The ethics perspective of physics department chairs

Our 10 most striking findings from the department chair survey are listed in the box on page 24. In many cases, the responses from PhD-granting and bachelors-only programs were not significantly different. Departments granting bachelors degrees only, however, are more likely to report that all faculty must take ethics training, less likely to offer a semester-long course on ethics, and more likely to report zero ethics violations. A number of department chairs from bachelors-only programs pointed out that because the survey questions seemed to assume that graduate students were in their program, they were difficult to answer, but we do not believe that that significantly affected the results.

Although the particular questions and the response rates from the early-career and department chair surveys were quite different, several responses seem consistent. We focus, however, on the discordant messages from the two sets of results.

Figure 1 contrasts how department chairs and early-career scientists responded to the question of ethics training. Nearly half the chairs reported that their departments do not provide ethics training. For those who said that training is offered, most singled out webinars or online modules as the main type. By contrast, only 4% of the early-career APS members said that they did not have ethics training. The difference comes about because 40% of the early-career respondents said that they received their ethics training through their research group. That is potentially problematic because many ethical dilemmas, such as pressure to falsify data, come from research supervisors.

In the early-career survey, graduate students and other early-career participants were asked to recommend the kinds of trainings that they thought would be useful. Among the most suggested types were courses, workshops, seminars, and discussions on such topics as how to treat people, organize a research record, and write a paper. Opinions on the value of Web-based trainings were mixed: Some said that they were useful, whereas others pointed to the value of in-person discussions to help illustrate the gray areas in ethics decisions. Several survey participants emphasized the value of having trainings for department members at all levels, including faculty. Here is a sampling of responses from early-career APS members:

Training practices across departments differ. The departments granting only bachelors degrees are more likely to have mandated ethics training for all faculty, whereas departments granting PhDs are more likely to offer a semester-long ethics course. According to written comments by the department chairs, training can mean various things depending on the institution.

An unsettling picture of the reporting of ethics violations emerged when we compared the information from the department chair survey with that from the early-career survey. In response to the question “Have you ever observed or had personal knowledge of ethical violations during your time as a graduate student or as a postdoc?,” some 288 (38%) early-career respondents who witnessed an ethics violation knew where to report it; the remaining 469 (62%) did not. They wrote 517 comments and described 527 cases of ethics violations. Respondents said that they themselves had provided institutional reports on 108 of the cases and that a total of 131 reports were filed, whether by themselves or someone else. Out of the 527 cases of ethics violations, 60 were resolved in a way that the respondent thought was satisfactory. Thus, according to the responses, roughly only 20% of ethics violations had institutional reports, and of those, only a bit more than half were resolved well in young physicists’ eyes.

Figure 2 shows ethics-violations data from the department chair survey. The 622 chairs who responded—yielding a response rate of 81%—reported more than 600 reported violations in the past five years, although most chairs said that they knew of none. In comparison, only 25% of postdoc and graduate-student APS members responded to the society’s early-career survey: Early-career APS members, who had obtained their PhD within the past five years, said that 131 violations were reported to authorities.

We are not sure how to reconcile the disparate findings, but one plausible scenario is that about 500 ethics violations may have been reported to authorities by early-career APS members if the response rate of 25% was extrapolated to 100%. That number of violations is not far below the 600 that chairs said they knew about. Department chairs may receive violation reports from others beyond graduate students and postdocs, so having early-career members report 500 estimated violations could be reasonable.

If the interpretation is correct, the majority of chairs lack effective communication channels for people to feel safe reporting ethical dilemmas they face or have witnessed. The majority of cases go unreported. The 66% of chairs in figure 2 who said that they dealt with no ethics violations in the past five years should be concerned by the possibility that cases occurred but were not reported. Here is a characteristic description of what students face:

The perspectives from the chairs and early-career scientists on ethics education and the experience of ethics violations in physics departments reveal important disconnects. Many early-career respondents have specific views on what types of ethics-related education would be useful to them beyond what they obtain through research-group activities. The need for ethics education has not risen to the same level of urgency among department chairs. They may have an overly favorable view regarding ethical practices in their departments, which is understandable if students and postdocs do not report 80% of the infractions that they experience, observe, or have been told about. The lack of reporting is because of a fear of retaliation—for example, not getting a good letter of recommendation, not receiving a PhD in a timely fashion, or being given less recognition or help in research—and the concern that justice would not be served by reporting. According to one respondent:

We did not survey physics undergraduates, and filling that gap would produce a more complete perspective of physics departments. Five years ago, however, undergraduate women attending the APS Conferences for Undergraduate Women in Physics were surveyed regarding whether they had been sexually harassed in the context of their physics education. The results showed that a shocking 75% of them experienced sexual harassment, most of which was never formally reported.² The Federal Policy on Research Misconduct, put out by the Office of Science and Technology Policy in 2000, does not include sexual harassment as one of its unethical practices, which it defines as fabrication, falsification, and plagiarism. The American Geophysical Union, however, has elevated sexual harassment to the same level of misconduct as the federal policy,³ and APS has included extensive material on harassment and bias in its 2019 ethics guidelines.⁴ Mistreatment of people is now considered by almost all scientific institutions, universities, and national and industrial labs to be a serious ethical breach.

