Like Superman,Institutional Review Boards Play an Important Role,But When Unchecked Can Cause Significant Collateral Damage: Five Reasons That it is Time to Reset Institutional Review Boards

#5 IRB Review Processes are Prone to Inconsistency That Can Result in Biased Decision-Making and Other Problematic Outcomes

The GOMusings guidelines say that I can’t gripe about things like emails telling me what words my IRB chair “doesn’t like,” so I’m not going to do that (did I just find a loophole here?). I will, however, discuss the overarching problem of IRBs leaving researchers to guess at their boards’ daily whims. While some argue that perceptions of IRB inconsistency are based on local idiosyncrasies (Stark, 2007), many IRBs fail to demonstrate internal consistency in their decision-making. At multiple schools, I’ve submitted protocols with previously approved language only to have the IRB ask me to modify the language IT PREVIOUSLY APPROVED.

In addition to researchers complaining about IRB consistency, they also raise serious concerns about IRBs showing favoritism toward some colleagues and allowing biases to influence their decisions (Keith-Spiegel & Koocher, 2010). This is likely because many IRBs don’t consistently evaluate protocols using objective standards, and when we don’t use objective standards, we increase the potential role of bias in evaluative processes (Obenauer, 2023). Such biases are often conceptualized as being identity-based (e.g., gender, racial), but within the context of IRB review, they may also be based on status (e.g., tenured vs. untenured) or work groups (e.g., psychology department vs. management department). For example, a lack of consistent review criteria may result in an IRB chair applying a less stringent review process to members of their own department than members of other departments. Inconsistent processes can also contribute to inefficiencies that cause researchers to waste time or miss time-sensitive data collection opportunities. Additionally, to the extent that inconsistent processes lead to decreased perceptions of procedural justice, they may negatively influence workplace outcomes such as organizational citizenship behavior (Mohammad et al., 2019). Within the context of academia, this may cause faculty who feel that they have been treated with procedural injustice by their IRB to be less likely to voluntarily contribute to the campus environment. So yeah, IRBs are protecting participants, but their lack of consistent guidelines opens the door to procedural inconsistencies, inequitable treatment of researchers, and other negative workplace outcomes.

Recommendation 1: IRBs need to set clear, objective standards and provide researchers with template protocol language.

#4 IRBs are Rewriting the Rules

Imagine a situation where a first-year journalism student and a faculty member with 20 years of experience working with human subjects want to distribute identical opinion surveys about a timely current event. One of these people can start data collection immediately with no oversight, but the other will spend a month dealing with their IRB’s bureaucratic processes and launch their survey just in time for people to have stopped caring about the event. If you guessed that the student journalist is the one who won’t be required to have their survey approved (HHS, 2018), you are correct!

In all fairness, the professor’s research should be “exempt from the requirements of [45 CFR 46]” (National Archives, 2023), and some have argued that the law allows researchers to make this determination on their own (Shweder & Nisbett, 2017), but many universities add restrictions beyond what the law requires that serve to delay and impede the research process. For example, after determining that research is legally exempt from IRB oversight because it is an anonymous survey, some schools require researchers to provide regular updates on their research (UIC OPRS, 2019), resubmit the research to the IRB if survey questions change (University of Washington, 2023), or request IRB permission to increase their study’s sample size (OVPRI, 2023). It is as if these IRBs somehow think that changing a survey question or increasing sample size impacts whether survey research still meets the exemption criterion of being a survey. Additionally, each resubmission slows the research process as reviewing research for exempt status can take five to ten business days at some schools (e.g., Cornell University ORIA, 2019), but others allow themselves two to four WEEKS (e.g., IRB-SBS, 2023).

These examples demonstrate how some IRBs are using their unchecked power to rewrite rules, which results in the impediment of research. I would call it a Lex Luthorish scheme, but his schemes have a point! Nevertheless, IRBs are restricting academic freedom by operating beyond the scope of their legal responsibilities. In some cases, IRB overreach can even contradict regulations (Klitzman, 2013). At unionized schools, a university’s failure to eliminate such overreach may violate the academic freedom clause in collective bargaining agreements.

