Blog Viewer

Volume LIII, Number 2

What Can Research Institution Administrators Learn from Researcher Experiences of the COVID-19 pandemic? Insights from a Survey of NIH and NSF Researchers

Stephanie Solomon Cargill, PhD MSPH
Center for Health Care Ethics, Saint Louis University
Chair, Castle IRB

Samuel V. Bruton, PhD
University of Southern Mississippi

Tristan McIntosh, PhD
Bioethics Research Center, Washington University School of Medicine

Andrea Blake, MPPA
Troy University

Jaime O’Brien, MS
Center for Health Care Ethics, Saint Louis University

Alison L. Antes, PhD
Bioethics Research Center, Washington University School of Medicine

Abstract

The beginning of the COVID-19 pandemic required research institution administrators and researchers to make rapid and unprecedented decisions about whether research should continue and in what form. In the fall of 2020, we conducted a national survey of 930 federally-funded principal investigators (PIs) who continued in-person research during the early months of the COVID-19 pandemic. We investigated researcher perceptions of what shaped their choices about conducting in-person essential research and managing personnel during this time. Several quantitative survey questions asked about PI perceptions of institutional policies serving as barriers and facilitators to decision making, and these responses to qualitative questions were coded when they made reference to the role of administration in their decisions. Using this subset of data, we analyzed how administrative decisions at the institutional level affected downstream experiences of researchers. By jointly interpreting the quantitative and qualitative data, we identified 10 concrete lessons that can inform administrator decision making and best practices in preparation and response to crisis shutdowns of research if and when they happen in the future.

Keywords: Crisis planning, best practices, COVID-19, research shut-down, stakeholder feedback, resilience

Introduction

The global pandemic caused by COVID-19 produced an unprecedented crisis in research communities.  At the outset, it was generally accepted that certain research activities should be continued in-person, on-site during the pandemic, especially those activities directly related to the crisis and “essential” research activities. However, in the absence of a unified national public health policy, applicable directives and guidance for the continuance of on-site research activities were left to be determined by state and local governments and individual research institutions. Administrative leaders at research institutions were required to make countless unique and rapid decisions, including:  a) which research projects could continue on-site and which could not; b) what restrictions and protections should be required for on-site research; c) who would be allowed to conduct research on-site; and d) how to provide necessary resources to comply with these determinations. These decisions were made during times of constantly changing informational and public health landscapes during which information about risks and protections was often conflicting and evolving. 

Prior to the pandemic, some scholarship and guidance about effective administrative leadership in times of crisis did exist. This literature emphasizes the importance of collaboration and social interaction (Raelin, 2016, 2018) as well as preparation, readiness of emergency measures, response, recovery, and reconstruction (Kapucu & Van Wart, 2006). It also makes the case that the prioritization of physical safety must be followed closely by the emotional and psychological wellbeing of the stakeholders involved (Hutson & Johnson, 2016). This literature also brings attention to the importance of resilient critical infrastructure, defined as systems that deliver critical services during times of crisis. It acknowledges that sometimes resilient strategies, such as built-in redundancy and stockpiling of resources, can come at the cost of efficiency, at least in the short term (Carvalhaes et al., 2020; Chester & Allenby, 2019). Qualities of effective administrative leadership identified in the literature include benevolence, reliability, competence, honesty, and openness (Sutherland, 2017). Finally, flexibility has been shown to be the cornerstone of organizations that adapt well to crises (Deverell & Olsson, 2010). 

More specific guidance for research administrators preparing for a crisis came out prior to the pandemic in a report called “Strengthening Disaster Resilience of the Academic Biomedical Research Community: Protecting the Nation’s Investment” from the National Academies Press (Carlin et al., 2017). Authored by a multidisciplinary committee dedicated to the topic, this report focuses on the importance of resilience, defined as “the ability to prepare for, absorb, recover from, and more successfully adapt to adverse events (p.138).”[1] This report emphasizes the importance of planning for disasters ahead of time, ensuring the preservation of research data, samples, reagents, and animals, as well as anticipating funding needs and changing governance structures.

In spite of this existing scholarship, most academic research institutions, like the rest of the world, were caught unaware and unprepared by the pandemic. They were forced to respond quickly with little guidance either from the government or from existing shared or internal policies. Additionally, administrators faced competing demands from different directions, including public health authorities, public opinion, institutional leaders, funders, as well as the faculty, staff, and students working within research institutions. These demands put administrators in the difficult position of needing to develop policies that balanced safety and productivity, need for interaction and need for isolation, need for process and need for expediency, all while managing limited resources made even more limited by the pandemic.  At research institutions, one key stakeholder group was the leaders of research teams that required guidance and policy from their institutional administrations. These Principal Investigators (PIs) also had to deal with competing demands of their own, including the safety and the financial solvency of their labs, the differing needs of themselves and their research personnel, and the need to pivot their research quickly and with transparency. PIs had to make decisions based on institutional directives that would affect their research and research teams, including decisions about staffing and conducting on-site research.

To better understand the experience of PIs during this crucial stage of the pandemic, our team conducted a quantitative and qualitative survey of federally-funded PIs who were conducting “essential research” in the spring of 2020. The purpose of the survey was to examine the experiences of researchers who were permitted to continue doing on-site research during this time. While the focus of the survey was on the perspectives and decision making of researchers, these choices did not happen in a vacuum. We found that the ways PIs made decisions about the continuity of their research during the early months of the pandemic were influenced by the decisions made at the institutional level. Quantitative survey questions about facilitators and barriers to making decisions frequently referred to the role of institutional policies.  When asked to reflect on best practices and lessons they had learned in their own words, PIs’ comments frequently focused on the role of institutional-level practices.  

