Equity and Diversity in Behavior Analysis: Lessons From Skinner (1945)

Abstract

Several authors have written about the disparity between our values statements concerning gender equity and diversity and the behavior of our professional organizations. In this article, I argue that this is a predictable by-product of our collective cultural learning histories, that we have access to the variables that must be manipulated to alter this behavioral trajectory, and that now is the time to apply the principles of behavior toward changing our current repertoire. As a case in point, I provide evidence regarding the current state of the efforts within psychology and behavior analysis to ensure gender equity, and end with a series of recommendations for institutions and individual leaders to enact toward the presumably valued outcomes of equity and, more broadly speaking, diversity.

This paper begins with a riddle. It would be useful to have a stopwatch on hand to record the latency to your solving the mystery that follows.

Imagine a boy and his father driving late at night through a densely wooded forest on a winding road. The car swerves and crashes into a tree. Paramedics arrive on the scene to find the boy’s father had died on impact. The child is rushed to the hospital, where emergency room physicians declare him in need of surgery. He is wheeled into the operating room. The surgeon opens the blue curtain, looks at the child, and says, “I can’t operate on this boy—this is my son!”

How could that be? What happened? I invite you to start your timer now and record the latency to your solving this puzzle.

If you are at all like me, you might have found yourself dismayed at the length of time it took you to supply the supplemental stimuli needed to solve the problem. In my case, I heard this brainteaser first when I was in high school, some 40 years ago. I had been at a friend’s house when I was given the problem and took the bus home, and it was not until I sat down for coffee with my mother that I was exposed to the supplemental stimulation that allowed me to find the solution.

The surgeon was the boy’s mother.

To me, it was shocking that I had taken so long to find the answer. My mother was an emergency room physician! I felt ashamed that the correct answer had not leaped out at me the instant I heard the riddle.

It was not until I read Skinner (1945) a few years ago that I began to view the latency to my response during high school with some self-compassion. Skinner’s paper, “The Operational Analysis of Psychological Terms,” provided a means to understanding the conditions under which the question regarding a “surgeon” stumped me. In truth, Skinner’s paper was not an operational analysis of current-day psychological terms. It was an extended definition of what it means to define terms—a definition of definition. Perhaps, too, it was an outline of his forthcoming book, Verbal Behavior, which has yet to be fully recognized by mainstream psychology for its importance. In essence, Skinner suggested that the meaning of words is found in the variables of which their use is a function. Words uttered by scientists to other scientists and words spoken by scientists to their friends and family are behavior, and the behavior of a scientist has to be understood with respect to the context in which the words are emitted. More simply, what you mean and what you understand are contextually determined. You might use the term respondent function when describing to another behavior analyst the effect of a sad story on a learner, but were you to use the same term to describe the effect of the story on your child to a group of neighbors at a cocktail party, you might not be invited back any time soon.

Skinner (1945) suggested that verbal context involves all the variables of which behavior is a function (motivative, discriminative, respondent, reinforcing, and punishing variables), each playing their respective roles simultaneously in selecting verbal utterances. Importantly, some of these variables are not observable in the present moment but could be part of a relevant learning history involving cultural selection and retention. For example, an interventionist may have been taught during training to say, “No, thank you. I brought water for myself,” when offered tea or milk by the family of a child receiving applied behavior analysis services. But a social rule the interventionist learned many years earlier—that it is disrespectful to refuse such an offering—could serve competing motivative functions in the presence of a congenial elder offering a beverage. Thus, in this circumstance, it is unclear whether “No, thank you. I brought water for myself,” would be the words this context selects.

Likewise, some behavior is unobservable and requires a verbal community to bring awareness to the speaker of the variables of which their behavior is a function so that the speaker can then provide useful information to others. In the previous example, a supervisor might ask, “Why didn’t you say what you’d been trained to say?” and, “Did something happen that made it difficult for you to behave as you’ve been trained?” The interventionist could then explain that in the presence of an elder offering a beverage, an earlier rule serving as discriminative for a different response exerted control and the competing contingencies resulted in nonoptimal behavior.

It is this earlier history that is useful to examine when assessing responses to the riddle about the surgeon at the hospital. The word surgeon is historically situated in frames of equivalence with White, middle-aged, straight, binary men. For example, there are all the familiar archetypes from television, such as Marcus Welby, Hawkeye Pierce, and Leonard McCoy. Because surgeon and physician are framed in coordination, other White, male, straight, binary medical archetypes such as Salk, Freud, Spock, and Jung all have respondent, discriminative, and reinforcing functions that come to bear when responding to the word doctor or surgeon in the present context.

It is this kind of history that is relevant to understanding the behavior that has resulted in the current optics of our behavioral university departments, editorial boards, and organizational leadership. And I will make the point that it is these variables that need be changed to bring our behavior in line with our organizations’ values statements with respect to gender equity and diversity. I will examine my own history as a case in point.

