Online influencers are increasingly powerful, but Benjamin Johnson explains that they present users with a choice when it comes to trust.
Benjamin Johnson, Ph.D. is an assistant professor in the Department of Advertising at the University of Florida College of Journalism and Communications. His research is focused on why and how people select and share persuasive messages in new media settings, particularly as it relates to psychological processes. The interview has been edited for length and style.
Tell us about yourself and your research.
The topic that I deal with is called media psychology. It is really focused on people’s thoughts and feelings when engaging media, the individual-level processes. In particular, I try to understand why people consume media in the first place, or why they choose certain messages.
| Click here to read about Benjamin Johnson’s research. |
A lot of research is looking at the effect. I try to understand what happens before that, how people end up in a situation where they could be affected in the first place. How do they come into contact with media to be influenced? Within that paradigm, I look at a lot of different content areas.
There is a classic approach, which we call selective exposure. People choose media that they agree with, so they tend to spend more time with stuff that is agreeable to them. I do some work in that area. I also do work in entertainment, why people choose certain types of films or how they curate their social media to be more or less entertaining.
Over the past few years, I have done more and more research with advertising influencers, why people might choose to engage with sponsored content or where they might choose to engage with something known to be persuasive. I have been looking at this collection of influencers, but also how those influencers are getting people to change their beliefs about particular brands or health behaviors. So, I work on a lot of different types of media content but it is usually trying to understand people’s thoughts and feelings as they approach or engage with media.
Tell us about some of the projects that you have been working on, particularly your work with the Consortium on Trust in Media and Technology.
One of the things that I am working on at the moment is thinking about ways in which expertise and trust might have some trade-offs. So, you might really trust your friends. They are very authentic and they are very genuine and real. But they are not necessarily experts.
When someone is an expert, you might think that they are very knowledgeable, they have training, they know a lot. But maybe they are not authentic in the same way. And so, historically, when we look at credibility and communication research, and psychology research, we always made a distinction between those two things: there is expertise and there is trust. We always made that distinction but we did not always separate them in terms of how we analyze them.
So, what I am doing is looking at situations where trust and expertise may actually exist in opposition to each other, and a lot of this is driven by some of the things we see with influencers. They really want to be authentic and real and genuine. They want to be like your friend or peer. They want to be a little bit rough around the edges, but they also want to be cool. They want to have some expertise. So there is this really interesting tension that exists. That is certainly something I want to look at in the political realm as well. I think you will probably see an increase in people putting their confidence in sources on the basis of being real, authentic or rough around the edges—and rejecting expertise.
Tell us more.
We are looking at the quality of production in influencer videos, if they are really carefully scripted and produced versus low budget. Do the personalities reveal a lot about themselves and feel really honest? Or are they careful about what they say? They are usually focused on some kind of interest or hobby or practice. What happens if you have an influencer who is not political, you have someone who has been known for video games or hunting and fishing or whatever? What happens if they talk about politics?
Will those people be especially effective deliverers of messages about politics because they have credibility, and because they have a shared interest with the audience? We are looking at the connection between non-political sources trying politics, and what happens when celebrities try politics. A lot of celebrities enter the political fray to talk about issues. Sometimes it is very effective. Sometimes it backfires. And so we are trying to tease out this influencer process of talking about politics, similar to celebrities doing endorsements or making political statements.
Let’s talk a little bit about how this relates to trust.
There is still a lot I do not know. It keeps me going on . I think one thing I want to look at going forward is that we have this term that gets kicked around: authenticity. What does that mean? I think there is still a lot of work to do in terms of understanding. What is the nature of credibility? What is the nature of trust? What are the different pieces? And some will matter more in certain contexts, like authenticity might really matter a lot. But it may not matter at all for celebrities. I think the trust that we have could look different for different sources.
I think there is a lot more work to be done in terms of mapping out, what is trust? What is credibility? How does it differ between different types of sources in different situations? And I think it is a moving target, too. I think trust and credibility have been altered in a lot of ways because of technological change. We have access to so many more voices. We have a bigger influx of people talking. There is more competition for attention. That obligates the end user to do a lot of work to figure out who is trustworthy. There has been a lot of upheaval in terms of credibility and trust. So I think we still have a lot of work to do.
What is your sense of how the end-user is coping with that burden?
I do not want to be the person who says we need individual-level solutions to social-level problems, as if there are no problems with regulation and technology and platform governance. That has to be addressed and should be addressed. But in the meantime, while we are still sorting that out, it is helpful for everybody if individual users are careful, they are reflective and using cognitive bandwidth to be critical.
