Research on Subordinate Populations

Because of their inherent vulnerability, subordinates should be included in research only when scientifically necessary, and researchers must adopt practices that minimize real or perceived undue influence or coercion and maximize privacy and confidentiality. It is important to understand that research on subordinates requires a full board review because the IRB is tasked with being "particularly cognizant" of participants who may be vulnerable to coercion or undue influence, particularly during the informed consent process, which must minimize the possibility of coercion or undue influence [45 CFR 46.116(a)(2)].

For example, subordinates (employees, students, etc.) may experience a perceived or real sense of coercion to enroll in a study. They may anticipate negative effects based on a choice to not participate, and they may feel pressured to participate in the study, particularly if data collection occurs during a work shift, course, or program. Foreseeable negative impacts could include fears about how managers, faculty, or peers will treat them. For example, declining to enroll in a study conducted by management could lead to a full spectrum of issues such as tense relationships, lower performance evaluation, job loss, etc., or the fear of such consequences. Students may be concerned about negative repercussions related to their position as a student (i.e., grades, future letters of recommendation, assistantships, etc.) if the investigator is an instructor, advisor, or active in their field of study.

Pressure is heightened if the researcher personally invites subordinates to participate or is directly involved with data collection—it is very difficult to say "no" to one's boss or instructor. Alternatively, a person may believe that participating will garner favor or benefit (e.g., better grades, favor with a supervisor, etc.). Finally, some circumstances beyond the student and employment context could also expose people to actual or perceived coercion or undue influence to enroll in a study. In short, researchers must pay attention to any power imbalance.

To sum up, because of their inherent vulnerability, subordinates in research can be complicated. Thus, if a subordinate population is of interest to you, you are encouraged to find a similar population outside of your organization. However, if your research must include people in a subordinate position, use the information that follows to design your study in a way that will reduce IRB concerns.

Confidentiality and Privacy Concerns

Implementing confidentiality and privacy protections can be complicated when protecting subordinate participants. Research often involves gathering information that participants would not normally share with their employers, supervisors, instructors, etc. For example, research participation may prompt the disclosure of negative statements related to their vulnerable context (e.g., assessments of their workplace culture or supervisor practices, opinions about instructor-effectiveness or teaching practices). Some studies request personal information and even seemingly benign information can feel risky to share with superiors (e.g., political opinions, religious practices, family background, purchasing habits, etc.). Personal discomfort or professional or legal risk is enhanced when participants can be linked to the information they provide. Anonymity is difficult to ensure when collecting information from persons known to the researcher. Basic demographic or background information is often enough for those familiar with the research setting to deduce identities. Moreover, some protocols prompt participants to provide examples or describe specific incidents. Even if participants don't specify names of individuals, organizations, or locations, people (participants or third parties) might be recognizable given enough related information.

Strategies to Mitigate Undue Influence or Coercion

The best mitigation strategies require that someone other than the researcher (and not in a supervisory role), recruit, gain consent, and collect data from the participants.

If you have concluded that it is scientifically necessary to invite participants who are vulnerable to coercion or undue influence due to their subordinate role, consider the following mitigation strategies:

Recruit participants broadly instead of relying only (or primarily) on subordinates (for example, send recruitment messages to students or employees of a whole department/college/organization instead of focusing only on those in the your classes or departments)
Use indirect recruitment methods that allow potential participants to initiate contact if they are interested (for example, post flyers or use mass emails instead of directly calling or emailing potential participants)
Design the study so that you don't know who has elected to take part in the research
Explicitly state that participation is voluntary in recruitment messages, consent processes, and throughout the study
- Make sure people clearly understand that their decision of whether to take part in the research, or to withdraw at any time, will have no impact (positive or negative)
- Include related language in recruitment and consent materials; for example, include language that conveys that their choice of whether or not to participate, to abstain from answering questions, or to entirely leave the study, will not have any effect on them in their role (as an employee, student, etc.), or their relationship with the organization
Allow people as much privacy as possible related to their decision to enroll in the study and provide participants plenty of time to consider and a private method to consent or decline
Minimize the information gathered to that which is scientifically necessarily; avoid detailed or specific background or demographic information when possible
Ensure privacy during data collection:
- Conduct the research out of sight by other employees/students whenever possible
- Avoid group settings (e.g., focus groups) for research that gathers information persons may not want their colleagues or classmates to know. If focus groups are conducted, include on the consent form, and reiterate prior to starting the group, that participants will protect group and individual privacy and confidentiality by not discussing who was present or what was shared (although note on your consent form that there is no guarantee confidentiality will be maintained)
- Work with fully de-identified responses (this may not be feasible in many cases)
- Advise participants to use personal (not work-issued) contact methods for study communication, such as personal devices and accounts
- Carefully assess information included in presentations and publications related to the study to ensure participants’ identities cannot be deduced by those familiar with the research setting; redact or recode information that directly or indirectly identifies people, locations, or organizations
- Consider whether it is necessary to share information that may disrupt workplace relationships (if a colleague or superior were to read it)
Employ a “member checking” process and allow participants the opportunity to review data collected from/about them and/or descriptions of them in publications/presentations, so they can point out potentially identifying information and/or information that may present risk
Inform participants -- acknowledge the unique privacy/confidentiality concerns and explain how you are addressing them:
- Explicitly inform participants that they can decline to provide information or complete procedures that make them uncomfortable
- Clearly inform participants of limits to privacy and confidentiality, along with any associated risks or discomforts
- Inform participants about any plans to present findings to or share reports with their workplaces, schools, etc., and what that might mean for them (e.g., a colleague or peer might figure out that they were involved in the study or the information they shared)