Many scientists argue that the lack of accountability in departments is problematic because it allows ethics infractions to be ignored, perpetrators to continue their harassment, and toxic environments to persist.⁵ In the survey, department chairs say that they refer reports of harassment to higher administration officials or Title IX offices, which places the investigations outside the department. That approach, to a certain extent, may be required by the college or university, but it can absolve a department chair from taking responsibility for holding anyone accountable for infractions. It may appear to those lower down in the academic hierarchy as a way of passing the buck.

Elevating harassment cases to other offices also discourages people from reporting ethics violations: One early-career APS member warned that “nothing will happen to address them.” In many institutions, the role of a department chair is temporary—often three to six years—and a chair might feel quite hesitant to bring one of their colleagues to account because of fears of retaliation. Although faculty members and department chairs are mandated to report sexual harassment, they are not required to inform authorities about research misconduct issues related to fabrication, falsification, and plagiarism.

Most department chairs were quite positive that APS could contribute considerably to ethics education. They were interested in APS webinars, case studies, and other materials that could start educational conversations about ethics in physics. Chairs may be eager to engage their department on ethics topics, but they feel that they lack resources and know-how.

Webinars and online ethics-training modules, however, may not be effective. Survey results showed that 55% of responding NSF graduate research fellows in several science and engineering disciplines felt that “mandatory ethics training left them unprepared to deal with ethical issues.”⁶ Because ethics violations almost always involve humans acting badly and the circumstances are not always clearly right and wrong, how to apply ethics training can be confusing. Designing effective training, such as discussion-based seminars and examination of real-life cases that can be widely shared, will help build the capacity across physics departments to prepare students more effectively.

As outlined in the ethics section of APS’s Effective Practices for Physics Programs guide, establishing a culture of ethical behavior with respect to teaching, learning, and research in a department is imperative.¹ The chapter includes detailed suggestions for getting buy-in from and raising the profile of ethics in the department. The early-career respondents emphasized the importance of good role models. For example:

In a hierarchical department, students may not feel comfortable talking to the chair about an infraction. It is often thought to be best practice to appoint a neutral party: someone who is well regarded, who may be from outside the department, and who can maintain confidentiality so that difficult issues can be explored and discussed. Some questions that may arise include the following: Did the person really commit a breach of ethics? Were they aware of the rules? Was the person exhibiting disrespectful behavior? Which options are available for dealing with the situation? Providing multiple lines of communication would also be valuable to accommodate diverse personalities and department structures.

Many institutions appoint an ombudsperson to serve as a neutral third party. In some cases the position provides a positive institutional contribution when problems arise. On the other hand, some worry that an ombudsperson has little power and therefore deflects problems. The respondents to our early-career survey provided some information on the two views. In the 517 comments described above on institutional responses to ethical failings, only two referenced an ombudsperson. In each of those cases, the respondent said that they were dissatisfied with the institutional response. Because of a lack of mention, we cautiously conclude, therefore, that the ombudsperson is often not effective in addressing ethics violations.

Is there any common thread binding together the 60 ethical problems that early-career APS members thought were resolved well? We have only 15 comments that provide more detail beyond that the matter was reported and resolved. In 12 of them, a group leader or PhD adviser led the resolution; in three, it was journal editors. A critical element for improving institutional response, therefore, may be the establishment of a trust network between and among students, faculty, postdocs, research advisers, and the department chair. Without trust, ethical concerns will not be reported. Developing trust demands open lines of communication and an institutional willingness to address problems.

Taken together, the results of the department chair and early-career surveys show an evolution of ethics awareness over the past two decades. Ethics education has improved: 95% of early-career respondents say they had some kind of ethics training—mostly in research groups or through institutionally mandated, and usually Web-based, tutorials. Many of the early-career comments, however, show that Web-based training is not valuable and does not equip the learners with the necessary tools to deal with real-life ethical issues. Physics departments should give more attention to ensuring that faculty, staff, and students have a clear and complete understanding of ethical principles, should consider other methods of teaching about ethics issues beyond Web-based modules, and should provide safe ways for people to raise concerns. Chairs listed positive actions that APS could take to help them build an ethical department culture. Most importantly, it is clear that dialog and open communication are essential to enabling everyone in the physics profession to continue to improve ethical behavior.

The authors thank the many people who have contributed to this work, especially the department chairs who responded in such great numbers to the most recent survey. Susan White and Julius Dollison of the American Institute of Physics’ Statistical Research Center administered the department chairs survey and analyzed the data; Jeanette Russo, APS corporate secretary, worked extensively on the surveys; and the APS Ethics Committee members contributed their ideas and suggestions. Michael Marder acknowledges many years of research support from the NSF Condensed Matter and Theory program for which work on ethics in physics has been part of its broader-impacts contribution.

Michael Marder is a professor of physics at the University of Texas at Austin. Frances Houle is a senior scientist at Lawrence Berkeley National Laboratory in California. Kate Kirby is CEO emerita of the American Physical Society.