Through the analysis of this subset of survey data referencing administration, we derived 10 lessons to guide future administrative decisions should a need arise to quickly and significantly alter the structure and function of research again. These lessons encompass the entire chronology of decision making, from preparing for emergencies prior to their occurrence, to the process of developing and implementing plans, to ensuring compliance and revisiting plans in light of changing circumstances. These lessons focus on issues of timing, fair processes, communication, and the content of policies. Further, such lessons have potential to equip administrators with flexibility and resilience when crises occur that impact the research within their oversight. Taking these lessons seriously can empower administrators to navigate future emergencies with forethought, fairness, and effectiveness

Materials and Methods

In the fall of 2020, our research team was funded by the National Science Foundation to create and administer a survey to examine the experiences of federally-funded researchers whose research continued on-site during the initial onset of the pandemic in the U.S.  This study was reviewed and approved by the Washington University Institutional Review Board (IRB# 202006012). We presented survey respondents with an informational consent form before they proceeded to complete the survey. 

Using online, publicly accessible listings of federal grant awards—the National Science Foundation Award Search database and the NIH RePORTER database—we created a list of actively funded NIH and NSF PIs. An email was sent to all PIs in this database soliciting participation to complete an anonymous fifteen-minute online survey asking about their experience transitioning to on-site “essential research” while other research was shut down or transitioned to being done remotely. The surveys were administered between September and November 2020. Demographic information about survey respondents is provided in Table 1.

Table 1. Participant Demographics

The survey inquired about perceptions of the institutional guidance investigators received about essential research work, the types of research activities that were classified as “essential” at their institutions, and their priorities and perceived effectiveness when making difficult decisions about lab operations. Two team members (AA and TM) analyzed the quantitative survey responses using descriptive statistics (i.e., frequencies and percentages). Participants were able to skip items or end the survey at any time. We retained participants in the final sample who completed ≥ 60% of the survey items. We report the effective sample size for each separate item to account for missing responses.

The survey ended with demographic questions and three open-ended questions: 

1. What key lesson(s) do you think principal investigators or institutions can learn from your experience conducting research during the pandemic?
2. What are your biggest concerns about ramping up research activities?
3. Is there anything else you want to comment on?

Of the 930 PIs who completed surveys, open-ended comments were provided by 706 (75.8%) of them. Two other team members (SSC and SB), along with two graduate assistants (AB and JO), analyzed these qualitative open-ended responses.  Upon review, the content of the three questions overlapped to such a degree that the decision was made to analyze the responses as a whole by respondent, not by question.  Qualitative responses were uploaded into Dedoose qualitative analysis software, and the two team members (SSC and SB) inductively identified key themes and concepts from a subset of 50 PI responses to create a set of descriptive codes. Their codes were compared and reconciled. After clearly defining each code operationally in the codebook, SSC and SB trained the two graduate students on the codebook and how to apply the codes to participant responses. Next, all four team members deductively coded the same set of 50 PI responses using the codebook. The team compared, discussed, and reconciled their code application, reaching consensus on the codes. This process of consensus ensured shared agreement on appropriate code application before continuing to code the remaining responses. The remaining PI responses were divided among qualitative team members with two people coding each set of responses and continuing consensus meetings to ensure coding reliability. The pairs of coders met weekly to discuss and reach consensus on unclear participant responses and code applications.  

This paper focuses on the results of quantitative and qualitative survey data related to the role of administration in shaping PI experiences. The key quantitative survey items about institutional guidance and institutional-level factors were framed as barriers and facilitators to PIs’ decision making during that time. Barriers were defined as factors that made decisions more difficult. Facilitators were defined as factors that made decisions easier to make. The qualitative data included were coded as comments that referred to, or had direct implications for, administrators at the respondents’ research institutions.  Twenty-seven percent (190/706) of PIs had excerpts coded in this category. The four qualitative team members each inductively and independently read the narrative text to which this code was applied and identified lessons emerging from these codes. The team then compared and discussed these lessons, which revealed significant overlap. Based on these deliberations, the lessons were then refined into a list of 10 lessons learned for administrators.

Results

A subset of survey questions specifically asked PIs to reflect on the quality and impact of institutional administration on their decisions about staffing and conducting on-site research during the pandemic. These questions and their answers are reflected in Table 2.

Table 2. Quantitative Survey Results Related to Institutional Policies N=930²

These results demonstrate several important insights about administrative decisions during the COVID-19 pandemic. Among the PIs who responded to the survey, a little over half (57%) believed that their institutions handled the situation well and that this facilitated their decision making during the pandemic.  A similar percentage (54%) trusted their administrators and even more (64%) felt supported through the process. On the other hand, it appears that PIs negatively perceived specific aspects of administrative plans and guidance on their decisions. Forty-two percent of respondents struggled with guidance that was too vague, and 25% said they lacked guidance altogether. In direct contrast, a smaller percentage struggled with guidance that was too specific (15%). More concerning, 28% said institutional guidance was not relevant to their type of research. These conflicting experiences appear to put administrators in a bind, since the more general the guidance, the more likely this guidance will cover all research, but also the more likely it will be vague and lack specific details for researchers (which 58% said was a facilitator of quality decision making). In any case, the larger issue seemed to be conflicting guidance, which 39% of researchers experienced as a barrier to making good decisions.  