I grew up during the 1970s. In those years, the textbooks I read were filled with images of middle-aged, White, straight, binary men. The answers to most questions in social studies, science, and literature classes were the names of those men. Faces on the front page of the New York Times, television news broadcasters, and top-billed movie stars were all White, middle-aged, binary men. It was not until very recently that women of color and people identifying as transitioning or lesbian, gay, bisexual, transgender, queer, and other nonbinary gender identities (LGBTQ+) became transmitters of cultural rules. Today, their stories of powerful engagement with the world are becoming news and entertainment norms. Political winds may be blowing in a way that will result in a more diverse and equitable representation in government offices. I hope so.

Also, today I am the picture of privilege: White, male, cisgender, straight, standard-dialect-speaking, overeducated privilege. Just wrapping my head around the extent to which my background makes me blind to the way I exert my privilege is a full-time job. It is like understanding what it is to be a parent of a child with autism—I cannot. But it is my responsibility to learn whatever I can and to share what I learn. This article is an effort in that direction.

The point is that there is an impact of decades of exposure to homogeneous (male, binary, White, straight) images purported to represent everything that is relevant, just, powerful, and courageous. Couple that with decades of reinforcement for behaving with respect to those norms and punishment or aversive self-stimulation for behaving outside those boundaries and the latency to solving that riddle about the surgeon loses its mystery. Of course the riddle takes most of us—women, men, and nonbinary individuals—time to solve. For that matter, of course our journals are authored and edited mostly by White men, our departments headed by White men, and accomplishments of mostly White men recognized by our organizations.

It is not OK by any means. It is not OK.

But it is predictable.

And there are things to be done about it.

One option is to apply a behavior-analytic lens to the matter of cultural change. The hallmark contribution of behavior analysis to psychological science is the establishment of a bar that is set higher than prediction. We aim to influence the future course of behavior. This requires a precise and functional definition of the response class slated for change and a clear evaluation of the conditions under which new behaviors are most likely to occur. Skinner (1945) suggested that terms such as diversity and equity ought to be evaluated by an examination of the conditions under which they are used. Needed is an examination of the variables of which these terms and the behaviors that correspond with them are a function. To date, the behaviors associated with diversity and equity have not been subjected to such an analysis. What follows is an examination of the ways in which one specific area of focus, gender equity, has been treated thus far in the empirical literature. Although analyses of other types of equity and diversity would be relevant to the discussion, there is a dearth of available research pertaining to these topics.

Gender Disparity in Behavior Analysis

Several authors have evaluated the status of women in behavior analysis (e.g., Iwata & Lent, 1984; Myers, 1993; Neef, 1993; Poling et al., 1983). Of note, Myers (1993) provided evidence of gender inequity in that the percentage of female full members of the Association for Behavior Analysis International (ABAI; 31%) was notably larger than the percentage of female first authors in the Journal of the Experimental Analysis of Behavior (JEAB; 15%). McSweeney and Swindell (1998) reviewed the participation of women as authors in JEAB and ruled out several hypotheses that explained away the inequities, such as the special skills involved in preparing manuscripts for publication and interest in the subject matter of the experimental analysis of behavior. Their analysis involved comparing female participation in the authorship of JEAB articles to that in journals with comparable subject areas and rejection rates. They suggested that because journals with similar selectivity and subject matter show higher rates of female authorship, the common answers that women are disinterested in the topic and undereducated cannot explain the finding of gender disparity. Similarly, McSweeney, Donahue, and Swindell (2000) showed that despite notable increases in the participation of women authoring papers in the Journal of Applied Behavior Analysis (JABA) between 1978 and 1997 (26%–38%) relative to comparable journals, a “glass ceiling” limits the participation of women in applied behavior analysis to lower levels of the profession. Of additional note, McSweeney and Parks (2002) showed a similar trend across nonbehavioral journals in the fields of developmental and social psychology. Odom (2000) subjected raw data visually inspected by McSweeney and Swindell (1998) to a statistical analysis and found less evidence of a glass ceiling and cautioned that emphasis should be placed on recruitment and retention as opposed to differential standards for men and women.

Despite these disparities, each of the articles cited previously suggests that the overall trends in the data were positive. Nevertheless, some evidence attenuated this optimism, such as the finding that publications authored by women were most likely to have been edited by women (McSweeney et al., 2000) and that as the selectivity of the journal increased, the participation of women decreased (McSweeney & Swindell, 1998; but see Odom, 2000, for a rejoinder). Also, Rehfeldt (2018) cited works indicating that women devote more hours to service work and occupy fewer associate- or full-professor positions than men in university settings. It is important to note that despite the meticulous analysis of data across multiple journals, each of the aforementioned articles presented conjectural hypotheses to explain the data, without subjecting them to an analysis of specific or general classes of controlling variables.

