03 · What You Need to Know
How Workplace Hierarchy Changes the Evidence Generated by Focus Groups
Why Does Hierarchy Matter in a Focus Group?
Focus groups generate evidence through interaction. Participants listen to one another, respond to statements, express agreement or disagreement, and sometimes revise their interpretations during discussion.
These interactions occur within social relationships that may extend beyond the research session.
Managers can influence employees' workload, performance evaluations, promotion opportunities, employment conditions, and access to organizational resources. Even when managers do not intend to exercise authority during a discussion, employees may consider these relationships before speaking.
For example, an employee might privately believe that a new institutional technology creates unnecessary work but hesitate to express that view when the manager responsible for implementation is present.
The resulting silence should not automatically be interpreted as satisfaction with the technology.
Kitzinger (1995) emphasizes the importance of interaction in focus group research. Where authority relationships influence that interaction, they become relevant to both data generation and interpretation.
Direct Supervisory Relationships Create Particular Risks
A direct supervisor has formal responsibilities affecting an employee's professional circumstances.
These relationships create a different interactional environment from a discussion involving employees and managers from unrelated departments.
For example, a junior lecturer may be unwilling to discuss difficulties with workload allocation when the department head responsible for those allocations participates in the same focus group.
Even a supervisor who explicitly encourages criticism cannot guarantee that employees will feel free to disagree.
The perceived possibility of future consequences may be sufficient to influence participation.
Where the research concerns managerial decisions, workplace dissatisfaction, institutional problems, or sensitive professional experiences, separate groups are often the more defensible starting point.
Employees May Modify Their Accounts Without Being Directly Pressured
Power-related influence does not require a manager to interrupt, threaten, or instruct an employee.
Employees may anticipate what managers consider acceptable and adjust their contributions accordingly.
They might emphasize positive experiences, avoid naming problems, or use cautious language when discussing institutional decisions.
For example, an employee who experienced difficulties with an AI-based performance monitoring system might describe the implementation as "still developing" rather than openly criticizing its consequences.
Such wording may represent the employee's genuine interpretation, an effort to preserve professional relationships, or both.
Researchers should avoid assuming motives without evidence. Nevertheless, the possibility of self-censorship should inform the design and interpretation.
Managers May Also Change What They Say in Mixed Groups
Hierarchical influence does not operate only in one direction.
Managers may present organizational decisions more positively when employees are present or avoid discussing confidential considerations underlying policy implementation.
They may also become defensive when participants criticize procedures for which they are responsible.
Conversely, some managers may welcome discussion and reconsider their interpretations after hearing employee experiences.
The presence of employees can therefore influence managerial accounts as well as employee contributions.
Researchers should examine how the group setting shapes both perspectives rather than assuming that managers provide unrestricted institutional explanations.
Separate Groups and Mixed Groups Answer Different Questions
The most important distinction concerns what the researcher intends to understand.
If the research question asks how employees experience a policy, separate employee groups may reduce some immediate pressures associated with managerial authority.
If the research question asks how managers and employees negotiate interpretations of that policy together, mixed-role interaction may be part of the phenomenon.
These objectives require different data-generation conditions.
| Consideration |
Separate groups |
Mixed manager-employee groups |
| Primary analytical opportunity |
Role-specific experiences and interpretations |
Cross-role interaction and negotiation |
| Immediate supervisory pressure |
May be reduced |
May remain influential |
| Direct disagreement across roles |
Not observed within the same discussion |
Can potentially be examined |
| Sensitive workplace disclosure |
May offer more favorable conditions |
May be constrained by perceived consequences |
| Contextual comparison |
Requires analysis across separate discussions |
Can emerge through immediate exchange |
| Ethical concerns |
Confidentiality and identification risks remain |
Additional concerns about hierarchy and retaliation |
| Moderation |
Attention to within-role differences |
Greater attention to authority and unequal participation |
Separate groups do not guarantee candor, and mixed groups do not guarantee meaningful negotiation. Their relative suitability depends on the research context.