From these quantitative results, it is clear that administrators faced difficult challenges to meet the often conflicting and multifaceted needs of their faculty. This invites us to explore more deeply how researchers were affected by administrators’ choices. The open-ended answers provided by the faculty PIs provide this insight, and a picture of best practices emerges from the themes identified. Table 3 presents a summary of these lessons. In the remainder of this section, we will define and explain each lesson, including illustrative quotes. Last, in the Discussion section, we will provide concrete recommendations for how institutional leaders can respond to recommendations.

Table 3. Lessons Learned

STEP 1: PREPARING FOR A CRISIS

Lesson 1:  Create standing crisis plans.

For many PIs, their open-ended comments indicated the importance of preparation before a crisis like a pandemic strikes. Some commented that while some types of emergency plans exist at academic institutions (e.g., natural disaster, biosafety, and fire), it is important that broader crisis plans for research shutdowns exist as well. Although this lack of planning and disaster preparedness is not unique to research institutions, the type of preparation needed is unique (Wigginton et al., 2020).

PI 304: This time my university was feeling its way through the crisis, doing the best they could. I hope the Covid-19 experience will lead to a more structured response next time. 

PI 342: There was a significant deal of uncertainty as the university shut down in stages. A contingency plan for a pandemic situation, should it repeat, would greatly enhance the ability to plan ahead. 

PI 18: Broadly speaking, advance planning is critical. I think my institution did a good job issuing guidance and allowing us to reopen in a safe manner. 

PI 90: The institution needs to have multiple plans prepared for different types of work conducted on campus.

Besides general comments about the need for preparation, respondents suggested that these plans needed to cover several specific issues. They should include not only contingency plans for how to pivot in times of crisis, but also plans for when and how to exit from that plan as time progressed. 

PI 608: Although it’s yet another layer of administrative burden, it would be good to have a required 'shutdown plan', just like IACUC and biosafety etc. 

PI 487: Having a return to work policy is critical because it was never obvious when to return to more "normal" work. 

PI 201: It's still not entirely clear how we can ramp up and with whom. This required a lot of back-and-forth.

Lesson 2:  Create a stockpile of basic resources for emergencies.

In addition, adjusting research policies required resources, often resources that were not available when they were needed. These needed resources varied and included personal protective equipment (PPE), COVID testing, technology, lab funding (including bridge funding), transportation, and parking. By knowing what basic resources are necessary to conduct research in emergencies, administrators can avoid imposing “unfunded mandates.” We heard from researchers that institutional policies that required specific practices and behaviors (e.g., mask-wearing and sanitation) failed to be effective when researchers lacked the required resources (e.g., masks and sanitizer) to be able to comply. Feedback from researchers can provide administrators with insights about what types of resources may run short in times of crisis.

PPE and Testing

A third of respondents in the survey said that they lacked resources for protecting personnel. This indicates that institutions were caught unprepared and that, in the future, PPE resources could be stockpiled and prioritization of testing could be emphasized to prepare for future emergencies. 

PI 199: No available resources from the institution for PPE and sanitization. I requested many items through the proper channels within the institution and was not provided with anything. At each level I worked through, no one had any idea the scale of what we needed (one box of disposable masks for a department, including research students isn't sufficient) or who was in charge of making decisions. Initially had to be out of pocket. 

PI 687: While safety was an obvious concern, we had stocked up on PPE before the shutdown despite the institutional shortage. 

PI 1006: We were fortunate to have regular testing services for asymptomatic people returning to campus for work. 

PI 963: TESTING TESTING TESTING. We still have concerns about whether people can get tested, how they do so, how they find out their results quickly, what the requirements are for reporting positive cases, how we can address noncompliance with guidelines among local labs, etc. 

PI 103: Regular testing at my institution and the wider university campus has kept the Covid-19 infection rate low, 0.01%.

Technology and Tech Support

Respondents also commented on the need for technology and technological support when required or asked to work from home, or when coordinating work with those working from home. This meant that those working remotely need to have access to the hardware and software required to conduct their work as well as information technology (IT) support to assist with connectivity issues.

PI 961: Importance of infrastructure (home computers, microphones, cameras, Zoom etc). 

PI 1183: At my institution, IT personnel were furloughed or offered early retirement to the point that critical aspects of remote work received no IT support, or support was delayed to the point of being a significant barrier to work.

Transportation

An unexpected resource issue that arose stemmed from shifting transportation needs. Taking public transportation to and from work and on campuses was risky, especially early in the pandemic, and many people lacked safer modes of transportation or parking at their institutions. While some of this is outside of administrative control, devoting resources to subsidizing parking spaces near the university for those who, in normal times, could use public transportation, or providing safe shuttles and ride-sharing options that posed lesser risk, may be helpful options.

PI 598: Biggest concern: transportation—I think the university is taking proper precautions, but getting from home to the university is questionable. Not enough parking for everyone to drive, public transit mask-wearing and distancing is unreliable. This is making people nervous, even if not yet directly leading to known spread of the disease. 

PI 125: My biggest concern was that public transportation was and still is required for me to get to work. 

PI 466: Institutions need to think about transportation issues, offer free parking. 

PI 78: We are in a major city, and most students don't own cars, but most people now (including me) don't feel safe taking public transportation, so that is another complicating factor. 

