The complex chemistry of diversity and inclusion: a 30-year synthesis

According to recent data (WISE, NSERC 2017) the number of male and female students enrolled in upper level high school chemistry across Canada is approximately equal or slightly higher for female students.¹ However, the proportion of women drops significantly at the post-secondary level, resulting in their significant under-representation in physical and chemical sciences (∼32%).² Although disaggregated data are hard to find specifically for chemistry-based occupations, the proportion of women in scientific professions requiring a university education is 23% versus 21% for scientific occupations requiring a college education (data from 2011).³ We could not find data on the number of women currently (2020/2021) enrolled in graduate programs in chemistry and who are professors of chemistry in Canada. Nonetheless, data from the Almanac of the Canadian Association of University teachers 2013–2014 report that the proportion of chemistry female professors ranged from 27.6% (assistant professor) to 9.8% (full professor).⁴ A 2018 study by the Royal Society of Chemistry (UK) reports that 44% of undergraduates, 39% of Ph.D. students, and 9% of professors in chemistry in the UK identify as women.⁵ Moreover, demographic data collected by the Open Chemistry Collaborative in Diversity Equity (OXIDE) 2016–2017 report that over the total faculty in surveyed Ph.D.-granting chemistry departments, only 20% are women (27.1% at the assistant professor level, 26.7% at the associate professor level, and 15.5% at the full professor level).^6,7 These reports further highlight that talented women interested in an academic career are leaving the sector before their full contributions can be appreciated and call our attention to the numerous extrinsic barriers that limit women’s progression in the chemical sciences including, but not limited to, exclusionary culture, narrow understanding and application of the concept of merit (which feeds into inequitable funding structures), and persistent and pervasive gender-stereotypes in home and work life that disadvantage women. When these data are viewed through an intersectional lens (taking into account race, economic status, disability, etc.), barriers and exclusionary behaviours are almost certainly amplified.

Dr. Margaret-Ann Armour was well aware of the exclusionary culture of chemistry. In 2018, she described to Dr. Imogen Coe her experience of debating with her departmental colleague Dr. Gordon Freeman the importance of encouraging young women to enter educational pathways and pursue careers in chemistry and that “he would have none of it”. She was tireless in her efforts to challenge these attitudes. The reasons for, and solutions to, an exclusionary culture in Canadian chemistry over the last 30 years have only occasionally been explicitly addressed (also with the help of Dr. Armour) within the scientific literature or beyond.⁸ The myth of meritocracy within academia is very strongly held in Canada despite ample evidence that systemic biases are still present⁹ (e.g., gender-bias in grant review)¹⁰ and despite the commitment by all universities to integrate equity, diversity, and inclusion (EDI) principles into their structures.¹¹ However, there is abundant proof, derived from extensive scholarship across a range of disciplines, that inclusive climates in science and medicine create more innovative and effective working cultures where there is increased productivity, retention, and outputs as defined by the traditional metrics of productivity, innovation, etc.^12,13 Therefore, it behooves Canadian chemistry to pay closer attention to creating more inclusive environments. Indeed, this requires methodical and precise referencing to expertise and scholarship, as well as a systems approach to organizational cultural change. It also requires additional intellectual work and acquisition of core competencies by all members of the community^14,15 and accountability and consequences for those who fail to meet the expectations.¹⁶

A powerful educational tool used widely in a number of disciplines (e.g., medicine, law, business) that can help people understanding “real-life” situations and developing analytical and critical thinking skills is the use of case studies.¹⁷ In chemistry, case studies are sometimes used to support inductive rather than deductive learning (for example, “what would you do in this situation” or “what does the analysis of this case tell you” about a particular situation). Following this paradigm, we present here two case studies of chemists in Canada. They both published work in scientific journals relating to diversity issues in chemistry. Furthermore, we ask what these cases can tell us about attitudes, context, process, rigour, expertise, and excellence. Analysis of these case studies can help to identify possible solutions for community members and organizations towards increased understanding and integration of inclusive excellence.

Traditional metrics (numbers and journal impact factor, awards, citations, value of externally funded research grants, educational pathways, etc.) would suggest that both Drs. Freeman and Hudlický are talented chemists with excellent training, who have been recognized and rewarded by their peers. Although less explicit, the scientific community can infer from these metrics that the two researchers would be expected to possess the highest levels and standards of analytical skills, experimental rigour, and ethical behaviour. In both cases, however, the analysis of their contributions, reviewed and published outside their areas of expertise, rather shows a lack of rigour, analysis and, in the case of Freeman, ethics. Both of them are chemistry experts who made claims on a subject matter that is different than their own without adequate preparation or collaboration, as exemplified by lack of appropriate references, absence of data in support of arguments, failure to compare and contrast with the state of the literature and research in the relevant fields, and lack of human participants research ethics approval (Freeman).

