Dependent Groups and Unequal Power
The definition has two limbs and the second is the wide one — an unequal power relationship with you, or with an organisation cooperating in the research. What the application then demands is not a promise about obligation. It is a mechanism that removes you from the approach.
The definition, and how far the second limb reaches
Most readers check themselves against the first limb, find they are not anyone's manager, and move on. The second limb is where the majority of project management studies actually land.
An unequal power relationship with an organisation which is cooperating in the research does not require you to hold any power at all. It only requires that the organisation letting you in does, over the people you want to talk to.
WHO THE TWO LIMBS CATCH IN A TYPICAL PROJECT STUDY
| Participant | Limb engaged | Where the inequality sits |
|---|---|---|
| Someone who reports to you directly or through one line | First — with you as researcher | You write or influence their review. Obvious, and the one everybody checks |
| A peer on the same delivery team | First, usually | No reporting line, but shared accountability and daily proximity. The source's own worked case is exactly this |
| Any employee of the organisation hosting your study | Second — with the cooperating organisation | Their employer approved the research. Declining can look like declining something the employer wants |
| A contractor or supplier to the host organisation | Second | Their commercial relationship is renewable and the host decides. Often the most dependent group in the whole study |
| A client or service recipient of the host organisation | Second | They receive something from the organisation, which is the classic shape of a right-to-service problem |
| A junior colleague in another part of the organisation | Second | You have no authority over them, but the organisation does and you arrived carrying its endorsement |
| A member of a partner or funding organisation | Second, if that organisation is cooperating | Cooperation is the trigger word, not employment. Anyone the cooperating organisation has leverage over is inside the definition |
The two-limb definition is the source's. The allocation of typical project participants against it is analysis added here.
What "do not feel under any obligation" actually demands
Read the object of the sentence carefully. The application does not ask you to ensure participants are not obliged. It asks you to ensure they do not feel under any obligation. Three consequences follow, and each one closes off a common answer.
Three things the wording rules out
A statement that participation is voluntary does not discharge it
Every information sheet says participation is voluntary. If a sentence were sufficient, the requirement would not exist as a separate question in the consent part of the form. The wording targets the state of mind that survives the sentence.
Signed consent is evidence of agreement, not of freedom
A dependent participant signs exactly the same form as a free one. Consent mechanics answer a different question on the same page — see Voluntary Participation and Informed Consent.
The verb is detail, which is the form's strongest demand
Where other parts of the application say state, outline or describe, this one says detail. That asks for specific arrangements with named actors and a sequence — the level at which an application either satisfies a panel or comes back.
What satisfies the requirement is therefore structural. You have to change who does what, in what order, so that a person who wants to decline can do so without the fact of declining reaching you. The full context of the form and its verbs is in Writing an Ethics Application.
The worked case: an evaluative study inside the researcher's own team
A student was undertaking an evaluative study within her own workplace, on the same team as the participants. Both limbs of the definition were engaged at once: she had a working relationship with every participant, and the organisation cooperating in the research employed all of them.
The arrangement that eventually satisfied the requirement was imposed by the workplace director, and it has four parts.
- The requirement setThe workplace director required that colleagues be approached individually by the university supervisor — not by the researcher.
- Before any approachA background information sheet, prepared by the researcher, had to be sent first by the supervisor, detailing what participation involved.
- The approach itselfEach colleague was approached individually by the supervisor. The researcher made no approach to anyone.
- If a colleague agreedThe director had to be notified before the researcher was informed.
- Through the committeeThe ethics application went back and forth to the committee five times for amendments before approval. The process took more than five months.
The other two cases the source supplies are treated in Ethics Application Failures: What Goes Wrong, which also draws out what all three have in common.
The mechanism underneath the arrangement
The specific arrangement in that case belongs to that workplace. The mechanism generalises, and it is worth stating plainly because it is the answer to almost every dependent-group question.
Insert a third party between researcher and participant, so that the researcher never learns who declined.
Everything else follows from that one property. If declining is invisible to you, there is nothing for a participant to fear from declining, and the feeling of obligation loses its object. If declining is visible to you, no amount of reassurance removes it.
How the mechanism is built
Someone other than you makes the approach
In the source's case that was the university supervisor. The essential property is that the person approaching has no stake in the answer and no ongoing relationship in which a refusal could matter.
Written information travels ahead of the ask
A background information sheet detailing what participation involves goes out before anyone is asked to decide. A person reading a document alone is in a very different position from a person being asked a question in a corridor.
The decision goes back to the intermediary, not to you
Agreement and refusal both travel to the third party. This is the step that makes refusal invisible, and it is the step most often dropped because it feels like an unnecessary layer.
Only agreement reaches you
You learn who is participating. You never learn who was asked and said no, and you should not be able to work it out from the size of the group you were told was approached.
The arrangement is written into the application
Each of these steps is named in the consent part, with the actor for each. This is what the verb detail is asking for, and it is why a general assurance of voluntariness comes back.