PI 971: The institute provided free parking so employees who were relying on public transportation could more safely use their own vehicles.

Bridge Funding for Labs

Although an obligation shared between institutions and research funders, many respondents indicated that they struggled to comply with administrative requirements without supplemental funding to cover the interruption in work and the funding to get started again.

PI 938: It feels that the University administration and leadership doesn't really care about how much it cost individual labs to shut down and then restart. Labs took a huge blow in terms of financials, supplies, animals, and progress for careers, yet there is nothing tangible that has been provided to help. 

PI 651: Institutions pass on "unfunded mandates" by requiring distancing, reduced staffing, use of PPE etc. but don't provide tangible resources to accomplish them. 

PI 1140: My Institute, [removed for anonymity], provided a bridging the gap fund (research grant) to junior faculties. Further, my startup is fully available. I do not need to furlough any of my lab members, I actually have a slight salary raise, instead of salary cut. Our safety officers provided free supplies of surgical masks and ensured high standard for safety. I feel very supported.

STEP 2: DEVELOPING A PLAN 

Lesson 3: Incorporate diverse input.

Understanding that one size does not fit all, the administrative challenge of developing plans for pandemics and other crises is complex. There are numerous stakeholders and expectations, and administrators can be (and are) criticized for being too specific or too vague, too early or too late, too directive or too hands-off. While the quantitative results of the survey indicated this challenge of satisfying the needs of everyone, the qualitative comments gave insight into how to navigate this challenge better. A common view among PIs was the benefit of gaining input from diverse stakeholders to design a plan that is as responsive to the needs of the community as possible. 

PI 173: Administration needs to be willing to communicate and listen to plans; mine was. When the initial "essential work only" call came out, plant research was not considered.  I responded quickly and immediately new guidelines were issued that included plant stocks as essential. I also had a worker dealing with a child care issue and the administration listened carefully to the issue and worked with me to find a safe solution so that the worker could attend to critical plant stocks. 

PI 172: It was chaos. . . I think having a more diverse group of people with diverse needs in positions making decisions would have been extremely helpful. 

PI 695: Listen to the actual folks spending time in labs when coming out with institutional guidance, and show some empathy that you understand how impactful this is for junior faculty, or postdocs and graduate students. 

PI 291: Institutions need to communicate with stakeholders ranging from faculty to staff to students with more clarity and engage them in decision-making processes.

Lesson 4: Be flexible.

The ambivalence found in the quantitative data regarding vague or specific, general or narrow administrative guidance was borne out and explained in the qualitative data. Even after using input from stakeholders to design a plan that does not omit important considerations, another common theme among PI responses was that no plan can fit all research contexts without flexibility and the ability to be adapted on the ground. 

PI 161: One-size-fits all measures administered in a top-down fashion are ineffective at universities because of the diverse nature of the research in individual labs and the diverse makeups of the labs (size, percentage of undergrads vs grad students vs postdocs). 

PI 8: One of the glaring issues was an assumption that all research was lab or bench research. Initial guidance from my institution was reasonable, but very focused on biomedical and bench related lab research. That left many PIs feeling like the guidance was simultaneously too specific, too vague, and not appropriate. 

PI 28: Flexibility by the many institutions I work with to offer me a platform for arguing whom among my team should be allowed on-site and who should not (as well as why) was effective. I appreciated the implicit understanding that I know my team best and that I was given autonomy to set priorities.

Likewise, universal requirements, such as the maximum number of people allowed in a lab, are not responsive to the different contexts on the ground and their implications for safety.

PI 575: It would have been nice if campus leaders actually walking (sic) through different laboratories and realized the complexity of different spaces, the needs for overlapping personnel for projects, etc. 

PI 497: I feel that most of the decisions at my university were based on arbitrary metrics. For instance, we were limited to a percentage of the workforce being able to come into lab, taking no account of the actual space available for distancing.

Finally, certain types of researchers felt that their unique circumstances were entirely disregarded by administrative policies. Most notably, respondents who were field researchers, plant researchers, or those who conducted qualitative research felt left out.

PI 8: Institutions need to have a broader understanding of the research being conducted—not just the most common type of research.  

PI 132: Guidelines must encompass the wide range of research that is conducted, not just traditional wet lab bench research. 

PI 221: Field research is fundamentally different from lab/on campus research but rare enough on my campus that the default guidance wasn't always applicable and I had to push to get the relevant permission/protocols.

Lesson 5: Trust, do not micromanage.

The significance of discretion and autonomy to a large percentage of respondents was clear in the quantitative results of our study. In their open-ended comments, many faculty PIs noted the diversity in lab personnel makeup, research functions, and various other socio-structural elements that are department-specific. While institutional-level policy and guidance was needed, PIs recommend that certain decisions should be left up to the faculty who know their local context the best. Sixty-five percent of respondents experienced being trusted by their administration as a facilitator to making good decisions. Conversely, the term “micromanage” came up frequently as a critique.

PI 695: Trust your faculty to make good decisions. Don’t start to micromanage them weeks and months after we figured out how to balance personal safety with some level of continued productivity (in that order). 

PI 128: Institutions should learn that their decisions should be logical and follow general established guidelines from CDC etc., leaving specifics to PIs, rather than trying micromanaging everything. 

PI 633: Allow PIs to make decisions about essential personnel as the PI (sic) know their lab members best. Blanket edicts that everyone has to work from home does (sic) not serve PIs whose research depends upon on-site performance. 