In 2018, Li, Curiel, Pritchard, and Poling published an updated evaluation of authorship and editorial board membership among women and men for seven behavioral journals from 2014 to 2017. Their analysis was consistent with those discussed previously in that women’s participation appears to be on the rise, but the authors cautioned that globally, women continue to face serious issues and that these problems mirror those of racial, ethnic, LGBTQ+, and other underrepresented groups. As with other behavior-analytic articles that have analyzed these trends, the authors did not translate the issues they hypothesized to be involved into behavior-analytic terms and did not make intervention recommendations, though they did mention that this would be a worthwhile endeavor.

Most recently, Nosik, Luke, and Carr (2018) conducted an updated analysis of overall trends in women’s representation in behavior analysis. They evaluated the percentage of women compared to men in nine achievement categories and listed these classes in relation to the career point at which each would be likely to occur. For example, the authors suggested that certification is expected to occur at a very early stage in one’s career, compared to invited speakerships, organizational leadership positions, and induction into the ABAI fellowship. As predicted, Nosik et al. found the proportion of women to men achieving milestones in the early stages of their careers was higher and declined as the categories focused on later stage career milestones. The authors agreed with Simon, Morris, and Smith (2007), who argued that some of these disparities could be explained by barriers such as family responsibilities and sexism, but they noted complications in defining what the goal would be if these barriers were lifted. Specifically, these authors suggested that when judged against a standard of parity, the state or condition of being equal, women appear to have reached or exceeded 50% representation. However, when evaluated with respect to proportionality, or the degree of correspondence in quantity of one thing to another, the number of women in the categories described previously could potentially be proportionally much smaller compared to the total number of women engaged in that activity. Because data on the number of women engaging in these activities are not currently available, an empirical evaluation of proportionality is not possible. However, the authors suggest that given current societal barriers, it is likely that these data, when available, will show that the glass ceiling McSweeney and colleagues discussed has not yet been lifted.

The ABAI website offers data that can be useful in evaluating the current state of our primary membership organization with respect to the recognition of women compared to men. A close examination of two particular categories related to the representation of women in the ABAI shows improvement and some lingering areas of concern. The top panel of Figure 1 shows the cumulative frequency of women compared to men in the role of president (ABAI, 2019). As of 2016, the total number of women presidents was 13, compared to men, which was 27. During our first 40 years (1976–2016), the percentage of women elected president of the organization was 32.5% (13 out of 40). Of note, the percentage of women increased between the earlier and latter half of this time period. Between 1976 and 1996, women comprised only 20% (4 out of 20) of the presidents, compared to 45% (9 out of 20) between 1997 and 2016.

Fig. 1.

ABAI presidents (top panel) from 1976 to 2016 and ABAI fellows (bottom panel) from 2004 to 2017.

According to the ABAI website, the fellowship program was started in 2004 “to recognize the most outstanding contributors to behavior analysis in one or more of the areas of research and scholarship, b) professional practice, or c) teaching/administration/service,” and current fellows nominate potential inductees (ABAI, 2019). The bottom panel of Figure 1 shows the cumulative frequency of inductees who are women compared to men into the ABAI fellowship. As of 2017, only 17 women had been recognized, compared to 82 men. The graph shows a steep acceleration of male inductees during the first 5 years, during which a large number of nominations was accepted for contributions made during the decades prior to the program’s inception when social factors sharply precluded women entering the sciences. But the trend has not decelerated, and in contrast, the trend of women inducted into the fellowship is notably flatter. As with the presidents of the ABAI, the first 7 years of the fellowship showed a more disparate number of men compared to women (6 out of 68; 13.2%) compared to the next 7 years (8 out of 31; 25.8%). This positive change notwithstanding, the overall comparison shows that the glass ceiling effect articulated by McSweeney et al. (2000) has not yet cracked.

The ABAI’s first diversity policy was accepted by the Executive Council in 1994. In 2018, the policy was redrafted, and as of this writing it reads,

The Association for Behavior Analysis International encourages diversity and inclusiveness in the field of behavior analysis broadly, and within the organization specifically. Diversity refers to differences in race, ethnicity, sexual orientation, gender identity, age, country of origin, religious or spiritual beliefs, ability, and social and economic class. (Association for Behavior Analysis International, 2019a, 2019b, 2019c, 2019d)

The addition of a definition of diversity is new to this latest policy. Of note, however, is the fact that inclusiveness mentioned as a separate class of behaviors from diversity has not been defined within this statement or elsewhere. More important to the current discussion is that no attempt is made to specify the variables of which diversity or inclusiveness is a function. The definition provides examples of topographies that are included but does not include any reference to controlling variables. Perhaps that would go beyond the scope of an organizational policy statement; nevertheless, it would be fitting for an organization dedicated to the science and practice of behavior analysis to provide some additional language pertaining to the way in which it aims to evaluate and foster these important goals. To this end, a task force has been launched and empowered to do just that.

There is evidence that the ABAI is committed to the values suggested in its diversity policy (Association for Behavior Analysis International, 2019a, 2019b, 2019c, 2019d). Table 1 depicts ABAI membership as of fall 1989 (Redmon, 1989). Thirty-five nations are represented. Notably, due to outreach efforts by delegations of ABAI leaders, membership was already beginning in Far and Middle Eastern nations such as China, Japan, Jordan, Korea, and Saudi Arabia. Unfortunately, differentiated data on the representation of women in the international behavior analysis community for that period of time are unavailable.