Recommendation 2: When universities permit IRBs to impose restrictions beyond what the law requires and such restrictions violate academic freedom clauses in collective bargaining agreements, faculty unions need to take appropriate action.

#3 IRBs Prioritize Regurgitating Information That They Don’t Understand Over Assessing Risk

Unlike Lex Luthor, IRBs don’t even fully understand the game they are playing. What I mean by this statement is that IRBs are supposed to assess risk (that’s “the game”), but rather than doing this, they often universally apply broad guidelines with little regard for how they actually relate to risk. For example, when considering discomfort in terms of risk assessment in social science research, IRBs frequently cite subjects such as sexual behavior, substance use, and mental health as introducing risk due to their potential to create discomfort. Research findings, however, question whether surveys related to these topics universally introduce more discomfort than what is typically experienced in daily life activities (Mustanski, 2011). While it is reasonable to expect that some survey questions may make participants uncomfortable and there could be situations in which particular questions could even serve as a trigger for participants with post-traumatic stress disorder (PTSD), these factors should be considered on an individual basis, and IRBs should (1) clarify their risk concerns, and (2) allow subject matter experts (i.e., the researcher) to clarify how they have addressed these concerns before requiring a generic solution such as language in an informed consent document. If an IRB does desire a generic solution, it should be able to clearly explain to researchers how such a solution actually protects participants (Fitch, 2005).

The most obvious example of IRBs just repeating what they don’t understand is when they say that IP addresses compromise anonymity. IP addresses are dynamic such that one IP address can be assigned to a variety of different electronic devices depending on when they are connected to a network. IP addresses can be used to narrow down the location of where an internet connection took place, but they cannot directly identify a device user. Many IRBs refer to IP addresses as identifying information because HIPAA’s Safe Harbor Rule requires IP addresses to be removed from health data for the data to be considered de-identified (HCS, 2022). Safe Harbor identifiers do not need to identify an individual on their own but could be used to identify an individual by combining them with other information (The Office for Civil Rights (OCR), 2012). In other words, to de-identify health data, one must remove IP addresses because when combined with health information, it is possible that an IP address can contribute to compromising anonymity.

So how does this translate to IP addresses compromising the identity of a person who completes an online survey about job attitudes? Simple…IT DOESN’T!!! In fact, some IRB members acknowledge that they place restrictions on digital data collection despite not understanding the related risks or potential harms (Huh-Yoo & Rader, 2020). Claiming that Safe Harbor identifiers compromise anonymity under all circumstances is lazy, but it has become accepted practice.

Recommendation 3: Rather than using universal logic such as, “X always compromises anonymity” or “Identity disclosure creates risk in all conditions,” faculty members on IRBs should strive to learn how different factors contribute to risk and apply risk concerns appropriately and systematically.

#2 IRBs are Creating Unnecessary Risk

Furthermore, IRBs often force researchers to warn participants of the risk of identity exposure, even when having their identity exposed would not result in a negative consequence. While there are certainly situations in which identity exposure can introduce risk (e.g., a study asking how often people steal office supplies from work), this does not apply to every study. For example, Briggs (2022) cited a benign protocol where the IRB required them to “tell people with PhDs that using their laptops in public might allow someone to see their screen.” If we consider that the two components of risk are “probability of occurrence and magnitude of those harms” (Cooper & McNair, 2015, p. 101), we can’t even begin to assess the risk associated with participating in Briggs’ study until we identify the specific harms that could occur if someone witnessed a professor filling out a benign survey. IRBs, however, are not required to articulate those harms, their magnitude, or their likelihood of occurrence.

I’ve reviewed publicly available IRB protocol and consent templates from schools around the country, and I am shocked by how many require researchers to warn participants that their identity could be exposed, no matter what the research is. A survey could be asking people’s favorite superhero, and IRBs would still require researchers to warn participants about the potential for identity exposure as if the result could be a heat vision blast from Superman (for those who don’t know, Superman is a fictional character…and he’s nice). When we tell people about a false risk, they start wondering things like, “Why should I care? What am I missing?” Such statements of false risk could make participants unnecessarily uncomfortable. Therefore, when IRBs force us to warn people of non-existent risks, IRBs are unnecessarily increasing the potential for participant discomfort and introducing risk to the research process.