PI 145: Important to trust PIs and give them a "voice" in managing their own laboratory.

Lesson 6: Prioritize general safety along with crisis-specific safety.

Quantitative survey responses reflected the importance of researchers’ perceptions of support and trustworthiness of their administration. Suspicion that institutional administrators were making decisions that did not prioritize safety came across in the qualitative data and likely undermined perceptions of support and trustworthiness. While many respondents experienced their institutions as prioritizing the safety of researchers, several respondents commented that the plans made by their administration appeared to prioritize profits over safety measures.

PI 527: My institution made safe parameters that I felt I could work in. 

PI 1195: My institution has prioritized buildings and endowments over people, especially young people. I think it should be the opposite. 

PI 503: Allowing people to safely work through the shutdown led my institution to apply for $150M in grants. 

PI 447: We went back to work too soon—for the sake of scientific productivity.  Now with the fall/winter rise of cases in the Midwest the situation is, if anything, far less safe than it was last spring. There is no indication that my institution will compromise and support further reductions in laboratory staffing. Sadly, there will likely need to be casualties for that to happen.

In addition, several comments indicated that in order to comply with COVID-19 safety measures, PIs were forced to put themselves and others at greater risk. This indicates that crisis policies should not lose sight of the fact that although one type of risk is increased, others still exist and should be weighed accordingly.

PI 181: COVID-19 safety has usurped other safety concerns in some instance: for example, we are still following strict 'critical' research only policies, and are limited to one person per vehicle. We had a serious accident recently that totaled a vehicle; thankfully the driver was fine. However, the incident may not have occurred if people had been traveling together. I will now be asking for permission for two people to travel together for travel safety reasons (with masks on, in front/back seat). 

PI 251: We had some safety concerns in the shop. We wanted to limit spaces to one person, but for safety reasons, we don't want people using cutting tools, like a table saw, when they are alone. 

PI 29: I was concerned about potentially dangerous work conditions due to heat in the greenhouse and alternatives to mask wearing when personnel work alone.

STEP 3: IMPLEMENTING A PLAN 

Lesson 7:  Be attuned to timing.

The timing of guidance is a distinct challenge during pandemics. Institutions need to communicate plans in a manner that makes sense with the timeline of unfolding events. While “not too early, not too late” seems to be the trite and facile answer, responses that addressed timing suggest possible strategies for administrators to determine when to convey information. 

PI 148: Institutions need to make quick, clear, and definitive decisions and keep PIs informed as the process of a safe campus return is evaluated.

Those who had concerns about guidance being too quick often took issue with premature directives that were irreversible and detrimental to research.

PI 675: My biggest regret is complying with the institutional requests to severely cull our mouse colony. This has done irreparable damage to multiple investigators' research programs, setting back progress of many trainees for months if not years, while having minimal benefit to the institution during the shutdown. Those PIs who chose to disregard these requests are much happier about their decisions.

Many PI criticisms about guidance being too quick were about the amount of time granted to be able to comply with institutional guidance.

PI 938: Need much more lead time for a shutdown operation than was given. Experiments were cut short or left unfinished. 

PI 181: Provide clear guidance and definitions and sufficient time to complete paperwork and receive permissions before restrictions go into effect.

Positive experiences reflected constant communication as well as time to prepare for required actions.

PI 1086: I was given an unofficial heads up about a week in advance to accept a shutdown of the labs. This gave the chance to wind down experiments, store/secure critical samples, and start making a plan for personnel. Early warning (even if just a few days) made a big difference in preparing. Ultimately they shut the labs several days before they had planned to close, but we were prepared because of the early and frequent communications.

On the other hand, several other comments noted that guidance was issued too late. PIs who commented on the guidance itself focused on the delay in mask mandates and other safety policies already recommended by the CDC, but not legislated yet by their state.

PI 667: We also did not have a mask mandate (or any policy on masks) in early to middle days in the spring, as the state did not have one yet. This caused confusion in the lab and led to several personnel arguments about safety and masks. We had one that didn't want to wear a mask at all until the governor declared the mandate.

Again, more comments about being too late focused on the inability to comply without adequate time to plan. 

PI 861: I feel their initial response to the pandemic was disorganized and immediate. Plans for return to research were not revealed until 7 days prior to a return and resulted in a great deal of stress and hassle for all involved.

Lesson 8: Communicate the plan clearly and consistently.

As indicated by the quantitative results, administrative guidance frequently lacked key information and was perceived as vague and inconsistent. Since trust is enhanced with transparency, in times of uncertainty and changing guidance, it is even more important to keep communication open and continuous. The clear impact of this strategy manifested in the starkly different experiences of PIs who said their institutions communicated well versus those who had more negative experiences.

PI 480: Our university did a really good job communicating. They told us in early March what options they were considering and what evidence they would use to make a decision and they gave us a deadline for making the decision. University administration stuck to that plan. Once we were fully stay at home with the exception of essential workers, they hosted a weekly zoom session to communicate updates and answer questions. The leader of these meetings was always clear about what was known and unknown and always followed up with updated information on the unknowns the following week. 

PI 334: My institution did a fantastic job. They clearly explained the decision to shut down in accordance with state guidelines, provided us with lots of information about what was required of us and the potential impact on research programs, processed requests to perform essential duties very quickly and worked with each laboratory and research program. Given the circumstances, I learned a lot about the values of my research institution.

