Interviewer Leading

Treat every recorded answer as the sum of the respondent's underlying state and the elicitation channel's own response function, since the act of asking — wording, presupposition, anchoring, order, interviewer affect — contributes content to the answer rather than transparently retrieving it.

Version

v3 · 2026-09-28 · History · 1 correction

Domain-specific #

565

Domain group

Social Sciences

Origin domain

Sociology & Anthropology

Subdomain

Research Methodology and Qualitative Inquiry → Sociology & Anthropology

Core Idea

Interviewer leading is the family of elicitation failures in which the act of asking — question wording, embedded assumptions, preceding-question frame, interviewer affect, response-option structure — systematically biases the answer in a direction the procedure introduces rather than the respondent's state supports. The structural commitment is the non-neutrality of the elicitation channel: a question is not a transparent pipe but a stimulus that participates in constructing the answer it records, so the answer is a joint product of state and channel.

How would you explain it like I'm…

Questions That Push Answers

If someone asks you, 'Wasn't that cake super yummy?', you might say yes even if you didn't love it. The way the question is asked pushes your answer one way. That push is called interviewer leading.

Questions That Steer Answers

When researchers ask people questions, they want to learn what those people really think or remember. But the question itself can change the answer. Choosing a word like 'smashed' instead of 'hit,' mentioning a number, or smiling at certain answers can all nudge people in one direction. So the answer is partly the person and partly the question. Good interviewers use careful, open questions to make that push as small as possible.

Question-Induced Response Bias

Interviewer leading is a family of errors in which the act of asking biases the answer in a direction that comes from the question procedure, not from what the respondent actually believes or remembers. A recorded answer is a joint product of the person's real state and the way the question was asked. The pathways include wording (Loftus found 'smashed' versus 'hit' changed speed estimates and false memories of broken glass), anchoring on any number mentioned, loaded presuppositions ('Have you stopped ignoring feedback?'), question-order effects, and demand characteristics, where people sense what answer is wanted. The fixes are shared: pre-planned wording, open-ended questions before closed options, randomized answer order, trained interviewer reactions, and structured protocols like the Cognitive Interview or the NICHD protocol for children.

 

Interviewer leading names the class of elicitation failures in which the question procedure systematically biases the response. The key structural claim is that the elicitation channel is not neutral: a question is a stimulus that helps construct the answer it records, so observed responses combine the respondent's underlying state with the channel's own response function, and only the former is of interest. Mechanisms include wording effects (Loftus's 'smashed' versus 'hit' changed speed estimates and false recall of broken glass), anchoring on any mentioned figure even when explicitly arbitrary, loaded presuppositions that force the respondent to accept or contest a framing, order effects from preceding questions, the structure of response options, the interviewer's vocal and postural reactions to partial answers, and demand characteristics of the whole encounter. Because all are manifestations of the question contributing content, they share a remedy family: pre-registered protocols and wording, open-ended prompts before closed alternatives, randomized option order, affect training, and structured protocols in high-stakes settings such as the Cognitive Interview, the NICHD child forensic protocol, and the UK PEACE model.

Scope of Application

Interviewer leading lives across the elicitation-methodology substrate — wherever a human respondent's reportable internal state is converted to a recorded answer through an asking-channel.

• Survey methodology — question-wording experiments, context effects, and AAPOR design standards.
• Eyewitness interviewing — Loftus's misinformation paradigm and the corrective Cognitive Interview and NICHD protocols.
• Clinical interviewing — structured-versus-unstructured trade-offs and avoidance of leading questions at intake.
• Market and user research — the "ask what they did, not what they would do" discipline.
• Intelligence interrogation — the PEACE and HIG models' non-leading discipline.
• Police interrogation — false confessions from leading questioning, the highest-stakes manifestation.

Clarity

Naming interviewer leading dissolves the illusion that a question is a transparent pipe and the answer a clean readout of belief or memory. Once the channel is non-neutral, a divergence between two instruments asking "the same" question stops being noise and becomes a measurable quantity — the channel's contribution. It gathers separately-named faults into one diagnosis with one fix-set, and its discriminator is precise: leading contributes to the answer's content, not merely how it is expressed.

Manages Complexity

Elicitation research has a long, scattered catalogue of named faults, each with its own remedy. Interviewer leading compresses the catalogue to one diagnosis and one decomposition: the channel is non-neutral, so every answer sums the respondent's state and the channel's response function, and each named fault is just a pathway adding content. The analyst tracks two quantities — the channel's contribution size and whether it can be controlled or must be modeled — from which handling reads off.

Abstract Reasoning

The concept licenses a diagnostic by inter-method divergence, reading a gap between differently-designed instruments as a direct measurement of the channel term. Its interventionist move applies one fix-set across all pathways, with a negative prediction that expression-only fixes fail. Its boundary-drawing move uses the two-term decomposition and a "where does the extra content enter?" test to place confirmation bias, Hawthorne, selection bias, and the statistical interviewer effect as neighbors.

Knowledge Transfer

Within elicitation methodology interviewer leading transfers as mechanism unusually broadly — the one diagnosis, two-term decomposition, and shared fix-set carrying across survey, eyewitness, clinical, market, and interrogation settings, only the surface pathway differing. Beyond it the mechanism genuinely recurs as co-instances (open-text fields, ML labeling rubrics, depositions, oral history), warranting the filed Elicitation Channel Contribution prime and its sibling Observer Effect, of which leading is the active-asking-channel sub-case; the methodology idiom and corrective protocols stay home.

Relationships to Other Abstractions

Current abstraction Interviewer Leading Domain-specific

Parents (1) — outgoing typed relations

• Interviewer Leading is a kind of Elicitation Channel Contribution Prime

  Interviewer leading is the human active-asking subtype of elicitation channel contribution, where wording, presupposition, order, options, affect, and register write channel content into an answer.

Hierarchy paths (2) — routes to 2 parentless roots

• Interviewer Leading → is a kind of Elicitation Channel Contribution → presupposes Observability

Show alternative path (1)

Neighborhood in Abstraction Space

Interviewer Leading sits in a sparse region of the domain-specific corpus (70^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Persuasion & Interpersonal Communication (17 abstractions)

Nearest neighbors

• Courtesy Bias — 0.85
• Social-Desirability Bias — 0.85
• Information Manipulation Theory — 0.83
• Message Sidedness — 0.83
• Central Route Processing — 0.83

Computed from structural-signature embeddings · 2026-10-25