Fighting the pandemic with solid messaging
UNC Hussman School’s Allison Lazard is leading a team to help the state health department design more effective communication.
If you have caught any of Gov. Roy Cooper’s press briefings, the ones featuring North Carolina Department of Health and Human Services Secretary Mandy Cohen advising her fellow North Carolinians about coronavirus safety measures, chances are you’ve encountered work done by Allison Lazard and her team at Carolina. Starting in March, Lazard, associate professor in the UNC Hussman School of Journalism and Media and a member of the Public Health Education Working Group advising University leadership, has led an initial communication task force helping to advise Dr. Cohen, who oversees 17,000 employees and a $20 billion annual budget.
The first phase of the group’s work lasted two months and focused on physical distancing communication. That 27-member team of faculty and student volunteers included Lazard; Kurt Ribisl, who chairs the department of health behavior at the Gillings School of Global Public Health; Allison Aiello, a Gillings School professor and social epidemiology program leader; Nabarun Dasgupta, an epidemiologist with UNC-Chapel Hill’s Injury Prevention Research Center; and Leah Devlin, a Gillings professor of practice who once served 10 years in the role Dr. Cohen now fills.
As the pandemic’s impact on North Carolina evolved and the focus shifted to mask use, Lazard and a smaller team received a state contract for Phase Two, which focuses on communications to encourage mask wearing.
The Well’s Logan Ward spoke to Lazard about her work with the North Carolina DHHS.
Tell me about your expertise.
I focus on health communication. I try to get the right message to the right people on the right channel at the right time to impact their health behavior.
Broadly, I put my work into two buckets. I do a great deal of message-design research. What are the words we should use? How should we frame messages? What channels should we use? What are the images we should include or not include? How should messaging be formatted? Should it be interactive? And then the other component of my research is audience analysis. We have to understand who we’re talking to. Put those two things together, and you can have an impact on behavior.
You refer to your work as “evidence-based health messages.” Explain what you mean.
We’re not just coming up with what we think will work. We’re not just using our own top-down expertise. We are going into the field and confirming what people think. Once we come up with messaging ideas, we show it back to those same people, or people who are representative of our audience, and ask, “Does this work the way we think it’s working?”
We start with methods and theories within the academy of what works in certain contexts, and then we develop our own evidence, which we gather through surveys and focus groups.
And you work side-by-side with public health experts?
Yes. I always partner with someone with a science background to back up what I’m saying. I communicate, but I rely very heavily on my public health colleagues to tell me where the science is on any issue. Early on, when we were telling people to stay six feet apart, I’d ask Allison [Aiello] all the time: “Where are we on the science? Why is it that we need to do six feet? Do we think that’s really going to be the recommendation going forward?”
We had epidemiology experts. We had health behavior experts. They needed someone who could just focus on communication, so I led that effort.
Were you teaching then?
I was teaching two classes. Six students from my undergraduate class volunteered to help, and they were very active on the team. So that’s really cool. They’ve done all kinds of stuff for us — collecting ideas, designing messages, coordinating campaigns. They did a ton of work. This was both beneficial to our effort with DHHS and a great addition to their CVs!
What exactly did the N.C. DHHS want, and how did you go about the work?
At that point, it was the stay-at-home phase, so they wanted us to figure out messaging to get people to stay home. We sent out surveys weekly or every other week. We asked people a laundry list of close-ended items, and we sent it out via consumer panels to people in North Carolina, just to understand what people were thinking. And then the next week, we took what they were thinking, and we asked new questions. And then we started testing some messages. We threw in some messages that were already out in the public to see what people in North Carolina thought about them. And then we started creating our own messages. Each week we would take our findings, show them to the health department and then get their feedback. We would refine the messages, get feedback, refine and get feedback.
Were there any surprises?
Some of the things that we thought would work did not work. That was really helpful for the health department to know, because if you just look at the literature, it says, “Oh, you should do X, Y and Z.” We tried a few of those things and found they didn’t work.
What’s an example?
One thing we know from the literature is that “tailoring” is a really good way to communicate with people so that they make informed health decisions. Tailoring means designing a message specifically for someone. A lot of times people are very responsive to individually-tailored messages.
We tried a couple of different levels of tailoring. In one, we asked people who they were most concerned about during this pandemic. We had them type in specific names, and then, using survey software that allows you to take that response and put it into another question further along in the survey, we piped that answer into a message, like, “If you stay home, you’ll be able to protect Sally.” People hated it.
We were shocked. While I don’t have data to support exactly what was happening, I would hypothesize that it was too personal. It became too real. No one wants to think about the most precious person in their life being sick right now.
In other states, they had a campaign where they were doing that kind of tailoring, but we recommended that North Carolina not go that route.
We also tried a different kind of tailoring — to get people to think of “specific others.” We didn’t pipe in names. We just said, “If you do X behavior, you’ll be able to protect your grandmother.” Or your friend with asthma or your neighbor with cancer. And that level of tailoring was really effective. That was our best rated message throughout many of our behaviors.
Other than the neighbor with asthma, we tested messages like, “protect your loved ones” and “keep your friends and family safe.” Iterations of those messages informed the taglines and other messaging used by the state in the press briefings and materials distributed throughout North Carolina. The underlying messaging framework from our work found its way into N.C. DHHS stuff.
When did that work end?
That iterative message testing lasted from late March to mid-May. And then we had a couple of weeks where we paused and reflected on our data. We delivered a few additional insights from our analyses. And then in mid-June, I talked with the health department again. And that’s when they said, “OK, we want to dig deep into face coverings. Can you help us with that?” This time, they had a small budget, so we submitted a proposal for a six-week contract.
We’re looking at four specific populations in North Carolina and trying to better understand why they’re not able or willing to wear a mask, what might motivate them to wear a mask or, if they do wear a mask, why. We’re specifically interested in people who identify as Latinx, Black or African American, living in a rural area, and young adults (ages 18-25).
Why those groups?
It’s a combination of reasons. These are the people we’ve heard the least from. These are the people least accessible in our big, general research mechanisms. These are also groups most vulnerable and the hardest hit by COVID-19. And there is anecdotal evidence that these groups may be less likely to want to wear a face covering or a mask. That is not our position, but we want to talk to them to find out if that’s true. And if it is, why is it true?
With the goal of then crafting a message that will convince them to wear a mask?
Yes and no. More than that. What if we find that it’s not that people don’t want to wear a mask, but that masks are not available in a community? That will be an important insight for the health department, in case they want to implement a structural change or some kind of other policy. The health department is very eager to learn about these audiences, to know what we can say to them but also to learn about them and their lived experience so that they might be able to better understand what can be done from a health department perspective to improve their health outcomes. Some of these communities are very hard hit right now.
What has your experience been like working with the health department?
It’s really good. I thought academics don’t really sleep, but those people really don’t sleep! They work tirelessly. The people we’ve been working with, primarily through the office of external affairs, are so sharp, so on the ball. Whenever we give them stuff, they give us awesome feedback. They’re timely. They’re direct. They’re appreciative. They incorporate what we give them right away. I am beyond impressed with the work they’re doing. I feel lucky to get a behind-the-scenes look. I wish more people in North Carolina knew what the folks at the health department are doing for us, because they work nonstop.
In addition to Allison Lazard, the current UNC Face Coverings Research team consists of Simone Frank, Marissa Hall, Isabella Higgins, Marlyn Pulido, Kurt Ribisl, Ana Paula Richter, Victoria Shelus, Sara Vandegrift and Rhyan Vereen.