Some, unfortunately, had the opposite experience.

PI 37: Although I generally trust the leadership at my institution, it is unclear what factors are guiding their decisions, and they have not clearly communicated with the faculty.  

PI 483: The institutional leaders who should have been taking the lead on providing guidance were all but silent at my institution. I was happy to have a department chairman who provided a clear and consistent message. 

PI 367: We received conflicting messages from leadership about the need to comply with essential research orders (institutional leadership) and the need to maintain high productivity (department chair level). I think this mixed messaging has added to the stress of faculty members at my institution, particularly those with child care issues. A more consistent message from the top down and acknowledgment of the stress that faculty are facing would have been welcomed and appreciated. 

Lesson 9: Adapt to changing circumstances and learn along the way.

Rather than inconsistency, many PIs appreciated policy changes which were perceived as welcome adaptations to changing circumstances. The plans enacted by institutions need to be able to adjust and reflect changes in the public environment. As time goes on, developments may require shifting of institutional guidance. The need for clear communication in how to adapt while also maintaining safety is key.

PI 172: I think that my institution, in particular, was scrambling to come up with guidelines that didn't have to be changed and in the end were too vague and no one really knew how to enforce them. I [also] think having flexible working plans codified into the university policy in the first place would have made this all better. 

PI 312: Institutions/administrators need to be clear about the logic/reasoning used when they make rules. They should quickly reconsider their rules and find alternatives if it becomes clear that rules instituted to accomplish one goal, end up penalizing a subset of the group. 

PI 667: Guidance needs to be a constantly evolving thing during a pandemic, as a pandemic is also an evolving thing. During the early days of the pandemic we also had protests occurring and there was no guidance as to how to address essential employees/students who were attending mass gatherings until weeks afterwards.

Lesson 10: Enforce policies fairly.

Some PIs remarked with a high level of vitriol that different labs and different members of the community complied with administrative guidance and requirements to varying degrees. This resulted in not only disjointed safety efforts, but also uneven impacts on productivity and feelings of unfairness. 

PI 596: Uneven following of directives led to some investigators continuing to work, whereas others completely stopped their work. This may lead to substantial differences in productivity and ability to submit grant proposals. 

PI 687: As with other crises, there were people who clearly abused the “essential research” designation to get ahead while others shut down. 

PI 285: Be aware of biases in the interpretation of the guidelines lead to disparities in who was approved to work.

While PIs indicated that they do not want to be micromanaged (Lesson 5), in most cases they did appear to expect their administrations to consistently enforce the policies that they developed to prevent this perceived unfairness.

PI 447: The fact that the other labs in the building were not enforcing masking and social distancing. The administration did not enforce the guidelines. There were cases of COVID spread. My lab folks were very frustrated with the people in these other labs. 

PI 667: . . .the institution took a "look the other way" attitude. There were some labs working at full capacity on regular research and it was encouraged by their department chairs (this was told to me secondhand). While other labs shut down completely. This creates unfairness at the institution and unfairness applied universally.

This indicates that simply developing policies is not enough; rather, these policies need to be consistently enforced to be perceived as fair and equitable. This is especially important given the competitive environment of research. Our open-ended comments indicate that those who were already compliant would welcome this enforcement and not perceive it as an overreach since the behavior of those outside of their labs (other research groups, staff, etc.) is outside of their control.

Discussion

The results of our survey indicated many ways that institutional leadership could, and often did, address the pandemic well in spite of its unexpected nature and severity. Many of the comments about successful administration referred to processes suggested in pre-existing leadership literature such as advanced planning, communication, and transparency, but made them much more concrete and actionable. While researcher experiences provide only one perspective about the decisions at research institutions, their lessons should be interpreted within the wider literature on leadership and adaptability produced both prior to and emerging during this pandemic. This evidence-informed interpretation of PI experiences can provide administrative leaders with best practices for all stages of a crisis. Below we discuss some of the ways their lessons dovetail with existing literature to provide concrete recommendations that can prepare research institutions for future pandemics and other crises that disrupt research.

Respondents to the survey called for a structured, institutional response based on a pre-existing contingency plan with nuance to address the types of research being conducted at the institution. This plan should not only include how and when to halt in-person research but also how and when to start it again. They also emphasized the importance of institutional resources to support such a plan, including resources for PPE and testing, technology support, transportation, and bridge funding for labs. 

Guidance has also emerged on how to plan for when and how to ramp research back up and stabilize after a pandemic’s risk has eased. Six U.S. research universities collaborated to share guidance documents for their phased ramping up plans (Wigginton et al., 2020). These policies have overlapping recommendations for adhering to public health guidelines, prioritizing the health and safety of the workforce, and establishing fair and transparent processes. Likewise, other universities and university systems have collaborated on phased ramp-up research approaches, many based on the Principles and Framework Guiding a Phased Approach to Restarting University Research Activity developed by University of California, Berkeley in partnership with Vice Chancellors for Research and Vice Presidents for Research from the University of Washington; University of California, Los Angeles; University of California, San Francisco; and Stanford University and borrowing liberally from planning documents at many other institutions (Research Recovery Committee, 2020).

When discussing how institutional leaders should develop a plan, researchers’ comments put flexibility, transparency, and input from stakeholders as central. This markedly reflects the current literature in leadership as well as the aforementioned literature on research institution resiliency, both of which make specific recommendations that reflect researchers’ insights. 

