Greatest Hits of Dissemination Impact: 2022 Behavior Analysis Articles in Social Media

If you are a serious scientist and saw “social media” in my title you may have winced. Science has high intellectual standards and focuses on objective solutions to important problems. Nothing sounds further from that than communication platforms that permit literally anybody to say pretty much anything, no matter how crude, vapid, subjective, poorly reasoned, or inaccurate. Heads up, skeptics: I’m going to try to persuade you that if you care about your science you should care about its attention in social media.

As I said in a recent post:

Most of the people who make most of the world’s important decisions are nonscientists. This includes government officials and the citizens whose votes and opinions influence them. It includes those who run large companies and nongovernmental agencies. It includes professionals in many fields who design goods and services, and the everyday people who consume those goods and services and share their opinions about them. To make much of anything happen in the real world, you have to go through nonscientists.

In short, science may not work like the everyday world, but it has to exist in it. If science is going to be funded, and if its insights are going to be embraced and implemented to solve important problems, everyday people will help to decide that. They will only do so if they notice and think favorably about what scientists do.

Dissemination impact means the attention that scientific work receives from non-science sources, and it is valuable feedback when you consider that science is beholden to societal support. Previously I’ve described how altmetric data can be used as a measure of dissemination impact. They summarize how often scholarly works like journal articles have been mentioned in places like news stories, blogs, policy documents, patent filings — and, yes, social media.

Social media posts are by far the most common type of altmetric attention (about 80% of all mentions), and the big dog social media application is Twitter [Yeah, yeah, I’m aware that Twitter has rebranded as “X,” but I’m here going to use the name that applied when I gathered my data].

Although altmetric data also quantify Facebook and Reddit mentions, by my estimate Twitter accounts for > 95% of social media mentions of behavior analysis articles. To give you an idea of the scope of Twitter, here are some basic stats as of early 2023: It has about 450 million users worldwide who visit the app about 7 billion times a month. Almost 60% of users are ages 25 to 49, and they are considerably better educated and more affluent than the world population as a whole. If you’re unfamiliar with Twitter and the short format of its content (formerly ‘tweets,” now relabeled simply as “posts”), you probably haven’t read this far anyway, so I won’t bother to explain.

What I will explain is the attention that 2022 behavior analysis articles got in Twitter, because this is one potentially important index of how much the science of behavior is breaking through onto a societal stage. According to a search using the Altmetric Explorer app on January 1, 2023, 401 different 2022 behavior analysis articles were mentioned by at least one tweet. That is less impressive than it might sound, because many articles receive a starter tweet from someone with a vested interest (author or publisher). So the real interest is in how widely tweeted a given article may be.

The table shows the Top 10 most-tweeted articles of 2022, ranked by total Twitter mentions. Also shown: number of unique Twitter users who tweeted about an article, and the total number of followers those users have. The latter is a measure of potential reach, in that Twitter functions as a network: Users can re-tweet a post they read to their own followers.

[By the way, you might be curious about the accuracy of my ten-month old search. Most tweets appear within a few weeks of an article’s publication, so not much change is expected after that. In my experience, after that critical period totals tend to drop slightly as some Twitter users delete their tweets or accounts. For those articles listed below, totals dropped an average of 9.9% between January 1 and October 1. Long story short, the totals are still informative.]

Twitter Mentions Unique Users Followers Article
65 48 48,266 Kirby et al. (2022). Humble behaviorism redux. Behavior and Social Issues.
64 56 343,181 Arnold et al. (2022). Acceptance and Commitment Therapy informed behavioral health interventions delivered by non-mental health professionals: A systematic reviewJournal of Contextual Behavioral Science.
51 48 107,097 Virues‐Ortega et al. (2022). Functional analysis patterns of automatic reinforcement: A review and component analysis of treatment effectsJournal of Applied Behavior Analysis.
47 45 298,177 Leaf et al. (2022). A call for discussion on stereotypic behaviorEuropean Journal of Behavior Analysis.
41 15 50,352 Garner et al. (2022). Exploitation, freedom, and coercion: The integration of applied behavior analysis in a capitalist systemBehavior and Social Issues.
39 40 70,341 Kishel, C., & Vollmer, T. (2022). An assessment of response to conversation cues of uninterest conducted via telehealthBehavior Modification.
38 27 29,161 Rickardsson et al. (2022). Caring for someone with an acquired brain injury: The role of psychological flexibilityJournal of Contextual Behavioral Science.
33 31 133,336 Rohrer & Weiss (2022). Teaching compassion skills to students of behavior analysis: A Preliminary investigation. Behavior Analysis in Practice.
33 29 62.873 Mead Jasperse et al. (2022). Problem behavior maintained by a precurrent relationBehavioral Interventions.
33 26 40,201 Miller et al.. (2022). How do people with acquired brain injury interpret the Valued Living Questionnaire? A cognitive interviewing studyJournal of Contextual Behavioral Science.