Table 1.

ABAI Membership, 1989

COUNTRY	MEMBERSHIP					REGISTRATION
	1989	1988	1987	1986	1985	1989	1988	1987	1986	1985
AUSTRALIA	7	8	5	9	2	2	2	1	4	0
BELGIUM	1	2	1	0	0	0	1	1	0	0
BRAZIL	9	13	6	13	6	8	5	10	5	2
CANADA	76	67	62	66	77	51	45	53	39	39
CHILE	0	0	0	0	0	0	0	0	1	0
CHINA	1	0	1	0	0	0	0	1	0	0
COLUMBIA	1	2	1	1	1	0	0	0	0	0
COSTA RICA	0	0	0	1	0	0	0	0	1	0
DOMINICAN REF	2	2	2	1	1	0	0	0	0	0
FRANCE	1	2	1	1	2	0	0	0	0	0
GREECE	1	1	1	0	0	0	0	0	0	0
GUADALAJARA	0	0	0	0	1	0	0	0	0	0
ICELAND	0	0	1	0	0	0	0	1	0	0
IRELAND	2	1	1	1	0	0	0	1	0	0
ISRAEL	2	2	0	0	2	2	2	0	0	2
ITALY	5	13	1	1	1	1	1	0	0	0
JAPAN	16	15	10	0	11	6	6	5	4	7
JORDAN	0	1	0	0	0	0	0	0	0	0
KOREA	1	0	0	0	0	0	0	0	0	0
MEXICO	6	4	3	2	0	5	4	4	1	0
NETHERLANDS	4	4	2	1	1	3	3	1	2	0
NEW ZEALAND	1	3	0	0	2	0	4	4	0	2
NORWAY	2	2	1	2	0	0	1	2	2	0
PERU	2	3	0	0	0	1	4	0	1	0
PORTUGAL	2	1	1	1	1	1	0	0	0	0
PUERTO RICO	2	2	1	0	1	1	0	0	1	1
SAUDI ARABIA	1	0	1	1	1	1	0	1	0	0
SPAIN	1	2	2	0	0	0	1	0	0	0
SWEDEN	2	3	2	4	3	0	1	1	1	0
UNITED KINGDOM	7	7	3	3	3	3	2	1	3	1
UNITED STATES	1824	1927	1731	1712	1713	1166	1340	1272	1287	1187
URUGUAY	0	0	0	0	6	0	0	0	0	0
VENEZUELA	1	1	1	1	1	0	0	1	0	0
WEST GERMANY	11	12	6	10	4	3	3	4	4	2
UNKNOWN						3	4
TOTAL	2009	2100	1874	1840	1840	1257	1429	1364	1356	1243

NOTE. From Inside Behavior Analysis,12,21

Gender Equity and Diversity in the American Psychological Association (APA)

Skinner (1957) argued that successful groups engage in cultural design by examining not only their own previous practices but also those of other groups. The APA is the largest scientific and professional organization of psychologists in the United States, with over 118,000 members, and is therefore a useful comparison to the ABAI (American Psychological Association, 2019a, 2019b, 2019c). Its code of ethics includes statements regarding justice, respect for people’s rights and dignity, unfair discrimination, sexual and other forms of harassment, and sexual relationships with students and supervisees. The organization’s vision statement is germane: “a strong, diverse, and unified psychology that enhances knowledge and improves the human condition,” as is one of its guiding principles, to “champion diversity and inclusion,” defined as furthering “the understanding and appreciation of differences and (being) inclusive in everything we do” (American Psychological Association, 2019a, 2019b, 2019c). There are 28 active APA committees, among which are three with focuses germane to this discussion. Specifically, these three committees report on ethnic minority affairs, sexual orientation and gender diversity, and women in psychology. Included in each of these committee’s annual reports is a statement regarding the group’s diversity representation and training.

In 2014 (the most recent report available), the Committee on Ethnic Minority Affairs developed proposals for no less than two designated seats for members of color on every APA governance group, a curriculum and leadership experience manual for new members of color joining the leadership structure, and a mentorship project for minority members in governance positions. Additionally, the group added language to the 2002 “APA Multicultural Guidelines on Multicultural Education, Training, Research, Practice, and Organizational Change for Psychologists.” This document suggests research-based practices for working with people of color and training students through the lens of race and ethnicity (APA, Committee on Ethnic Minority Affairs, 2014).

The Committee on Sexual Orientation and Gender Diversity presented two awards in 2018 for outstanding achievement and hosted an event called the “Queering of APA” at the annual convention. The committee hosted webinars and drafted evidence-based guidelines related to psychological practice with transgender and nonbinary clients. Additionally, the group proposed revisions to numerous documents that offer practice guidelines with respect to intersex conditions, sexual orientation, and homosexuality. Of note, the committee developed a draft rationale for APA data collection on members’ sexual orientation and gender identity (APA, Committee on Sexual Orientation and Gender Diversity, 2018).

