point because media campaigns are expensive. The competitive environment is very intense.  At best we’ll be able to say that X% of young people heard our message and report XYZ. The meth campaign story illustrates this point dramatically (below.)

Theoretical models of fear processing are cool!

Witte and others talk about the evolution of models that describe how people process and act on fear. These reveal that it’s not just about scaring someone straight – that for fear to be motivating, the recipient needs to perceived that they personally are seriously at risk and that it is seriously harmful to them. And that’s not enough: the recipient also needs to believe that they can actually do something about it and that the thing you suggest they do will really work.

Witte claims that when

• It’s a very scary thing (to you) that really could happen (to you) AND
• You know how to avoid it/deal with it and believe that it really is possible to deal with or avoid it 
  … a fear message can work.

Witte says “a persuader should promote high levels of threat and high levels of efficacy to promote attitude, intention and behavior change.” (604) The concepts in this model likely sounds familiar. It’s a basic communications strategy to include a clear, do-able call to action. Thinking through what that action should be, what helps accelerate it, what gets in the way of it…these are all marketing issues that must be dealt with.

These models are of interest to me because they help us examine the larger process through which young people adopt or reject behaviors. While there are excellent behavior change models that imitate commercial marketing (aka “social marketing”), our public health community tends to focus solely on “messages.” The fatal flaw in this approach is that, unlike advertisers selling a can of almonds, our “product” is rarely concrete, hard to find/access, may have no immediate benefit or value and may not really be available to us at all!

The big assumption in our approach to behavior change (one that Witte seems to miss – but other researchers note at length (Brennan, Pechmann, Keller)) is that we may not ever be able to offer the critical second part of the equation for effectiveness: for many issues, we can’t really offer adolescents an effective next step that works the majority of the time. In advertising, you can’t broadcast a commercial for your product unless it is available in X% of the markets your message reaches. This ensures that folks seeing an ad for super-absorbent paper towels can go to their local grocery store to buy them. Is this true for health concerns?

• Can we really tell mothers in poor urban neighborhoods that the best way to keep their kids healthy is to feed them fresh, organic food when she has a 2-burner hot plate and can only shop at the 7-11?
• Can we really tell young people to make sexually healthy choices when they are bombarded 24x7 with images and messages that make the exact opposite the norm?
• Can we really tell young girls to avoid pregnancy by using birth control when we know that clinics might not see them, policies may limit options and their health actions will surely be shared with their parents?  

As Brennan puts it: 

“… the size and scope of the problem people are being called upon to resolve with their actions must also be within the capacity of individual achievement." (145)

I am frustrated by Witte’s pat conclusion that “changing teenagers attitudes toward birth control, decreasing perceptions of invulnerability and motivating teens to act are the keys to successfully lowering the rate of teenager pregnancy.” And I’m not the only one. Bradley’s editorial is explicit:

“These ads are trying to effect behavioral change in a community already lacking in resources to sustain that change. Inundating this community with fear about their lifestyle choices, coupled with a lack of means to make fundamental changes and improvement, is far from ethical. Doing so simply marginalizes this community further and it may induce different degrees of medical nihilism.”

The unintended consequences are SCARY

In addition to these ethical concerns, the research documents a number of potentially negative consequences. Even Witte describes the risk: “Strong fear appeals and high efficacy messages produce the greatest behavior change, where as strong fear appeals with low-efficacy messages produce the greatest levels of defensive responses.” What are “defensive” responses? Witte notes that fear messages could motivate the recipient to deal more with their fear than the initial danger proposed triggering “denial, defensive avoidance and reactance.” Translation: the worst case isn’t just an ineffective campaign – we could actually contribute to the problem by using fear as a tactic.

Brennan, who looked at how media messages impact reporting compliance among those on assistance, is more forceful in his condemnation:

“Fear, guilt and shame have been shown to be differentially motivating; with an overuse of fear message resulting if fight more often than not, and shame resulting in flight from the message. The participants in this study were overwhelmed by guilt, and messages invoking guilt were likely to invoke self-protection rather than courage action.” (145)

If we are talking about fear messages, we have to know that feelings of shame are a possible result – as evidenced by the teen pregnancy and obesity prevention campaigns we’ve seen in the past few years. And again, research indicates that when the feeling of shame is evoked, negative results follow.