I wanted a sympathetic researcher’s perspective on all of this so I asked Michigan State’s Matthew Brodhead, who actively shares about his research on social media. Here’s why he values this kind of attention:

I love social media because it increases the visibility of my work. It is a very easy way to share my research with an audience much broader than the readership of the journal it is published in. Also, social media is great for sharing pre-prints (e.g., PsyArXiv) of papers, allowing you to get the word out while you wait for the slow-grinding wheels of the final editorial and production processes to finish.

Matt will offer some further comments below, but here what he’s saying is that the boundaries between “scholarly attention” and other kinds of attention are more porous than they used to be. Matt has serious scholarly impact goals, but he sees multiple pathways to pursuing them.

Matt Brodhead

Now consider a few takeaways from the Top 10 article list:

  1. Most Twitter attention is from non-scientists (see the graph below). Twitter lets users self-identify as Members of the Public, Scientists, Practitioners, or Science Communicators. Most who tweeted about the Top 10 articles were Members of the Public. This as it should be if the goal is dissemination impact.
  2. Unsurprisingly, what’s getting noticed is mostly applied work. If everyday people are going to notice science, it will be science that addresses their everyday needs and interests.
  3. Journal of Contextual Behavioral Science had three of the Top 10 articles. As I’ve mentioned in other posts, this journal does an unusually good job of addressing topics of general interest.
  4. Most of the attention is positive. Tweets about journal articles typically either call attention to the article and provide a link, or provide a brief quote or synopsis of the article’s message. For instance, a tweet about Leaf et al. (2022) noted, “Treatment hours ought to be in service of the client’s needs [rather than] billable hours.” I scrolled through all of the tweets received by the Top 10 articles, and I estimate 95% of those represent positive attention.
  5. The amount of per-article attention (between 33 and 65 tweets) is not terribly impressive. For instance, during one six-month period in 2015 the most-noticed articles in the journal Nature received between 806 and 3664 tweets — and those are old data. According to the database, the overall volume of article-mentioning tweets has more than doubled since 2015! Suffice it to say that recruiting social media attention is something for which behavior analysts have room for improvement.
For the articles in the table, percentage of Tweets from different kinds of people. Critical point is that most are not scientists, practitioners, or science communicators.

In closing I’d like to make three points.

First, Altmetric attention (including tweets) is an objective way to measure the social validity of science. You’d be wise to monitor it for your own work, and in a recent article, Dennis Dixon and I explained how to do this using a free tool called the Altmetric Bookmarklet. What’s particularly interesting about social media attention is that it’s the quickest to accumulate after an article is published, with most of the action typically occurring within a few weeks. This makes attention in places like Twitter a potential canary in a coal mine that may suggest whether an article is likely to get noticed by other sources (e.g., news, blogs, etc.) that accumulate more slowly. social media attention can actually catalyze other kinds of attention. For instance, my daughter, a professional journalist, tells me that journalists sometimes comb through social media in search of good stories, so there’s potential for an amplifying effect: Social media can sometimes lead to news coverage.

Second, there are ways for interested authors can generate social media attention for their work (see the end of this post), but this is a bit of a tricky topic. On the one hand, attention can be great feedback about dissemination impact. On the other hand, there are ways to game the social media system to increase attention in ways that probably don’t count as genuine dissemination impact. Consider a parallel concerning scholarly citations. It’s well known that one way to attract a lot of citations is to be controversial — for instance, to criticize respected scientists or well-established findings. If people cite you mainly to say you’re wrong, that’s not a desirable kind of attention. Similarly, any quest for social media attention has to be undertaken responsibly, which probably means not employing all of the tactics that typical social media users do. Matt Brodhead notes that, while attention is valuable, it’s also important to “stay true to yourself on social media. Use it how you want to use it, and don’t feel like you need to be or act like someone you’re not.”