In 2012 (the most recent report available), the Committee on Women in Psychology hosted a leadership institute for women in psychology with emphases on political campaigning for APA governance roles. The group coordinates data collection on the level of representation of women; ethnic minorities; lesbian, gay, bisexual, and transgender individuals; and individuals with disabilities serving on and running for office in APA governance. To this end, the committee releases an official document, “Women in the APA,” providing data on the numbers of women in leadership roles of APA divisions, state associations, APA governance boards, and in the publications process as editors, associate editors, and reviewers (APA, Committee on Women in Psychology, 2012).

A visible structural element that reflects the APA’s response to concerns about gender disparities is the Women’s Programs Office, located in the Public Interest Directorate (Scarborough & Rutherford, 2018). Established in 1977, the office publishes information regarding women and psychology and provides staff support to the committees listed previously. Reports issued by the committees are distilled into public interest pamphlets and web pages that are published prominently on the APA website and distributed at conferences (American Psychological Association, 2019a, 2019b, 2019c).

Historically, election to fellow status in the APA has followed the same trajectory as in the ABAI. In 1980, 22.3% of the fellows elected were women, up from 8% in 1970. However, a decade later, in 1989, the total percentage of elected fellows had only risen to 23.4% (Scarborough & Rutherford, 2018). The most recent “Women in the APA” report supplied by the committee is dated 2006. According to this report, the overall proportion of women in the association was 53% compared to 47% men, yet the proportion of women inducted into the fellowship was only 26% (74% men). By contrast, in 2018, the APA released a list of 81 individuals that had been awarded fellow status for making unique, outstanding, and national contributions to the field. Fellows are nominated to the APA’s Board of Directors by an APA division and the board recommends fellow status to the association’s council of representatives, which formally appoints the slate. Of this cohort, 40 inductees were women, and 41 were men (American Psychological Association, 2018). This was an important new trend, because as of 2017, although 58% of the APA’s members were women, only 30% of the fellows were (APA, Committee on Women in Psychology, 2017).

Scarborough and Rutherford (2018) traced the history of women’s participation in governance and leadership roles within the APA from its inception in 1892. Similar to the ABAI, behavior analysis journals, and university behavior analysis programs (e.g., Li, Curiel, Pritchard, & Poling, 2018; Nosik et al., 2018), women were peripheral to the APA during its early years. Women gravitated to applied work, which was regarded as second rate, because academic careers were unavailable to them. This led to the establishment of 44 organizations by academic women, among which was the Association for Women in Psychology, the Women’s Programs Office, and the Division of Psychology of Women. The lattermost of these, APA Division 35, includes sections on Black women and feminist professional training; standing committees; task forces; award mechanisms; and a feminist, scientific, peer-reviewed journal (Scarborough & Rutherford, 2018).

Subsequent to the formation of these groups with their concerted effort toward training and mentoring women into academic and governance positions, the last two decades have seen notable rises in the median percentage of female doctoral faculty members (13%), internship supervisors (10%), department chairs (18%), and journal editors (14%; APA, Committee on Women in Psychology, 2017). But there has been only marginal increases in the overall percentage of women in full, associate, or assistant professor ranks (4%) and proportions of women on councils, boards, committees, and governance positions (8%; APA, Committee on Women in Psychology, 2017), offsetting the aforementioned gains. Of additional concern, the proportion of women editors of APA journals has increased only 6% in the last decade; in fact, fewer than one in five APA journal editors are women. Nevertheless, in comparison to other underrepresented groups in APA governance with fewer professional training opportunities, standing committees, task forces, and award mechanisms, there have been dramatic increases among women holding leadership positions (12% vs. 3% overall change in the distribution), according to the 2017 APA gender composition task force report (APA, Committee on Women in Psychology, 2017).

Taken together, gains in the distribution of women across multiple dashboard measures within the APA suggest the likelihood that women’s mobilization strategies have been effective. Task forces and committees compile numbers, draft reports, distribute public interest bulletins, and establish mechanisms for recognizing meritorious performance. These tactics make it difficult for controlling agencies within an organizational culture to remain unresponsive. Apparatuses such as these are used by other underrepresented groups in the APA, but in comparison, there is a relatively slim number of mechanisms that have emerged within the ABAI for promoting gender equity, other forms of equity, and diversity. If redressing imbalances in the composition of behavior-analytic journal editorships, academic leadership, and organizational governance is important, then lessons gleaned from the APA’s 125-year experiment in culture design are worthy of examination.

Behavioral Definitions of Equity and Diversity

From the data in Table 1, it seems fair to conclude that during the 1980s, the ABAI focused on expanding behavior analysis to other regions of the world. Given the historical and cultural context, our field showed itself responsive to selection pressures to grow along the lines of social evolution that were sustainable for the times. The practices the ABAI engaged in were suitable then. But the organization was perhaps not yet influenced by other determinants that now shape different definitions of diversity and inclusiveness. Skinner (1945) suggested that just like any other response, the use of a scientific term can shift in its stimulus function for an audience when emitted in novel contexts. Likewise, the same response can function to produce different outcomes for the speaker based on altered environmental determinants.