In principle, both articles refer to interesting issues, such as the participation rate of women in chemistry, the impact of diversity initiatives on outputs and innovation, student conduct, and fraud in science. These are all substantial and important questions that are, indeed, worthy of academic research. Surely, it is possible for a physical scientist to acquire training and expertise (and even pursue research interests) in these areas, which are typical of the social sciences and humanities, through a thorough study of the field’s state-of-the-art. Alternatively, a viable option is to collaborate with established experts in the field whose expertise is demonstrated through previous work. The cases of Freeman or Hudlický, however, do not fulfill these requirements: in fact, there is no evidence for either of them having training or research experience in social sciences and behaviours or having ever conducted scholarly or pedagogical studies outside of chemical processes, nor do they appear to have sought out the assistance and insights of academic experts in social sciences, despite both being located at large institutions with excellent researchers in relevant areas. Indeed, over the last few decades, social sciences based scholarship into the social context and dynamics of the human endeavour of scientific research has grown extensively, and there is a vast body of literature on the relevant social and ethical topics. Anyone (including scientists) addressing these topics in their contributions would thus be expected to refer to such body of knowledge. Indeed, we expect ourselves and our students to reference correctly and thoroughly when writing a manuscript or preparing a presentation. Why should these cases be different? Clearly, neither Freeman’s nor Hudlický’s aforementioned pieces meet basic standards of scholarship. Despite Hudlický’s statement that “New ideas, and open mind and attention to detail will see us through to greater inventions and a return to high integrity”, his essay demonstrates an absence of scholarly evidence, references, or data in support of statements relating to the “preferential status” of certain groups and (or) individuals, preferential hiring (that overrides qualifications), changes in participation rates for “women and minorities” (without apparent awareness or recognition that many people belong to both groups), mandatory diversity seminars, and training. His comments about the Gordon Conference Power Hour reflect a complete misunderstanding of this component of the Gordon Conferences.³⁸ Undoubtedly, an established researcher as Hudlický knows well how to effectively present arguments, especially since other statements in his essay are adequately supported by references. Nevertheless, his lack of attention to a well-supported argument on EDI is exceptionally puzzling given his major thesis of a plea for a return to “high integrity” and experimental rigour in chemistry. Why, we might ask, has such un-rigorous work been allowed to slip through peer review? Could it be, as he warns us – with an apparent lack of self-awareness – because of the “preferential status” of certain groups? Interestingly, Seebach’s paper from 1990 is extraordinarily thorough in terms of supporting evidence for arguments with 14 figures, 33 schemes, and 568 references, many of which contain subreferences.²⁵

Although we discussed, so far, that neither Freeman’s nor Hudlický’s contributions can be considered of scholarly nature, it might be argued that such contributions are opinion pieces, which are, perhaps, not subject to the same standards as academic submissions (although editorials and opinion pieces are often peer reviewed in several science journals). Nonetheless, opinion pieces can be either informed or uninformed. We suggest that it is a reasonable expectation that even opinion pieces from established academics, with a track record of publications in peer reviewed scientific journals, be as thoroughly informed as possible and held to the high standards of academic rigour as for scientific submissions. Beyond this, if scholars chose to make their opinions publicly available (e.g., on blogs or in newspapers), it should come as no surprise that those pieces will be subject to rigorous analysis and critique by the broader community, even more so if the opinion is uninformed. Presumably, that is indeed the point of scholars presenting informed opinions and debating different points of view that are supported by evidence. Although Freeman is reported as rather enjoying his notoriety following the publication of his article,²¹ it appeared to come as a surprise to Hudlický that his essay was criticized and held to a high standard of scholarship. Moreover, he claims that this incident represents a “slide back to Calvinism and burnings at stake”.³⁹ However, perhaps there is another interpretation: many of us in academic science are aware that there are members of our community who possess opinions that differ from, or are even offensive to, those of the broader society, particularly with respect to complex issues such as diversity and inclusion. Publicly expressing opinions that “challenge orthodoxy” means that, especially if an academic, one should expect to have those opinions contested and counterargued. Undeniably, the scientific training and expectations of scientists at any level (from principal investigators to postdoctoral fellows, graduate students, and honours students), the interactions at conferences, and the evidence-based discussion of the scientific literature are all based on challenging each other with the ultimate goal to maintain the highest of standards. That is why we require our trainees to “defend” their honours or masters theses and doctoral dissertations. We expect their experimental work to be rigorous and ethical and their conclusions to be supported, extensively, with evidence and reliable, reproducible data. As academics, we have exacting and demanding intellectual expectations and standards. Hudlický was surprised that his contribution was found wanting in these regards. He claims to have suffered some professional impact, although he continues to enjoy the rare and considerable privilege of the protection of academic freedom and tenure, which allows him “to express opinions that challenge orthodoxy”.³⁷ This is not to say the privilege of expressing opinions that challenge orthodoxy is without consequences, either in academia or in the arena of public debate. Hudlický will not suffer economic consequences as an academic scientist. He will not lose his academic position or his research funding. He may not be asked to serve on hiring committees and he may be watched with concern in terms of his supervisory responsibilities, but he will never be subject to the bias and prejudice that have excluded others from pursuing a successful career in chemistry or other sciences. Ironically, relieving a faculty member of these types of duties without a concomitant rebalancing of their service work, may give them more time for research and, ultimately, accelerate their career even further. This will be accomplished mostly at the expense of those who do practice inclusive behaviours (and are often from historically excluded groups). It is therefore of crucial importance that accountability and consequences are carefully considered and equitable to those harmed.