Third, speaking of negative attention, I fudged just a bit in constructing the Top 10 list. I intentionally omitted one outlier article that received more than twice as much Twitter attention as any other article in the list. Unlike for those other articles, however, the majority of its attention was negative — in some cases, deeply, disturbingly unpleasant. In a future post I’ll dive deeply into the story of that article to see what lessons can be learned from it. Stay tuned.

Postscript: Some Social Media Tips for Scientists

Here are a few tips for recruiting social media attention for your research, but this is just kindling: Keep in mind there are plenty of resources available online (like here and here and here) that get into the specifics much more deeply than I will.

  1. Get social media accounts (duh). Notice that’s plural. I’ve emphasized Twitter in this post because that seems to be the current epicenter of altmetric attention for journal articles, but each platform has its own users, so more platforms means more potential sets of eyes. There are, to be sure, social media platforms that specifically target researchers (like ResearchGate), but these less likely to reach the general public.
  2. Build up a network of followers who can help to spread your message. The most common way a message circulates on Twitter is when others re-tweet it (basically, forward it to others). Therefore, the more people who are guaranteed to see your original message the better. Online you’ll find lot of tips for how businesses can accumulate Twitter followers, but some of the same principles apply to individuals. The obvious starting place is to follow others, as Matt Brodhead notes: “Don’t be afraid to follow people, as you’d be surprised how willing people are to give you follow back. And even more generally, don’t be shy about sharing your work, even if you only have a few followers, as building your base of followers takes time.”
  3. Decide what’s worth posting about. If your research addresses theoretical questions using laboratory methods and complex quantitative models, there’s probably not much of a pre-made audience for this. Many of the same features that make research interesting to news outlets also can attract attention on social media, for instance if the research…
  • Addresses issues that affect large segments of the population
  • Is relevant to issues already being discussed in the news or on social media
  • Focuses on social justice issues
  • Produced surprising or unexpected findings
  • Is likely to intersect with the focus of special-interest groups with a big social media presence

Keep in mind that, as Matt Brodhead notes, you’re always guessing about what will capture another’s fancy:

There have been moments where I’ve thought “People should be really excited about this” and then I share it and as it turns out, nobody cared (that is, I receive few re-Tweets and likes). Then, there are moments where I feel less enthusiastic about the message and next thing I know, the Dean of my College is re-tweeting it and it’s getting a bunch of attention. So if there’s anything to be learned here, it is that you have to engage in behavior in order for that behavior to be shaped (to be honest, I have known this for a while just in case anyone is questioning my behavioral wherewithal). So take chances and learn from your successes and failures. Or, if you get stuck, just copy what Derek Reed does because he’s been quite successful with his dissemination efforts (he’s also a swell person).

  1. Say it with panache. Social media posts can’t sound like one researcher talking to another. They need to be direct, clear, and understandable by the wide spectrum of individuals who use social media. They need to get right to the “So what?” (see #3 above). Usually brevity matters. Adding an easy to understand graphic (hint: single-subject graphs probably don’t count here) can make your post more noticeable.
  2. Consider common tips for getting attention for your social media posts, but as suggested above be aware that some of these probably don’t align with your values as a researcher.
  3. Provide a link to the full article for those who might want details. Yes, social media attention increases visits to the full text.
  4. Many posters make for light work. If you have co-authors, post separately, because you’ll likely have different followers. If your journal’s owner or publisher engages with social media, make sure they post as well, for the same reason.
  5. Be prepared for an intermittent reinforcement schedule. Most non-research-focused social media posts don’t go viral, so you won’t hit a home run every time when you post about research. The more you post, of course, the more that consequences (amount of attention) will shape your repertoire.
  6. Whether Twitter will continue to be the dominant force in social media discussion of research publications remains to be seen, but this doesn’t change the general principle that it’s important to get non-scientists talking about behavior science.