The behavior-analytic studies cited previously discuss equity and diversity as matters of importance to the field of behavior analysis. Yet neither equity nor diversity has been defined so as to facilitate a functional analysis of the issues, to my knowledge. In the following sections, I define each in technical terms and provide examples of how behavior analysts might manipulate variables to produce desired changes in these areas. In common language, I recommend that behavior analysts view diversity with respect to visibility at the highest levels of the enterprise of people from groups that are served by and compose the ranks of Board Certified Behavior Analysts worldwide. Then equity can be examined in terms of the level of the contribution and acknowledgment granted to each group so identified. To the extent that the discussion is limited to binary notions of male and female gender representation, establishing equity in the field is relatively straightforward, even if it is not at all simple. By contrast, to the extent that diversity is seen as a prism of multiple marginalized identities with multiple burdens, equitable representation and acknowledgment become more challenging. I will discuss equity first and then diversity.

Equity

Skinner (1945) suggested that every problem faced in science requires its own conceptualization and functional analysis. He was deeply influenced by Bridgman (1927), who argued that a physicist measuring the length of a small object defines “length” differently when the item is stationary compared to when it is in motion. Like Bridgman, Skinner suggested that new contexts lead to novel ways of defining problems with respect to the unique controlling variables involved.

Based on Skinner’s analysis of conventional behavior, it seems fitting to apply a more precise behavioral definition to the notion of equity. A useful approach related to Glenn’s (1988) metacontingency analysis would be to conceptualize equity as a cultural practice involving interlocking contingencies of reinforcement for making contributions relative to the composition of the group. When the output of pooled, reinforced actions by one group is discriminative for the pooled behaviors of the next group within an organization, behavior resulting from these interlocking contingencies can serve to generate a broader class of reinforcers for the whole enterprise, provided these actions promote sustainability and survival. In other words, if the behaviors of individuals in one group result in corresponding behaviors of individuals in another, we would say that they occur within interlocking contingencies of reinforcement. If this pattern is selected over time, the metacontingency will have maintained two sets of corresponding behaviors.

A few examples of equitable practices might include equalizing the gender, race, nationality, gender identity, and gender expression representations of (a) members officially recognized for their contributions and (b) candidates nominated for leadership positions proportionally to a conservative estimate regarding the number of individuals from these categories that could be selected. If the contributions of individuals from underrepresented groups result in recognition and nomination, and if this results in increased membership of other individuals from underrepresented groups, the metacontingency will have produced a broader class of reinforcers that serve the survival of the larger group, as well as the interests of its constituent groups.

Putting equity into practice based on balanced representation is not without complication. The study of interacting identities with respect to gender, gender expression, race, and other factors has been called intersectionality (Crenshaw, 1989). This critique of antidiscrimination efforts suggests that those with multiple identities are multiply burdened and that political action focused on any one factor minimizes the need for wholesale remediation. The issue for a science of behavior is that there is currently no empirical method of assessing the marginalizing effect of compound identities or the level of need for inclusion into what one hopes would be a more equitable enterprise. Moreover, Odom (2000) cautioned that with respect to women, whose status is already on the rise in behavior analysis, such practices could prompt some to ask whether a woman’s success is due to her gender. The same could be said of any other group included on the basis of its members’ identity, even if in conjunction with merit. However, given the number of groups that are currently underrepresented in academic compared to applied domains of behavior analysis, it seems that something along the lines of affirmative action could be beneficial, at least for a brief interval, to jump-start the behavior of celebrating diversity and promoting equitable opportunities for advancement. Initiatives established in the APA that ensure that seats in governance positions are reserved for women and members of color are examples of this type of affirmative action (APA, Committee on Women in Psychology, 2017).

Such conceptualization and organizational strategy allow for an analysis of the prediction and control that the strategy affords. To this end, an organization simply needs to try these behaviors to evaluate their effect on the growth and retention of members from underrepresented groups. Based on the evolutionary principles of variation, selection, and retention, it is possible that these practices would be selected and retained if they were shown to produce aggregate outcomes that contribute to the group’s survival.

Diversity

To promote diversity (including that related to gender, ethnicity, and other demographics) within behavior analysis, it would be useful to make more precise use of the metacontingency construct in alignment with Skinner’s suggestion that terms are a function of their controlling variables. To this end, four steps are detailed in the following paragraphs and outlined here. First, diversity can be viewed for our purposes here as a set of conditions that result in people with unique skill sets contributing to and gaining recognition within a group. Second, those conditions include practices of valuing, evoking, and reinforcing novel contributions of others with uncommon repertoires. Third, when evoking and reinforcing others’ contributions become practices that are reinforced by the verbal community, they may more likely be selected and retained. Fourth, a group is most likely to engage in these practices when its survival appears to depend on accessing the skills of those with uncommon repertoires. Fifth, make ample use of appetitive contingencies.