When Are Separate Focus Groups Preferable?
Separate groups are particularly appropriate when the study requires participants to discuss organizational experiences that may involve criticism of authority.
Examples include performance management, workload allocation, workplace discrimination, institutional misconduct, employee well-being, and dissatisfaction with administrative decisions.
They may also be useful when managers and employees have substantially different responsibilities and need opportunities to explain role-specific experiences in detail.
For example, teachers and academic administrators may interpret an AI policy differently because they encounter different implementation challenges.
Separate discussions can explore those perspectives without requiring participants to defend them immediately before people occupying other positions.
However, researchers should still consider informal status differences within each group.
When Might Mixed Groups Be Justified?
Mixed groups may be appropriate when cross-hierarchical interaction is central to the research question.
For example, a study might investigate how managers and employees collectively negotiate responsibility for using generative AI in organizational decision-making.
The researcher may need to observe how participants explain competing priorities, respond to authority claims, and interpret organizational responsibilities together.
Mixed discussions can also be useful in some participatory research or formative policy-development settings.
However, a participatory workshop and a research focus group are not automatically equivalent. Their objectives, consent arrangements, and intended use of the resulting material should be distinguished.
The potential analytical benefit of mixed participation must be weighed against foreseeable risks to participants.
Are Managers from Different Departments Less Problematic?
Sometimes. A manager who has no direct authority over another participant may create fewer immediate concerns than that participant's supervisor.
However, formal reporting relationships are not the only relevant source of influence.
Managers from other departments may still participate in institutional decisions, influence reputations, or have access to senior leadership.
Employees may perceive them as representatives of management generally.
Researchers should therefore assess actual and perceived authority rather than relying solely on whether participants belong to the same department.
What About Teachers and Academic Administrators?
Educational institutions contain authority relationships that are sometimes less visible than those in conventional workplace hierarchies.
Department heads may influence teaching assignments, promotion recommendations, performance reviews, and access to professional-development opportunities.
Program coordinators may have limited formal authority but substantial informal influence over instructional practices.
A researcher investigating institutional AI governance should therefore distinguish teachers' experiences from administrators' policy responsibilities.
Where the study concerns how teachers experience administrative requirements, separate discussions may be preferable.
Where it concerns negotiation between organizational roles, mixed discussion may be analytically relevant if the risks are acceptable.
Why Confidentiality Is Especially Difficult in Workplace Focus Groups
Researchers can protect recordings, transcripts, and identifying information under their control. They cannot guarantee that other participants will keep the discussion confidential.
Employees may be identifiable through descriptions of their responsibilities, incidents, or departmental experiences even when names are removed.
For example, a participant describing a particular dispute over AI assessment policy may be recognizable to colleagues familiar with the event.
Mixed groups may increase the consequences of identification when managers have authority over participants.
Researchers should explain confidentiality limitations before consent and avoid promising that statements cannot affect workplace relationships.
The decision about including participants who already know one another should account for these risks.
Can Ground Rules Make Mixed Groups Safe?
Ground rules may establish expectations for respectful discussion, noninterruption, and confidentiality.
They can help clarify how participants are expected to interact.
However, ground rules do not eliminate organizational authority or guarantee that information will not be repeated outside the session.
Nor can a moderator ensure that managers will never interpret employee statements through their existing professional relationships.
Researchers should not use ground rules as a substitute for evaluating whether mixed participation is appropriate in the first place.
Can a Neutral Moderator Remove Hierarchical Pressure?
An independent moderator may reduce some concerns about institutional influence, particularly when participants believe the research team is connected to management.
Neutral prompts and balanced facilitation can also create opportunities for different perspectives.
However, moderator independence cannot remove the authority participants hold outside the session.
Even when managers remain silent, employees may anticipate how their contributions could be interpreted.
