Reflections on Behavior Analysis and Society: Busting Out of the Utopian Ghetto

Quick! Can you guess what these four journal articles have in common?

The articles address different topics in different journals with different missions and subscriber communities. Three are full length articles while the other is a brief report. One is a discussion article, while the others are primary empirical papers. Aside from appearing in 2022 in behavior analysis journals, the articles would seem to have little in common.

But despite their many differences, what unites the four articles is that they are all among the most impactful ever published, in any field, in terms of what has been called dissemination impact, which has to do with the non-scholarly attention received by scholarly work. I’ll explain this claim, but first let’s talk a little about why this kind of attention matters.

A Miraculous Technology

Behavior analysts like to say that their science can save the world (e.g., see here and here and here). Their enthusiasm was bequeathed directly from B.F. Skinner who, across his professional lifetime grew increasingly adamant about the practical value — and importance — of the science of behavior. In the 1948 utopian novel Walden Two, Skinner first imagined that technology based on the science of behavior could have far-ranging everyday impact. In 1953’s Science and Human Behavior, he took a step further, suggesting that behavior science is relevant to every corner of human existence. By 1976, emboldened by early returns from applied behavior analysis (ABA), Skinner asserted that “a special behavioral science… can take the place of wisdom and common sense and with happier results.” By the time of his final paper, Skinner felt comfortable proclaiming that no other science offers “more fascinating things to explore” or “a greater potential for solving the problems that face the world today, above all saving Planet Earth.”

Skinner himself was influential, but today we also have mountains of evidence to support his vision of a science with far-reaching practical implications. For instance, recently some colleagues and I listed 350 domains of socially important behavior to which behavior analysis has been applied. The inset shows a few examples (a sampling from “S” in our “A to Z” list). If you know the behavior analysis literature, the coolest thing about a list like this is how incomplete it seems. When we showed early drafts of our paper to colleagues, for instance, their reaction was not to be wowed by how many domains we’d identified, but instead to blandly observe, “I can think of more than that.” And they were right. If you know where to look, you can find behavior analysis being used to understand and improve just about anything imaginable.

Our “Ghetto”

And yet you do have to know where to look. Much of the most interesting applied work done by behavior analysts has appeared in publications that, from the perspective of mainstream society, are fairly obscure. In many cases the work has been little noticed outside of the tight-knit club of behavior analysts.

This calls to mind another comment from Skinner’s final paper, in which he opined that, “We have been accused of building our own ghetto… Rather than break out of the ghetto, I think we should strengthen its walls.” Skinner presumably was concerned with quality control in a developing science of behavior, that is, the potential for mainstream ideas to dilute its unique and powerful properties. Yet if a ghetto is what we inhabit, that cannot be helpful to a science whose practitioners have world-changing aspirations.

The first, eponymous, ghetto was a Venetian island used during the 16th Century to segregate Jews from the rest of the city’s population. Today, that word still connotes separation and powerlessness. A curious trope of utopian fiction, therefore, is the routine ghetto-ization of the miracle technology being promoted. Commonly, the technology is discovered on some isolated island — literally, as in Bacon’s The New Atlantis, or figuratively, as in Walden Two’s remote Ohio community. As Ronnie Detrich and I pointed out in a recent paper, this narrative device uses geography to explain why the reader has not so far heard of such impressive technology. But it serves another, more troubling, purpose: To absolve the author of having to show how the miracle technology can be imported into our great big world and become scaled for the benefit of all.

Miracle technologies that are not implemented help no one. And outside of a few areas of application (like early intervention for autism and Positive Behavioral Interventions and Supports), behavior analysis seems to remain a hidden secret. Indeed, many people who are not behavior analysts to find the very label “behavior analyst” confusing. My colleague, Val Farmer-Dougan, took to calling herself an “animal behaviorist” because she tired of being confused for a psychoanalyst. For a long time, my own daughter thought my job was to solve serial homicides for the FBI. If people “out there” lack a basic understanding of what we do, they have little hope of appreciating the progress we are making toward doing it.

