Home› Groups› Data Enumerators› Data Collection Experts› When you recognize a respondent

When you recognize a respondent

Started by Rhoda Nakhosi Sep 16, 2026 4 replies 👁 7 views
Rhoda Nakhosi Admin Community Champion (1,500+ points) Community Champion
Sep 16, 2026 at 7:51 pm

 In close-knit communities, enumerators sometimes realize mid-interview that the respondent is a neighbor, a relative, or someone they know from another part of their life. The protocol says proceed if consent is given, but the dynamic has already changed. Has anyone withdrawn from an interview for this reason, and how do you make that call in the moment without embarrassing the respondent or revealing why?

Fortunate Member Thought Leader (400+ points) Thought Leader
2 weeks ago

This is an important issue because consent alone may not fully address the power dynamics created by a pre-existing relationship. If an enumerator recognizes the respondent, I think they should have the option to pause and discreetly consult the supervisor rather than automatically proceeding.

Rhoda Nakhosi Admin Community Champion (1,500+ points) Community Champion
↩ replied to Fortunate 1 week ago
@Fortunate, your instinct to pause and consult the supervisor is right but I would push on what happens when the supervisor is not reachable. In a lot of field settings, enumerators work alone or the supervisor is hours away, so "consult first" becomes a dead letter in the moment that matters. The enumerator is left improvising, and improvisation under pressure is exactly when respondents get embarrassed or the relationship gets revealed. What has worked for me is giving enumerators a pre-agreed neutral line they can use without asking anyone. Something like "the study allows us to reschedule or assign another interviewer, no reason needed." That way the pause isn't an escalation, it's just a normal option being offered and the decision stays with the respondent. Do you think though that line should be scripted into the protocol, or does scripting it make it sound rehearsed in a community where everyone notices patterns?
Dr. Magoba Member Thought Leader (400+ points) Thought Leader
2 weeks ago

This is an important ethical and methodological issue because consent alone may not fully resolve the change in power dynamics when an enumerator discovers a pre-existing relationship with the respondent. In close-knit communities, continuing the interview may influence what the respondent is willing to disclose, even where consent is given, particularly for sensitive topics such as VAWG.

I would suggest that the protocol should allow the enumerator to pause discreetly and consult the supervisor, rather than automatically proceeding. The respondent should also be given a neutral opportunity to continue, pause, or request another interviewer without being required to explain why. For example, the enumerator can simply say that the study permits interviews to be rescheduled or conducted by another team member if the respondent prefers. This avoids embarrassing the respondent, revealing the reason for the change, or inadvertently disclosing a pre-existing relationship.

The decision should therefore be guided by confidentiality, voluntariness, privacy, and risk minimization, rather than consent alone. The research team and ethics committee should establish the protocol and escalation pathway beforehand, while the respondent retains the right to decide whether the interview continues. Where the relationship could reasonably affect disclosure, confidentiality, safety, or perceived freedom to decline, reassignment to another trained enumerator may be the most appropriate option. Importantly, the situation should be documented as a protocol event without recording unnecessary personal details that could identify the relationship.

Rhoda Nakhosi Admin Community Champion (1,500+ points) Community Champion
↩ replied to Dr. Magoba 1 week ago
@Henry Magoba, I agree with your framing. Consent alone does not resolve the power shift, and documenting it as a protocol event without unnecessary detail is exactly right. But I would add a question about who decides the relationship is significant enough to reassign. From the enumerator's seat, that judgment is being made in real time, often in front of the respondent, with no clear threshold. One enumerator withdraws at "neighbor," another proceeds at "cousin," and the data reflects the difference.
So the protocol may need to name what counts as a disqualifying relationship not just leave it to in-the-moment discretion. Have you seen a study that actually specified that threshold upfront, or does it usually stay a judgment call the enumerator has to own alone?