Both the cases discussed above are classic examples of scientists self-identifying as “objective” and utterly failing to recognize they are human and, thus, fallible.⁴⁰ Both Freeman and Hudlický wandered into areas of scholarship where they had no acquired or demonstrated expertise and presented uninformed opinion with no apparent self-awareness. These case studies highlight the importance, especially for the dominant demographic^2,3,5,7 of scientists, to consciously and intentionally apply the same rigour to observation of themselves and the chemistry culture as they do to the practice of chemistry. Expertise, by definition, reflects knowledge of a specific domain: being an expert in organic synthesis, physical chemistry, polymer chemistry, etc., does not inherently make one an expert in anything else, particularly the social sciences. However, we can use scientific approaches to understanding ourselves (hence the value of reference to and engagement with social science research and scholars). We can, and should, analyze our own behaviours to identify and correct for bias, just as we would identify a machine in the lab or a step in a process that is inefficient or mis-calibrated, and correct for it to ensure that our outputs are correct. We illustrate this process by using an analogy in Fig. 1. If we consider the series of actions that need to be undertaken to identify a specific molecule of interest from a mixture of proteins, the usage of mis-calibrated equipment will inevitably result in inaccurate data. If we replace “molecule of interest” with “excellence” in Fig. 1, we can see that many of the processes currently used to determine “excellence” are demonstrably flawed (e.g., evaluation of the CV, committee evaluations, assessments of quality of contributions) due to mis-calibrations, which, when repeated at various stages in a process, are inevitably amplified. Ultimately, the faulty process will lead to significant errors in outputs (e.g., selection towards homogeneity, subjective success rates of grants or awards). Computer simulations of repeated rounds of bias show that over-representation of one demographic relative to another of up to 20% can occur.^41,42 Scientists pride themselves on their objectivity but, being human, are really one more component in a system that requires attentive calibration, similarly to what is applied to our laboratory equipment. We must continually calibrate ourselves in terms of our biases to ensure that we are assessing our science and our scientific community as rigorously and accurately as possible. This can be done through education and awareness raising, implicit bias training, and many other approaches that have been well described. We should also be intentional about ensuring the inclusion of talent from across the spectrum of humanity (Fig. 2). Humility and the ability to acknowledge error are hallmarks of great scientists, as described by Seebach himself whom, in his concluding remarks (which expressed the hope that his piece had not overpromised) quoted Teresa of Avila (1515–1582) “Teach me the glorious lesson that occasionally it is possible that I may be mistaken”.²⁵

It is clear that peer review failed in both cases. The overall scholarship for both submissions is not of a standard that warrants publication. Both pieces could have been rejected outright or, at a minimum, sent back for major revision by an editor. Both incidents reflect poorly on the peer review system and do not engender trust in the editorial process, particularly from traditionally marginalized scholars. Indeed, failure to promptly address such egregious abuse of the system hardly instills confidence in a structure that is routinely held up to be the gold standard in ensuring rigour in science. Responses from the reviewers would have been expected to comment on issues including, but not limited to: the clear lack of evidence of thoughtful analysis of the relevant published literature (either in 1990 or 2020); the failure to place issues of culture, diversity, and equity in the broader context or with reference to the established large body of scholarship on the subject by experts in these fields; the lack of adequate references, as well as evidence and data in support of statements; the poor (unethical) research design. The failure of scholarship in both cases is compounded by failure of peer review. Together, they represent the cumulative failure of the system. It is deeply problematic for the broader community when poorly supported scholarship is published in the scientific literature because it provides legitimacy to what is presented. Nonetheless, providing legitimacy to such intellectually inadequate contributions does undoubtedly a disservice to the broader community. This viewpoint was well presented in 1992 by Catherine Warren who stated, in a contribution to New Scientist following the publication of the Freeman article:

“in publishing a paper, a journal automatically legitimises its authors and endorses their work. When these ideas are then popularised, often deviating wildly from their source, they somehow retain this hallmark of ‘truth”’.⁴⁶

It is therefore a priority to ensure that appropriate training for reviewers is designed to promote an understanding of unconscious bias and how it can affect the peer review.