Wall Builders

So who is responsible for our discipline’s relative obscurity? For a clue, look no further than Skinner’s “strengthen its walls” comment, which has inspired others to propose that behavior analysis must distance itself from all other enterprises, including by avoiding common disciplinary descriptors like “Psychology” and by relying on custom-built (and entirely unfamiliar) substitutes such as praxics and behaviorology. Consistent with this view, behavior analysts tend to publish in behavior analysis journals, attend behavior analysis conferences, and converse preferentially with other behavior analysts. When we do engage with “outsiders,” we often talk in ways that other people don’t understand. Unlike the Venetian Jews, who were forcibly isolated, we have banished ourselves from the very world we wish to save.

Given this tradition, it must be wondered how the amazing stuff that behavior analysts do every day can escape from our self-created ghetto. Many thoughts on this have been shared (e.g., here and here and here), so I will simply repeat a point: If people “out there” lack a basic understanding of what we do, they have little hope of appreciating the progress we are making toward doing it. And they certainly are not likely to support the broad scale implementation of our best solutions.

In contemplating what aspects of ABA may garner societal attention, it is useful to begin by acknowledging that what excites behavior analysts and what excites other people can be two different things. If you build a better mousetrap, the world may not beat a path to your door. Some years ago, when I was an officer of APA’s Division of Behavior Analysis, the association’s President approached me for some examples of how behavior analysis was solving the world’s Big Problems. I began to explain how applied behavior analysis was the most effective evidence-based treatment for autism — but he waved his hands and, having immediately lost interest, walked away. The reason is something Pat Friman explained in a now-famous 2006 newsletter article (for a quick synopsis, see here): Autism is a serious condition but affects a relative few, whereas many of the world’s Big Problems affect a great many. Thus it is possible that as, in my conversation with the APA President, our amazing work in promoting the happiness and well-being of persons with autism is scarcely noticed by those with no direct connection to autism.

Given the possibility that behavior analysts and everyday people value different kinds of solutions, we cannot rely on intuition alone to determine what members of society may be interested in. The only reliable guide would be hard data showing which behavior analysis solutions have become part of conversations with the potential, eventually, to lead to acceptance and implementation.

Dissemination Impact

And now, finally, we return to those four exemplary articles mentioned at the outset, which have achieved a noteworthy level of dissemination impact. To unpack this point, let’s start by noting what dissemination is not. Most readers know about citation impact, with citations being the trail of breadcrumbs that scientists leave behind when notice and are influenced by a published work. We should care about the attention behavior analysis receives from other scientists. But citations can tell us nothing about the attention that scientific work receives outside of the relatively small community of scientists. And 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.

The question of interest, therefore, is how often is our stuff is noticed by people who aren’t behavior analyst and aren’t scientists. We need to know what kind of work is getting the most attention, and what specific kinds of attention it’s getting. This is the gist of dissemination impact.

Before explaining how dissemination impact can be measured, a word of caution is in order. For sciences in general there is a years-long lag between the development of new solutions and their ultimate adoption. Diffusion of innovation theory describes a series of dominos that must fall along the way — the first two of which involve people becoming aware of, and forming a positive impression of, some innovation. With behavior science still fairly new compared to established disciplines like biology and chemistry, we should be unsurprised if not everything we develop has so far tickled the public’s fancy. But at least some of it may have, and there is a way to find this out.

Since 2012, the company Altmetric has been tracking the attention that scholarly writing receives  from a variety of non-scholarly digital sources, including social media, news stories, Wikipedia pages, blogs, patent filings, and policy documents. [Usage note: The company name, Altmetric, is capitalized, and the name of the type of data they tabulate, altmetrics, is not.] To date Altmetric has tracked nearly 42 million articles and other sources, of which about 26 million have received this kind of attention.

In future posts I’ll say more about specific kinds of attention, but for now let’s stick to the big picture. Articles that get mentioned in various kinds of sources build up an Altmetric Attention Score, which aggregates various kinds of attention into a single, interval-scaled score. Altmetric Attention Scores range from 1 upward, with higher scores indicating more attention and a higher relative rank among all tracked articles.