Commenting on the recent NYC campaign, Time magazine reporter Svalatiz cites research proving that shame is associated with relapse among re-covering alcoholics and feeling embarrassed about one’s weight prompts “obese” people to gain weight. While Svalatiz doesn’t cite a study specifically about adolescents and shame messages, she does connect the dots by reminding us that places where abstinence-only messages and programming dominate also have the highest rates of teen pregnancy.

The online medical journal editorial adds to the danger list by including victim blaming (triggering shame, also) and diminishing returns. Because viewers eventually become sensitized, fear message that rely on shocking imagery or content will need to continually escalate the shock value.

As many of you have collaborated with me on applying framing/communications insights to our work, you know that another significant “unintended” consequence (sometimes I doubt the unintended bit) is that these fear messages frequently frame the teenager as the problem – they don’t eat right, exercise, use contraception and then they drink alcohol, drive badly and do drugs. Net message: teenagers are the problem. More on this in the discussion of the meth campaign.

Developmental Insights!

These readings offered a number of developmental insights.  I’m still trying to process the fear in adolescents/brain research I read! (Since it doesn’t fit with this narrative, go here for a brief summary and links). What caught my attention most was the research that showed how, for young people, social consequences trump health consequences.

The question I get asked most frequently is “how do I get young people to care about … dental care, condom use, physical activity, eating right …” And it turns out that my default answer actually has peer-reviewed evidence behind it.

Keller clear documents that for young people, concern over (fear of) social consequences are better at motivating health intentions than health intentions. This is not the case for most adults, specifically men, who are far more motivated by health outcomes. I found the Tuong study about motivating sunscreen use aligned here as well. This study found that young people were more compelled by appearance-based concerns, compared to health concerns.

By social consequence, we are talking, of course, about peer groups. These are the “smoke and your friends won’t like you” messages. Pechmann agrees with Keller:

“We find that social risk severity and vulnerability are distinguishable from their health risk counterparts and that social risk severity perceptions are especially predictive of adolescents behavioral intentions. (15)

Pechmann’s extensive analysis of message testing concluded that three message-types best triggered young people’s concerns with social disapproval and thereby increased their intention not to smoke. These were:

• Smoking endangers others (think about your little sister!)
• Smokers aren’t cool (negative life style)
• See this cool gal who turns down a smoke? (role model/refusal)

While the Pechmann article was illuminating, the developmental home run in my reading was “The Utility of Understanding Adolescent Egocentrism in Designing Health Promotion Messages” by Greene. Could there be a better article for my interests – and yours? I’d suggest that you go read this article immediately. Yes, it is a bit old (1996) but it illuminates the role of risk taking and adolescent egocentrism in development in a very applied way.

His frame on risk-taking aligns with ours: it is a natural and necessary part of adolescence and the idea that providing facts and information will impact adolescent risk taking is … bunk. 

“Focusing on a recipient’s sophistication of processing and understanding information is very different from furnishing information and facts, as if found inmost current health campaigns.” (149)

Green’s discussion of egocentrism explains why fear of social consequences and promotion of social norms can be so effective. The “imaginary audiences” phenomena – where because young people necessarily spend a lot of time thinking about themselves while their perspective taking skills are still in development – assume that everyone else is thinking about them all the time, too! This pre-occupation makes them very sensitive to these appeals.

Most of these studies recommend “social norms” campaigns that offer young people information about exactly what their peers are and aren’t doing. There’s a whole body of research on those campaigns that I didn’t touch.  

The meth campaign

Here’s the story: Successful Wyoming businessman (Siebel of software analytics fortune) wants to do something about terrible meth epidemic. He creates a foundation and decides to do an ad campaign. Works with a big, fancy ad agency that hires famous directors and cinematographers. They produce high-production value, narrative adsthat are very intense and scary. If you haven’t seen these ads, I recommend that you view with caution – they are triggering in all sorts of ways – particularly for us “youthy” types. Everybody notices – ooh, ahh – look at those ads!  Foundation issues a report with all sorts of good results. Everybody piles on: state starts funding it, feds give it awards (White House!) and more money, several other states pick it up. More ads produced by more big name directors. Ads win tons of awards – even Cannes. Meth usage rates are dropping. Woot! What a success story!

Except it isn’t.

And we probably wouldn’t even have realized it, if not for an Australian researcher (Erceg-Hurn) who thought – that doesn’t seem right.