The National Academies report mentioned in the introduction emphasizes that: 

The principal investigator (PI) is the central focus of research efforts. PIs and their laboratory members are in the best position to understand the specialized needs of their specific research. To promote a resilient laboratory and protect their research, PIs should actively engage in disaster planning with institutional leadership (Carlin et al., 2017, p. 9). 

Likewise, the concept of “collective leadership” has risen to prominence in leadership literature (Cullen et al., 2012; Day et al., 2004; Friedrich et al., 2009; Gronn, 2002). Everyone within the organization may not be a leader in the sense of having a formal leadership position or title, but institutions can be led effectively by engaging the relevant expertise of multiple individuals at appropriate times. This approach requires sharing of information, collaboration, and joint decision making in order to make best use of the diverse skills and expertise of individuals (Friedrich et al., 2009). While this type of group leadership and decision making can be challenging when time is of the essence, if integrated into the planning phase, it is much more feasible. Advanced input from stakeholders would allow the plan to reflect their diverse situations and concerns. Doing so would also address the problem of researchers feeling that their particular circumstances were neglected by institutional plans. Likewise, collective decision making enables transparency and allows for expectations to be set appropriately such that details about the timing, type, and priorities of the plan would be known and agreed upon ahead of time. This would also enable local research teams to develop their own plans, which fit within the plans of the institution as a whole. These types of transparent and trustworthy processes have been shown to increase compliance (Martinson et al., 2006; Schweitzer & Gibson, 2008; Tyler & Blader, 2003). 

While the details of a pre-developed plan would have to leave out the specifics of a particular emergency, many of the predictable crisis issues such as financial planning, communication plans, priorities and goals of the plan, as well as appeals and enforcement of the plan could be specified in advance and adapted to specific circumstances. Likewise, the process for modifying the plan (e.g., who would be involved, when and how it would be done) could be specified in advance.

Many of the lessons about implementing a plan should follow directly from the first two steps, if done properly. Planning ahead when there is time to plan would enable administrators to cultivate stakeholder buy-in, as well as a coordinated process, for navigating pandemics and other emergencies. Pre-articulated communication plans that are formed collaboratively with numerous stakeholders should determine what will be communicated and when and should set expectations for researchers accordingly (Giorgini & Mumford, 2013; Marta et al., 2005; Mumford et al., 2017). Pre-formed chains of communication between what the NAS report calls “parallel authority streams” would avoid the challenge of researchers receiving different messages from different parts of the institution. Advanced planning would enable researchers to know how to obtain more resources, on what timeline, and with whom to communicate. Business continuity plans paired with a robust, stakeholder-informed understanding of intricacies within different research areas would enable administrators to build in specification and avoid reliance on a “one-size-fits-all” plan. Finally, while being adaptable and flexible amid changing circumstances is a challenging aspect of plan development, having a process to expediently assess changing circumstances with pre-articulated goals and priorities would enable this process to be done more consistently and less haphazardly. 

The thorny issue of fair enforcement could be addressed in two ways. First, involving research stakeholders in the planning process increases buy-in as well as the perception that the rules and policies are fair. Empirical research shows that people tend to violate norms and rules more frequently when they believe them to be unfair (Martinson et al., 2006; Schweitzer & Gibson, 2008; Tyler & Blader, 2003). Second, how an institution enforces its plan, through incentives or penalties, locally or from the top-down, through increased documentation or other surveillance mechanisms, can all be determined ahead of time. Giving researchers a voice in how the plan will be enforced makes it more likely that the plan will be responsive to their needs while simultaneously reducing the need for punitive approaches to enforcement (Morrison, 2011). 

Conclusion

Our findings have direct and useful implications for research institution administrators who want to better prepare for future emergencies when research operations are disrupted. Effective research administration during crises requires attention to the diverse needs of research faculty and staff members, as well as the incorporation of their perspectives into planning and the execution of plans. By examining surveys of researchers during the pandemic, it is clear that research institutions were faced with a daunting task of addressing diverse needs of faculty, both in terms of their professional and their personal situations. Our survey provides insights into potential blind spots, such as transportation challenges and broader (i.e. non-crisis) safety issues and provides guidance into how the conflicting and multifaceted needs of researchers can be planned for and accommodated in a practical and effective way. The pandemic has shown both how resilient, yet vulnerable, researchers and research institutions are during times of crisis. The complete impact of the COVID-19 pandemic on researchers and the research enterprise has not yet been fully realized. It is the obligation of administrators and researchers alike to learn from this experience and prepare so that when the unexpected happens, they can mitigate the harms that we are only now beginning to understand. Administrative leaders at academic institutions hold great power and thus bear great responsibility to thoughtfully lead their institutions through future crises. We hope that this discussion can provide guidance toward that end

Authors’ Note 

The primary contact for this article is Stephanie Solomon Cargill, who is also the primary author and takes responsibility for all the content within.  She can be reached by email at stephanie.cargill@slu.edu or by phone at (314)977-1061.  Dr Burton, Ms. Blake, and Ms. O’Brien contributed to the data coding and analysis and to the writing of the manuscript.  Dr. McIntosh and Dr. Antes (the PI of the grant) provided crucial feedback on the manuscript.  This study was funded by the National Science Foundation RAPID Reward mechanism (McIntosh, Antes, PI #2031851). We wish to thank the principal investigators and personnel who filled out our surveys as well as those who volunteered to provide feedback on the initial drafts of our surveys. TM and AA gratefully thank Kari Baldwin and Matt Wroblewski for their assistance with this research. AB thanks her former employer, the University of Southern Mississippi, under which this work was done as a graduate assistant.