The kind of impact conveyed in an Attention Score is functionally distinct from citation impact. At the level of individual articles, altmetric attention correlates only weakly with citation counts. It is possible, even common, for an article to receive considerable altmetric attention but few citations (or vice versa). For instance, the Behavior Analysis in Practice article with the journal’s all-time highest Attention Score has been cited only 15 times. One of the most-cited Journal of Applied Behavior Analysis articles of the past 10 years has received little altmetric attention.

Although dissemination impact might be operationalized in many ways, altmetric attention is a convenient and readily-available means of quantifying one form of it. Mentions of your work in social media, news stories, and the like do not assure that the work will be widely accepted and implemented at scale. But it seems certain that work which is not noticed will not be accepted and implemented.

See here and here for extended comments on using altmetric attention to quantify dissemination impact. For present purposes, simply note that such data can provide valuable objective feedback for those involved in a given area of study. For instance, if data suggest that society is interested in some of your projects but not others, you can make an educated decision about where to most productively invest your effort. Also, in future posts we’ll discuss some strategies for bringing more attention to your work. Altmetric data can make clear which of these really work.

Four Altmetric Classics

Now, about our four sample articles. These four stand out as what might be called “Altmetric classics” because they rank at the 95th percentile or above among the millions of articles tracked to date by More impressively, as 2022 publications, they achieved this status in the briefest of intervals. If you’re interested in how to get your work noticed, you could to worse than by inspecting success stories like these.

In future posts, i’ll examine the 2022 behavior analysis articles that people were “talking” about (outside of scholarly citations). What topics did they address? In what journals did they appear? What kind of attention did they receive? From whom? I’ll also be inviting authors of some of the most-noticed articles to comment on the significance of their work and the possible reasons why people who are not behavior analysts took notice of it.

As we examine articles that broke through into the public consciousness in a measurable way, some readers, inevitably, will want to know: How much attention is “enough”? For instance, if I tell you that one 2022 behavior analysis article was mentioned in 207 tweets (note: I wrote this before the Twitter social media platform rebranded as “X;” the current term for tweets is now “posts”), is that a lot or a little? And the answer is that it depends. I can tell you that fewer than half of scientific articles are tweeted. I can tell you that 207 is the most tweets received by any 2022 article in a behavior analysis journal. But I can also tell you that, within the massive scope of the Twitterverse overall, 207 tweets scarcely counts as a rounding error. For instance, one 2020 Nature Medicine article on the origins of the SARS-CoV-2 virus has been tweeted over 63,000 times. If you want an additional frame of reference, see here for a list of the most-noticed recent articles in another discipline.

Rather than split hairs over “how much is enough?” let’s simply agree that, where societal attention is concerned, more is better. The more eyes that are on your work, the greater the potential to persuade people out there that behavior analysis has something useful to offer. But there is a caveat. Only positive attention can promote your work and the discipline of behavior analysis, and in a future post we’ll take a peek at some instances where the articles received attention, but not of the desirable type. The problem with social media and other platforms not controlled by behavior analysts is that contributors are free to conclude, and say, what they wish. But even negative attention is useful feedback. If people misunderstand or object to your article, then clearly, if your goal is to promote dissemination, you have work ahead of you.

Coming Soon: 2022’s Greatest Hits of Dissemination Impact

Check back here for a series of posts focusing on the behavior analysis articles people talked about in 2022. We’ll dig into which papers got which kinds of altmetric attention, and discuss some of the possible predictors of this kind of attention. And we’ll hear from authors of some of the most-noticed papers who will explain, in their own words, why they think their paper is important and why it’s getting attention.

This is, in a sense, an exercise in functional analysis. As behavior analysts interested in saving the world, we must acknowledge that the “collective organism” known as society is always right: It likes what it likes, and it’s our job to develop effective solutions that it can notice and appreciate. Magical thinking, along the lines of if you build it, they will come, is not going to grease the wheels of societal acceptance. But solid examples of work that broke free of the behavior analysis ghetto just might provide valuable insights into how to make that happen.

Additional Reading (paywalls may apply)

General introduction to altmetric data and their uses.

Discussion of some shortcomings of altmetric data.

Note: This article is a bit dated but the authors’ general point still holds: Any “one size fits all” metric can be misinterpreted. The same concern, it should be noted, is routinely raised about citation counts as a measure of scholarly impact.