Behavior analysis is a field that has not generated substantial variation with respect to diversity practices so defined. Promising trends are notable in the literature and in the analysis of current data procured from the ABAI website. But these trends do not constitute variation. Rather, they seem to be the continuation of a pattern that is progressively better for women over time, but not reflective of variation in interlocked behavior across the organization that would result in contact with new, broader classes of reinforcement. What follows is a list of details regarding the recommendations listed previously for fostering such changes in our current repertoire.

1. Start with the necessary conditions for diversity and evaluate whether these are sufficient. Sciences usually seek to uncover the necessary and sufficient factors required to produce an outcome. Operant theory suggests that reinforcement is an important, if not always necessary, determinant of behavioral selection. The stimulus conditions under which people with unique skill sets will risk contributing ideas to an organization or manuscripts to a journal would likely involve a history of reinforcement in past or current environmental contexts. Because behavior analysis organizations such as the ABAI, JEAB, and JABA cannot control the historical environments from which people come to the field, it is important that these organizations evaluate their immediate environments with respect to motivative, discriminative, respondent, reinforcing, and punishing variables. If the leaders of these organizations do not look like you or talk like you, the environment would seem risky for venturing a contribution of novelty. In plain language, the optics matter. Organizations seeking contributions from members of diverse groups should demonstrate their commitment to diversity by inviting people from diverse backgrounds to apply for senior leadership positions and by recognizing the contributions of members with unique and uncommon skill sets. In this way, diversity and equity practices are dependent on each other. The aforementioned research suggests that these conditions are necessary. But in themselves, the research also suggests that these measures have not been sufficient for generating ongoing contributions from diverse populations, at least from women in behavior analysis. Additional factors seem to be needed. Specifically, task forces and special committees such as those established by the APA should be founded in the ABAI to draft reports that evaluate the status of underrepresented groups and propose new mechanisms for redressing imbalances.

2. Value, evoke, and reinforce novel contributions. Operant theory suggests that unreinforced behavior will not maintain, that behavior reinforced on very lean schedules is accompanied by emotional responding, and that conditions that are indiscriminable produce behavior that takes longer to reach stability. Behavior analysis organizations may find it beneficial to find ways to demonstrate their stated commitments to diversity and equity in ways similar to those enacted by the APA. This special issue of Behavior Analysis in Practice is one step: The announcement itself, posted on many social media sites, is discriminative for behavior that has a good chance of contacting reinforcement. The editor-in-chief’s statement that he will personally mentor contributors from diverse backgrounds wishing to submit manuscripts shows a commitment to valuing diversity (Tarbox, 2018). Should articles contributed by those from underrepresented groups get published in this issue, this reinforcement could produce the conditions under which more such behavior is evoked over time.

Another example of commitment to the diversity policy is a statement in the ABAI Annual Convention Call for Papers Handbook that reads,

The Association for Behavior Analysis International values diversity of all kinds in the composition of the convention program. We encourage submissions by presenters from diverse groups that have traditionally been underrepresented in the ABAI program and in behavioral research. In addition, presenters organizing symposia and panel discussions are encouraged to make efforts to include presenters from diverse and traditionally underrepresented groups. It is anticipated that enhancing the diversity of the ABAI convention program will not only make the program more equitable but also enhance the variety of perspectives represented, thereby enriching the convention program overall. (ABAI, 2019)

Commitment such as this can easily be evaluated for its effects on conference submissions. However, a litmus test will be in the continuation of efforts such as this in years to come and their resultant effect on submissions over time. Maintenance depends on ongoing periodic reinforcement opportunities. Generalization to the practices of other ABAI organizations such as JEAB, JABA, PBS, and TPR will be another litmus test. Should such novel practices lead to lasting changes in the demographic that is represented and acknowledged for its contributions, it would be fair to say that the adoption of these practices led to selection and retention across the broader culture of academic behavior analysis.