Researchers should therefore treat moderation as one safeguard within a broader design rather than a complete solution to hierarchical influence.
Could Anonymous Written Responses Help?
Researchers sometimes use private written reflections before group discussion to allow participants to formulate views without immediate social influence.
This may help identify issues that participants would otherwise hesitate to introduce.
However, written responses collected within a mixed group are not necessarily anonymous, particularly when handwriting, examples, or contextual details are recognizable.
Moreover, the subsequent group discussion may still expose participants to pressure.
Private written contributions can supplement a mixed discussion, but they should not be presented as eliminating the risks associated with hierarchical participation.
Should Separate Groups Be Conducted Before Mixed Groups?
A staged design may be useful when the study requires both role-specific accounts and cross-role interaction.
For example, researchers might first conduct separate discussions with employees and managers to understand their perspectives.
A later mixed session could examine how participants negotiate selected issues.
However, researchers should not automatically bring the same participants together after they have disclosed sensitive experiences privately.
Participation in a later mixed session requires appropriate consent, risk assessment, and careful decisions about which information may be introduced.
The staged arrangement should be justified by the research question rather than adopted as a universal procedure.
How Can Mixed Groups Affect the Interpretation of Consensus?
Suppose a manager describes an organizational policy as effective, and several employees agree.
The researcher should not automatically conclude that the employees independently endorse the policy.
Agreement may reflect shared experience, politeness, uncertainty, perceived authority, or other interactional circumstances.
Likewise, disagreement does not necessarily indicate unrestricted expression of opposing views.
Researchers should examine how power differences shape focus group data and interpret the resulting interaction in context.
The analysis should distinguish what participants explicitly expressed from interpretations about why they expressed it.
What If Managers Dominate the Conversation?
Managers may contribute extensively because they possess institutional knowledge or are accustomed to leading meetings.
The moderator can acknowledge their contributions while creating opportunities for other participants to speak.
However, simply asking employees to disagree with managers may expose them to additional pressure.
Moderators should avoid turning quiet employees into publicly identified dissenters.
Approaches for managing dominant and quiet participants can support balanced participation, but they cannot resolve every structural concern.
If hierarchy prevents the study from obtaining relevant evidence, a change in group composition may be more appropriate than continued moderation.
How Should Managers and Employees Be Recruited?
Recruitment procedures should minimize inappropriate pressure.
Employees may feel obligated to participate when invitations come directly from their supervisors, particularly if the study is described as institutionally important.
Where feasible, independent recruitment procedures may help distinguish voluntary research participation from workplace expectations.
Researchers should explain that participation is voluntary and clarify what information, if any, managers will receive about participation.
However, anonymity of participation cannot always be promised when employees attend visible group sessions.
Recruitment and consent arrangements should reflect the actual organizational context and applicable ethics requirements.
How Should the Researcher Decide Between Mixed Groups and Individual Interviews?
Some questions may be poorly suited to group discussion regardless of whether managers and employees are separated.
For example, research involving allegations of workplace harassment, retaliation, or identifiable misconduct may require greater privacy than a focus group can provide.
Individual interviews may allow participants more control over disclosure and reduce immediate peer influence.
However, interviewer authority, confidentiality limitations, and institutional pressures can still affect individual interviews.
The choice between individual interviews and focus groups should follow the evidence required and the conditions necessary for ethical participation.
How Should Group Composition Be Reported?
The methods section should explain whether managers and employees participated together or separately and why the arrangement was chosen.
Researchers should describe relevant authority relationships, recruitment procedures, moderation safeguards, and confidentiality limitations.
If mixed groups were used to study organizational interaction, explain how that interaction informed the analysis.
If separate groups were used, clarify how perspectives were compared without assuming that each group represented an entire organizational category.
Where hierarchy may have constrained disclosure, acknowledge the limitation rather than claiming that the procedure eliminated all participation bias.