From Erceg-Hurn's report to the Governor, his chart shows
that contrary to the Meth Foundation's claim, teens
disapproval of meth usage was already high and dropped.

Turns out, the Meth Foundation’s press releases and reports weren’t entirely “honest.” Hid some data, didn’t fully disclose and mis-represented the facts. Erceg-Hurn documents the atrocities in detail, so I won’t recount them here except to say that based on self reports the campaign didn’t actually change young people’s perceptions of the dangers of meth or their intentions to use/not use it. Until Erceg-Hurn’s article, there was no peer-reviewed commentary on the campaign. (Note: In 2012, an article in the Journal of Marketing Research reported on a study that re-tested the ads with college students and determined that “disgust” was a key in engaging these young adults in the topic, showing a reduction in their self reported intent.)

That is not to say that the campaign was not a success.

It was a success if you worked for the Meth Foundation: the emerging darling of anti-drug work. The ad agency (Venables Bell & Partners) and the commercial’s directors received numerous awards and likely a trip to Cannes. Politicians hopped on board to enjoy the positive PR generated from the campaign. HBO took advantage of the buzz and created their own documentary-style programming. Private foundations benefited by funding this positively perceived campaign as it was adopted for use by other states. State and federal legislators and public health departments in leveraged the “successful” campaign for their own use, securing additional funding in the process. (States included Montana, Arizona, Illinois, Idaho, Wyoming, Hawaii, and Colorado.)

Why did we – the public health community – so readily ignore our instincts and over-look the lack of scientific evidence -- our ultimate standard?

For me, the answer to this question is the reason I will never recommend media campaigns that take this approach – it’s easy and it feels good. Associating ourselves with this type of campaign puts us on the “right side” – it conveys our moralistic approach to the individual choices that result in damage and destruction. Drug users make bad choices, do bad things therefore drug users are bad. And because the campaign made it about teens and meth the campaign reinforced our culture’s dominant belief:  teens are bad.

Where, I wonder, is the research that examines how the messages in these ads impact public perceptions about young people? Or public perceptions about who is at fault for the situation?  If, as a culture, we continue to make the problem an individual one when will we address the deeply seeded inequities that blossom into drug epidemics, teen pregnancy and obesity?

And since we know that ACEs (Adverse Child Experiences) are linked to drug abuse, where is the research about how these messages are received by young people who turn to drugs as a response to their experience of trauma?

Finally, while I can’t find a total figure for the amount of money spent on this campaign, Erceg-Hurn claims in 2007, Montana spent $2 million on the campaign. From my own experience, production alone was at least $1 million. I can’t help but wonder how that money could have been better spent? How many Boys & Girls clubs? Sexual health organizations? Food banks? Work training programs? Mental health practitioners? Drug counseling programs?

I’ll end my rant here since my conclusion is obvious.

Adolescent health advocates -- when it comes to fear-based media campaigns: just say no.

Brief Brain Research Summary

Siobhan, Pattwell, Casey & Lee. (2013) The Teenage Brain: Altered Fear in Humans and Mice. Current Directions if Psychological Science. 22, 146-151.

Glynis’ reading of the findings:

You put a kid in a room. Show him a blue card and at the same time a scary noise. And do it over and over again until they are scared of the blue card.

Then you put them in the room and show them the blue card with no scary noise.

Adults and kids stop being scared at the blue card faster/earlier than teens. Teen stay scared by the “cue” longer.

Teens stop being scared by the room – “context” – sooner/faster than adults and kids. But later, when these teens are past adolescence, the fear associated with the context comes back! (so it was only temporarily suppressed).

The researchers explanations:

“It may not be surprising that this temporary suppression of contextual-fear expression coincides with the developmental period encompassing the transition into and out of adolescence, during which animals engage in heightened exploratory behavior required for sexual reproduction and survival. For adolescents, exhibiting heightened levels of contextual fear during this period of exploration may prove maladaptive. Specifically, adolescents may be unwilling to leave the safety of their parents’ niche and explore new environments.”  (149) 

 “If an animal is afraid to venture out of the home environment, then it may exploit and deplete food in the home environment and fail to find a mate. However, if the animal ventures out and is attacked by a predator, it will be no more likely to procreate or survive than if it had stayed at home. Thus, the animal needs to be highly vigilant to cues of threat in new environments, which may explain the heightened fear response to cues of threat.” (150)