References

 Carlin, E., Brown, L., & Benjamin, G. C. (2017). Strengthening the disaster resilience of the academic biomedical research community: Protecting the nation’s investment. National Academies Press.

Carvalhaes, T., Markolf, S., Helmrich, A., Kim, Y., Li, R., Natarajan, M., Bondank, E., Ahmad, N., & Chester, M. (2020). COVID-19 as a harbinger of transforming infrastructure resilience. Frontiers in Built Environment, 6, 148. https://doi.org/10.3389/fbuil.2020.00148

Chester, M. V., & Allenby, B. (2019). Toward adaptive infrastructure: Flexibility and agility in a non-stationarity age. Sustainable and Resilient Infrastructure, 4(4), 173–191. https://doi.org/10.1080/23789689.2017.1416846

Cullen, K. L., Palus, C. J., Chrobot-Mason, D., & Appaneal, C. (2012). Getting to “we”: Collective leadership development. Industrial and Organizational Psychology, 5(4), 428–432. https://doi.org/10.1111/j.1754-9434.2012.01475.x

Day, D. V., Gronn, P., & Salas, E. (2004). Leadership capacity in teams. The Leadership Quarterly, 15(6), 857–880. https://doi.org/10.1016/j.leaqua.2004.09.001

Deverell, E., & Olsson, E.-K. (2010). Organizational culture effects on strategy and adaptability in crisis management. Risk Management, 12(2), 116–134. https://doi.org/10.1057/rm.2009.18

Friedrich, T. L., Vessey, W. B., Schuelke, M. J., Ruark, G. A., & Mumford, M. D. (2009). A framework for understanding collective leadership: The selective utilization of leader and team expertise within networks. The Leadership Quarterly, 20(6), 933–958. https://doi.org/10.1016/j.leaqua.2009.09.008

Giorgini, V., & Mumford, M. D. (2013). Backup plans and creative problem-solving: Effects of causal, error, and resource processing. The International Journal of Creativity and Problem Solving, 23(2), 121–147.

Gronn, P. (2002). Distributed leadership as a unit of analysis. The Leadership Quarterly, 13(4), 423–451. https://doi.org/10.1016/S1048-9843(02)00120-0

Hutson, H., & Johnson, M. (2016). Navigating an organizational crisis: When leadership matters most. ABC-CLIO.

Kapucu, N., & Van Wart, M. (2006). The evolving role of the public sector in managing catastrophic disasters: Lessons learned. Administration & Society, 38(3), 279–308. https://doi.org/10.1177/0095399706289718

Marta, S., Leritz, L. E., & Mumford, M. D. (2005). Leadership skills and the group performance: Situational demands, behavioral requirements, and planning. The Leadership Quarterly, 16(1), 97–120. https://doi.org/10.1016/j.leaqua.2004.04.004

Martinson, B. C., Anderson, M. S., Crain, A. L., & De Vries, R. (2006). Scientists’ perceptions of organizational justice and self-reported misbehaviors. Journal of Empirical Research on Human Research Ethics, 1(1), 51–66. https://doi.org/10.1525/jer.2006.1.1.51

Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. Academy of Management Annals, 5(1), 373–412. https://doi.org/10.5465/19416520.2011.574506

Mumford, M. D., Todd, E. M., Higgs, C., & McIntosh, T. (2017). Cognitive skills and leadership performance: The nine critical skills. The Leadership Quarterly, 28(1), 24–39. https://doi.org/10.1016/j.leaqua.2016.10.012

Raelin, J. A. (2016). Imagine there are no leaders: Reframing leadership as collaborative agency. Leadership, 12(2), 131–158. https://doi.org/10.1177/1742715014558076

Raelin, J. A. (2018). What are you afraid of: Collective leadership and its learning implications. Management Learning, 49(1), 59–66. https://doi.org/10.1177/1350507617729974

Research Recovery Committee. (2020). UC Berkeley guidance for the ramp-up/ramp-down of research activities during the COVID-19 Pandemic. UC Berkeley. https://vcresearch.berkeley.edu/sites/default/files/inline-files/Final_Research_Recovery_Committee_Report_June_16%2C_2020-UCB_Guidance_for_Ramp-Up_Ramp-Down_of_Research_Activities.pdf

Schweitzer, M. E., & Gibson, D. E. (2008). Fairness, feelings, and ethical decision-making: Consequences of violating community standards of fairness. Journal of Business Ethics, 77(3), 287–301. https://doi.org/10.1007/s10551-007-9350-3

Sutherland, I. E. (2017). Learning and growing: Trust, leadership, and response to crisis. Journal of Educational Administration 55(1), 2-17. https://doi.org/10.1108/JEA-10-2015-0097

Tyler, T. R., & Blader, S. L. (2003). The group engagement model: Procedural justice, social identity, and cooperative behavior. Personality and Social Psychology Review, 7(4), 349–361. https://doi.org/10.1207/S15327957PSPR0704_07

Wigginton, N. S., Cunningham, R. M., Katz, R. H., Lidstrom, M. E., Moler, K.-A., Wirtz, D., & Zuber, M. T. (2020). Moving academic research forward during COVID-19. Science, 368(6496), 1190–1192. https://doi.org/10.1126/science.abc5599