Designing the buffer for your own study
Not every study can use a university supervisor as the intermediary, and the choice of buffer changes what you can promise. The options below are synthesis rather than the source's prescription; the source describes one case and states one requirement.
Choosing a mechanism for your situation
What a panel is looking for, and what it costs
The requirement is answered with specifics or it is not answered. The checklist below is what the verb detail is asking you to be able to state.
Evidence that obligation has been removed
- You have stated which limb of the definition applies — with you, with a cooperating organisation, or both
- The person who makes the approach is named by role, and it is not you
- That person's relationship to the participant does not depend on the participant's answer
- Written information detailing what participation involves goes out before anyone is asked to decide
- Refusals travel to the intermediary and never reach you, and you cannot infer them from what you are told
- Anyone excluded on ethical grounds appears in your exclusion criteria with the reason
- Nothing in the recruitment material implies the organisation expects or values participation
- Confidentiality is described separately for collection, analysis and reporting, including who inside the organisation learns anything
- Withdrawal after agreement is described, and it also runs through the intermediary
Build the time in as well. Approval sequencing, and why nothing collected beforehand can be used, are covered in Ethics Review and Approval Processes — and where your employer is also the sponsor, the independence questions in Sponsors and Researcher Independence apply to the same arrangement.
What to carry forward
- The second limb of the definition catches far more people than the first. If a cooperating organisation has power over your participants, they are a dependent group even if you have none.
- The requirement targets a feeling, not a fact. A voluntariness statement, a signed form and a good intention all leave the feeling intact.
- The verb is detail — the form's strongest demand. Named actors and a sequence, not an assurance.
- The one mechanism that works is a third party between you and the participant, built so that you never learn who declined.
- Written information travels ahead of the ask, and refusals travel back to the intermediary. Dropping that second step removes the whole benefit.
- Five rounds and more than five months describes one application, of the hardest kind. It is a reason to start early, not a number to plan against.
- The buffer often runs through someone with authority over the participant. Notice what that does to confidentiality and say what you will do about it.
Frequently asked questions
Are my own colleagues automatically a dependent group?
If you work for the organisation cooperating in the research, the second limb of the definition is engaged regardless of whether you manage anyone. The source's own worked case is a researcher studying her own team as a peer, and it was treated as a dependent-group application. The safe assumption for any workplace study is that the requirement applies and has to be answered with a mechanism.
Is a statement that participation is voluntary enough?
No. The application asks you to detail how you will ensure participants do not feel under any obligation, which is a separate question from whether participation is formally voluntary — and every information sheet already says it is. What satisfies the requirement is a described arrangement in which a person can decline without you learning that they did.
How long does a dependent-group application take?
The supplied material gives exactly one figure: an application that went back and forth five times and took more than five months. That is one case, from the hardest category of application, and the source offers no average, range or typical duration anywhere. Treat it as a reason to begin the application early rather than as a duration to plan against.
Who can act as the intermediary?
The source does not say. In its worked case the workplace director required that the university supervisor make the approaches, with a background information sheet sent first. The property that matters is that the intermediary has no stake in the answer and no relationship with the participant that a refusal could damage. Where no such person exists, raise it with your supervisor before designing around it.
What if my group is too small for refusals to be invisible?
Then say so in the application rather than claiming a protection you cannot deliver. Reduce what turns on it instead: report in aggregate, avoid attribution by role or project, and do not state how many people were approached. A panel can work with an honestly stated limit; it cannot work with an assurance that the arithmetic contradicts.
Should I exclude my direct reports from the study?
Often, yes, and the application gives you a natural place to say it. The participants part asks for selection and, where appropriate, exclusion criteria — an exclusion made on ethical rather than methodological grounds shows the panel you identified the problem and dealt with it. Losing a few participants is cheaper than an arrangement no buffer can rescue.
References and source attribution
- Cooper, D. & Schindler, P. 2008, Business Research Methods, 10th ed., McGraw-Hill — the text the supplied source draws on for its treatment of coercion and participant rights.
- Quinlan, C. 2011, Business Research Methods, 1st ed., Cengage Publishing, Chapter 3 — the prescribed reading accompanying the ethics material.
- Trochim, W. M. K. 2006, Research Methods Knowledge Base — the source of the excerpt on voluntary participation and the system of ethical protections that this requirement operationalises.
- Swinton, J. 2009, 'Why bother with research ethics?', Journal of Community Nursing, vol. 23, no. 9, pp. 25–27.
- The supplied teaching source: consolidated weekly teaching notes and slide material on research ethics, which supplies the dependent-group definition quoted here, the eight-part application structure it sits in, and the three case studies stated to be based on fact with fictitious names.
Suggested questions for Ask KEVOS
- My participants are peers on my own delivery team. Draft the arrangement I should put in the consent part of my application.
- Work out which limb of the dependent-group definition applies to each participant group in my study.
- Design a recruitment path for contractors of my host organisation where refusal never reaches me.
- My team is only six people. How do I answer the obligation question honestly when refusals cannot be invisible?
- Write the exclusion criteria that remove my direct reports from the study, with the ethical reason stated.