What else are you studying?
One article that just came out speaks to how partisans make sense of political books on Amazon, like Republicans seeing Republican or Democratic books on Amazon and then looking at the ratings.
Perhaps Bill O’Reilly’s books are really highly rated. Or imagine there is a Nancy Pelosi biography that has really low ratings. What we did was we manipulated those ratings in some experimental sleight of hand.
We wanted to look at a lot of psychological responses and to see if any of those would be connected to people polarizing and having more negative feelings toward their opponents. What we found for Republicans and Democrats is that if the ratings were against them, and it is a book they liked, they said it must be manipulated.
If the ratings were agreeable and consistent with your worldview, people thought that the people who were providing the ratings were fellow partisans . This is just one little book. But people see messages all the time. What people believe about those rating systems could activate in-group and out-group emotions.
When you say in-group and out-group, what do you mean?
In-groups are the groups you belong to. If people have a political affiliation, their political party is one of their in-groups, they are part of that group. These groups are part of identity. In the U.S right now, political identity is one of the powerful, overriding identities. People have really strong feelings about that, like really negative feelings toward the out-group party.
Obviously, you have been studying this for a long time and you have a lot of insight into sort of trust in the online world. Describe the environment that you are seeing.
I think one thing that came out of this summer was that more whites were supportive of Black Lives Matter than when BLM had its first protests after Ferguson in 2014. There was a change in American trust toward police departments. And that has been very, very rapid. Those beliefs were not very malleable in 2014. There was not a lot of change that came out. But after the tipping point, there was a lot of change.
I think in our current moment, we see that this in-group, out-group thing really matters in terms of trust. People who are alike trust each other and they distrust people who are different. I think when those processes reinforce each other, there is a spiraling effect that can lead to a lot of social cleavages.
When you restrict your worldview, you lose perspective, and it damages the ability for us to function as a state or a country or a species or whatever. Being able to cooperate with people who are different is really important. When you have trust in people who are different, who have different goals or different functions or different perspectives, being able to build that trust is adaptive. It keeps us alive and helps us survive when we are able to cooperate with other people. So, I think we have to have it.
You are working to make changes in your field. Tell us a bit about that.
In social psych, about 10 years ago, there were some cases of fraud, which impacted the day-to-day practices of psychologists and social scientists more broadly.
But then, there are also some instances of classic studies that did not hold up later on. When you go to try to recreate the classic experiment, it is not reproducible. Being able to replicate findings is a fundamental part of science, but it is often something that is not done as much as it should be in social science. Some of that is because the things we study are more fuzzy. People have free will, people often act in unpredictable ways. So it is not like physics. There is a lot more noise in explaining and predicting how people behave. But if we say that we can explain certain behaviors, it should be able to hold up over time.
So there was a big push about 10 years ago for more replication in social psychology, and that made people even less confident in the findings and the methods and procedures. So there has been a big reform toward what we call Open Science, which is basically being radically transparent. So sharing data, sharing material. If you have a questionnaire or you have images, sharing those with other researchers so they can look at what you did in excruciating detail. The other thing is also to make a stronger distinction between prediction before data are collected and the post-hoc explanation.
Explain what you mean about prediction and explanation.
So basically, you go into a study with expectations about what you will find. You often do not find that. You find something different. Some researchers do go back after the fact and rewrite the hypotheses so that it tells a better narrative. But you are claiming that you made a prediction on the post-hoc information.
So there has been a big movement to be extremely clear about what happens before a study is launched. That is called pre-registration. It is similar to clinical trials in medicine. Basically, you file something in advance, and then that essentially becomes immutable so that it cannot be altered the day after.
You have been involved in promoting these ideas. How did you come to participate?
I have been part of a group for the past few years that is trying to adopt and implement some of these reforms. We published an agenda for science and communication research. It has 37 authors from, I think, 30 universities in eight different countries. We published it in the leading journal in communication.
I was in a PhD program when all this reform started bubbling up. I thought that it was something that would make my research better, make me a better researcher, and make my work a higher caliber. I started tinkering with some of these solutions, with my pre-registering and sharing data. This serves as my default process now.
If we are doing social science right, we have to try to prove ourselves wrong. If there is something that we thought was true and then the data show that is not true, we want to show that. We need to be very clear about when we make predictions, what we plan to measure, who we are going to include, and how we are going to analyze all that. This can be a bit rigid, but it gets us a bit closer to the truth.