Recommendation 4: Faculty need to push back on IRB requests to list false risks.

#1 IRBs Perpetuate Discrimination by Impeding Researchers’ Efforts to Learn More About Diversity, Equity, and Inclusion (DEI) Issues in the Workplace

One purpose of IRBs is to prevent the exploitation of racial minorities, such as what happened in the Tuskegee study (Corbie-Smith, 1999). This doesn’t, however, magically immunize IRBs from having discriminatory effects or impeding efforts to achieve equity. While much of the literature on IRBs impeding equity has focused on healthcare outcomes (e.g., Friesen et al., 2022), in this section, I will illustrate how by forcing researchers to overstate risks, IRBs may be altering participant behavior in such a way that they limit management scholars’ ability to develop knowledge related to DEI in the workplace (see Figure 1).OPEN IN VIEWER

Following a series of recent failed replications (e.g., Obenauer & Kalsher, 2022; Ubaka et al., 2022) in the DEI space, Obenauer (2023) reasoned that social desirability effects might be masking discrimination in the lab, but does that really make sense? I mean, who cares about behaving in a socially desirable way in an anonymous survey? I don’t know…maybe people who were intimidated by a long, scary informed consent document. Doesn’t it seem possible that research participants are behaving in a socially desirable way because we tell them that they shouldn’t trust us to protect their identity? Beyond that, some schools require researchers to disclose the use of deception by omission before a study takes place (e.g., UPenn, 2023). So basically, we tell participants, “Hey, the research question here is so sensitive and damning that we can’t actually tell you what it is…oh, and by the way, we might not keep your identity secret,” and then we expect people to respond with honest answers. How can a field full of PhDs studying human behavior expect that this implied message won’t influence participant responses?

I appreciate the need to protect participants, and I can see how, perhaps, one could argue that leaking certain attitude scale scores could pose a risk to participants, but here is my issue: IRBs are forcing researchers to prioritize people’s rights to mask their biases over society’s need to address workplace discrimination and, in many cases, they are doing so when identity disclosure is unlikely to cause material harm to the participant (e.g., a between-subjects vignette study where participants rate employee competence). Beyond impeding efforts to make meaningful change, the impact of this overstating of risk is that every experimental study that fails to capture workplace discrimination in a way that accurately reflects society has the potential to feed the false narrative that workplace discrimination is no longer a problem. IRBs are supposed to protect participants, but does using consent forms to scare people into hiding their discriminatory tendencies protect participants, or does it protect discrimination?

Final Recommendation: Management faculty need to conduct and publish research on the unintended consequences of IRB overreach in our field.

Closing Tirade

The problem is that 20 thousand people can be discriminated against in the workplace, and no one will ever look at an IRB and say, “If you hadn’t impeded important research, this could have been avoided.” An IRB won’t be held accountable for slowing down the research of a person who didn’t get tenure. If a student has a stress-related event because they’re overwhelmed by an unnecessarily scary and complex consent form, there will be no blowback on the IRB. Just like Superman after he places a villain’s yacht in the middle of a city street in an unchecked attempt at humor, IRBs are not accountable for the collateral damage they cause as a result of overreach.

As societal leaders, however, we must recognize that our responsibilities go beyond identifying problems—instead, we should be solving them. More than 10,000 scholars typically attend the Academy of Management’s (AoM) annual meeting. Perhaps it is time to introduce professional development workshops focused on identifying solutions for these concerns. Imagine a round table session where elected legislators engage in dialogues with researchers about IRB overreach. Faculty could also use large conferences such as AoM and SIOP to organize letter-writing campaigns where attendees can simply sign prepared letters addressed to their congressional representatives regarding these issues. Unionized campuses are often affiliated with the National Education Association (NEA), which is the largest labor union in the United States and has considerable lobbying power (Noe et al., 2019). Faculty should ask the NEA to advocate for legislation limiting the power of IRBs. Furthermore, we could take a page out of the marijuana legalization playbook and leverage local relationships in an attempt to limit the boundaries of IRB influence through state-level legislation.

Real Final Recommendation: Faculty need to lobby to have legislation passed that limits IRBs’ ability to overreach.