Much of the discussion, flamed by the fans of a whirring social media machine, focused outrage on the nature of the opinion expressed in the piece, which ran counter to current foci in universities on the power and potential of increased diversity and inclusion to drive innovation and leverage talent. Between 1990 and 2020, social media has significantly changed the nature of debate about science⁴⁷ and, especially, about science and social justice issues.⁴⁸ We propose that this is generally a good thing, as the voices of the marginalized and the work of a broader community can now be shared. The negative consequences, however, are a rush to response, which ultimately can weaken the argument in support of, for instance, a more diverse chemistry community. In fact, although moral outrage can certainly have positive social consequences, social media outrage may reduce the effectiveness of collective action and limit participation in the public square.⁴⁹ The reasons for and the solutions to an over-representation of one demographic in chemistry are complex.⁵ Moreover, the solutions require a systems approach, which is not easily presented in a barrage of short, sometimes anonymous, comments on the internet (e.g., a “Twitterstorm”) within 24 hours of a poorly researched article being published (although using social media to rapidly share deeply felt opinions about newly published science has become standard and should be expected). A thorough and substantive response might not be so well suited to the social media machine and should not be considered a mechanism for systems change. We need to embrace all forms of interaction (social media response, rapid editorial and institutional responses, disciplinary society commentaries, etc.) to ensure vigorous debate in appropriate settings and the highest standards of integrity and ethical behaviour, just as we expect from those whose opinions we might find uninformed and offensive.

This should be a requirement for every leader in academic science (principal investigator, chair, dean, VP, President), as well as editors of journals. Evidence of core competencies (i.e., actions that have contributed to inclusive excellence and training) should be provided by anyone serving in hiring, selection, tenure, promotion, or awards committee. Opportunities for professional development towards acquiring these core competencies are the responsibilities of leaders and organizations. Consequences for failing to demonstrate inclusive leadership should be clear and enforced (e.g., removal from the committee, removal from supervisory roles) and behaviors that are offensive and harmful need to be disciplined without hesitation and with hard measures (e.g., the case of Dr. Gianluigi Veglia at University of Minnesota Twin Cities who, despite being sanctioned for sexually harassing lab members, still holds three R01 awards from the NIH).⁵⁰

Those responsible for the supervision of trainees (undergraduate and graduate students, postdoctoral fellows, research assistants) should be educated in cultural competencies that enable and facilitate the professional development of trainees across cultures.⁵¹ This is a skill set that can be learned (based on evidence, relevant scholarship, and recommendations from experts in the field),⁵² repeated at regular intervals, and formally required of managers before they are allowed to supervise trainees.

Graduate programs in scientific disciplines should include, as core mandatory elements, the development of professional skills and core competencies in the application of the principles of equity, diversity, inclusion, and accessibility. Graduate programs can also provide education about the systemic barriers faced by individuals from historically excluded groups, as is now often the case in medical school where student learn about social determinants of health.⁵³ Practical actions include, for example, speaker seminars and workshops focusing on discrimination, gender and racial bias in the scientific landscape, and implicit bias and microaggression training modules to be completed within the first three months of joining the program.

Editors and editorial boards should actively and intentionally be working towards increased diversity in their reviewing and editorial boards. For scientific journals, this increased diversity may include broadening editorial boards and potential peer reviewers to include those with scholarship and background in areas such as science and society, science and gender studies, science and policy, organizational cultural change, etc.

Award structures need to recognize and reward inclusive excellence. Although the evaluation of “excellence of a researcher” and “merit of research” can still be informed by traditional metrics, there is now a much broader understanding of excellence and impact that should be factored in evaluations for external awards (including funding).⁵⁴ Traditional metrics should be viewed through an EDI lens to remove bias.⁵⁵ Contributions to, for instance, science policy and science communication can be highly significant and long-lasting at local and national levels. Indeed, NSERC Discovery Grants Peer review manual highlights that involvement in public outreach activities (e.g., organizing promotional events, taking on leadership positions in science outreach, and contributions to the promotion of equity, diversity and inclusion in the research enterprise) are to be included as evidence of stature in the field.⁵⁶ These types of assessment are not at odds with fundamental research and can be held to the same levels of rigour and ethical standards as other aspects of a scientist's work.