References

Arrillaga-Andreessen & Chang.  (2009) Thomas and Stacey Siebel Foundation Meth Proejct. Stanford:CA. Stanford Graduate School of Business. Retrieved on May4, 2014 from https://gsbapps.stanford.edu/cases/documents/SI-114%20Thomas%20and%20Stacey%20Siebel%20Foundation%20%20060113%20(1).pdf

Bradley. (2011). Ethical Considerations on the Use of Fear in Public Health Campaigns. Clinical Correlations: NYU Langone Online Journal of Medicine. Retrieved on April 17, 2014, from http://www.clinicalcorrelations.org/?p=4998

Brennan & Binney. (2010) Fear, Guilt and shame appeals in social marketing. Journal of Business Research.  63: 140-146

Erceg-Hurn. (2008). Drugs, Money and Graphic Ads: A Critical Review of the Montana Meth Project. Prevention Science. 9: 256-263

Erceg-Hurn. (2009.) Statement of David Erceg-Hurn to Governor Schweitzer and Montana Legislature regarding funding of Montana Meth Project in 2009 Budget. Retrieved on April 21 from http://bloximages.chicago2.vip.townnews.com/billingsgazette.com/content/tncms/assets/v3/editorial/f/a3/fa3502c6-6981-11de-a692-001cc4c002e0/fa3502c6-6981-11de-a692-001cc4c002e0.pdf.pdf

Frameworks Institute. (2000-2008) Framing research series on public perceptions/messages about Adolescents. Retrieved in 2001 through May, 2014 at http://www.frameworksinstitute.org/adolescence1.html

Greene, Rubin, Hale & Waters. (1996) The Utility of Understanding Adolescent Egocentrism in Designing Health Promotion Messages. Health Communication 8(2) 131-152.

Keller & Brown. (2002). Media Interventions to Promote Responsible Sexual Behavior

The Journal of Sex Research. 39 (1): 67-72

Keller & Lehmann. (2008). Designing Effective Health Commuications: A Meta-Analysis. Journal of Public Policy and Marketing 27 (2): 117-130

Montana Meth Project. (2011) Fact Sheet. Missoula: MT. Montana Meth Project. Retrieved on April 21 from http://montana.methproject.org/documents/Montana%20Meth%20Fact%20Sheet11-7-11.pdf

Morales, Wu & Fitzsimmons. (2012). How disgust enhances the effectiveness of fear appeals. The Journal of Marketing Research. 49(3): 383-393

Nauert, R. (2012) More Difficult for Teens to Block Out, Overcome Fear. Psych Central. Retrieved on April 16, 2014, from http:/psychcentral.com/news/2012/09/28/more-difficult-for-teens-to-block-out-overcome-fear/45285.hmtl

Pechmann, Zhao, Goldberg & Reibling. (2003) What to Convey in Antismoking Advertisements for Adolescents: The Use of Protection Motivation Theory to Identify Effective Message Themes. Journal of Marketing. 67(2) 1-18

Prevention First. (2008). Ineffectiveness of Fear Appeals in Youth Alcohol, Tobacco and Other Drugs (ATOD) Prevention. Springfield: IL: Prevention First.

Siobhan, Pattwell, Casey & Lee. (2013) The Teenage Brain: Altered Fear in Humans and Mice. Current Directions if Psychological Science. 22, 146-151.

Svalavitz. (2013). Why New York’s Latest Campaign to Lower Teen Pregnancy Could Backfire. Time Magazine Online. Retreived on April 16, 2014 from http://healthland.time.com/2013/03/28/why-new-yorks-latest-campaign-to-lower-teen-pregnancy-could-backfire/

Tuong, Armstrong. (2013.) Effect of appearance-based education compared with health-based education on sunscreen use and knowledge: A randomized controlled trial. Journal of the American Academy of Dematology. 70(4). 665-669.

Witte. (1997). Preventing Teen Pregnancy Through Persuasive Communciations: Realities, Myths, and the Hard-Fact Truths. Journal of Community Health. 22(2).  137-153

Witte, Allen. (2000) A Meta-Analysis of Fear Appeals: Implications for Effective Public Health Campaigns. Health Education and Behavior. 27: 591-614