• 3.Reinforce leaders for valuing, evoking, and reinforcing novel contributions. Leaders who step outside their established repertoires and go out on a limb to value, evoke, and reinforce contributions from people with unique skill sets will continue to perform this way provided their actions contact reinforcement from the verbal community. Cultures that offer rich reinforcement for leaders who push the envelope can expect those leaders to continue to advance in this way. Individuals who introduce actions such as the ABAI Diversity Dialogue; the ABAI Diversity, Respect, and Inclusion Task Force; the current special issue of Behavior Analysis in Practice on diversity and equity; and the statement in the ABAIAnnualConvention Call for Papers Handbook need their actions to be promptly recognized and reinforced. It is up to the verbal community to observe the behaviors, make the observations visible to the community, and encourage others to reinforce these behaviors. In lay terms, it is up to us to shout it from the rooftops if we want others to hear about and reinforce these practices.
• 4.The selling point is survival. If people seem disinterested or unwilling to engage in behavior that values, evokes, and reinforces participation from a diverse pool, it may help to point out the contingencies. Species that do not respond to shifts in contingencies are unlikely to engage in behavior that will be optimally reinforced. Their survival prognoses are more limited unless variation that contacts new sources of reinforcement is occasioned. The same is true for organizations. Those whose leaders take risks by encouraging staff to try new things become organizations whose members sample reinforcement for novel behavior in the midst of the inevitable shifts in cultural contingencies. When individuals contact rich pools of reinforcement after engaging in new behavior, they tell others who begin trying out these new behaviors. The newly selected behaviors are likely to be retained over time if the reinforcement continues to be available and if other organizations that do not make these shifts in practice begin to show signs of weakness. Selling the verbal community on the need to reinforce leaders for taking these risks may require talking about the survival advantage conferred on groups that try out new behaviors.
• 5.Arrange appetitive contingencies. Because verbally able adults frame events in opposition without formal instruction or reinforcement (Hayes, Gifford, Wilson, Barnes-Holmes, & Healy, 2001), it is likely that talking about “survival advantages” will bring behavior under aversive control. That is, people might do what is suggested to avoid the group’s disintegration. However, under such negative reinforcement contingencies, the group is unlikely to continue reinforcing behavior that fosters diversity and equity unless the negative reinforcer is periodically presented. Because the negative reinforcer is a decline of the group, this would not be perceived as desirable. The alternative is to establish positive reinforcement for those who encourage leaders who value, evoke, and reinforce behavior emitted by people from diverse backgrounds. In other words, recognize those who recognize others. Reinforce those who use reinforcement. On a broader scale, it would be useful to establish appetitive consequences for groups within the wider verbal community that actively value, evoke, and reinforce practices that stimulate diversity and equity in our field.

Discussion

The previous recommendations rest on an interpretation of existing data from both psychology and behavior analysis related to diversity and equity. The interpretation is based on an examination of the motivative, respondent, discriminative, reinforcing, and punishing variables that would likely be the necessary and sufficient conditions for increasing diversity and equity across the field of behavior analysis. But it is important to note the as yet unstated assumption inherent in this article and the extant literature on this topic. The assumption is that diversity and equity are desirable. In truth, this has not been subjected to an empirical evaluation. The question as to whether the long-term survival of our scientific and service organizations would be more viable if we acted to abolish nondiverse, noninclusive practices remains unanswered. It is a question that has not yet been asked in any substantive empirical fashion. The question of whether the practices outlined in the aforementioned recommendations would be “good enough” to promote equity and diversity is related to the broader question of whether such aims are fundamentally “good.” My values and everything I have stood for in my life lead me to be optimistic about the answer to both these questions. As a student of B. F. Skinner, I am cautious to say that without data, this is speculation at best. These remain empirical questions, and ones that I do hope we try to answer if for no other reason than if it benefits our field, our data-based strategies could be a model for other disciplines.

Fifty-eight years ago, my mother was a medical student when she became pregnant with me. At that point, she did what other women were forced to do at that time in history: She left school. But my mother, who had escaped Hungary to come to the United States, refused to give up. She completed a doctorate in biochemistry and joined my father in conducting and authoring a series of endocrine experiments. Her work was related to the transfer of free fatty acids from obese mice across the placenta to make otherwise typically developing infants more susceptible to morbid weight gain. She and my father replicated and extended their work with rats, cats, and dogs. Today, epigenetics researchers are picking up the threads of my mother’s early efforts to identify critical pathways to childhood diabetes.

At conferences during those years, people routinely approached my parents and greeted them saying, “Hello, Dr. Szabo. Congratulations on your work. Hello, Mrs. Szabo. So nice to see you again.” After 18 years of being the principal investigator on over 30 empirical studies, my mother got tired of the abuse. She left research, went on to get a medical degree, and spent the rest of her life working as an emergency room physician.

Given that history, I still feel some shame when I recall how long it took me to answer the “surgeon” riddle. Perhaps you felt similarly awkward at your latency to solve the problem. But there is no need to allow feelings such as these to push us around. Likewise, we need not blame or feel overly ashamed of the current status of people from diverse backgrounds in our flagship organizations. All of us, including our leaders, are products of our histories. If this history is producing ineffective behavior, we can and ought to manipulate environmental variables to promote better stimulus control over our behavior with respect to gender, gender identity, gender expression, race, ethnicity, nationality, and ability. This is most likely important, given the increasing diversity among behavior analysis clinicians and clients (Fong & Tanaka, 2013).

Skinner (1945) suggested that when we are concerned about a problem, we should define it in terms that fit the context and evaluate our definition on the basis of how well it guides us to effective action in that context. A different set of definitions and variables is needed for the next context, and the next one after that. In the 1980s, the ABAI did much to promote a kind of diversity that was fitting for that era. New definitions of equity and diversity, with attention to their current controlling variables, are called for today. It is in the hope that we will identify and make use of these controlling relations that I have offered this discussion.