Elicitation technique¶
A method for drawing specified knowledge or preferences from people's responses or observed behavior in an inquiry.
Core Idea¶
An elicitation technique is a method for making relevant human knowledge, preferences, requirements, or behavior available to an inquiry. The analyst specifies a question and human source, chooses questions, an assigned task, or an observation protocol, and records what the person says, chooses, or does. The method may be an interview, questionnaire, focus group, naturally occurring field observation, task analysis, scenario, or structured expert probe. Direct verbal report and assigned performance are possible routes, not universal conditions: planned observation can reveal practice without interrupting it.
Hoffman and colleagues distinguish interviews, usual-task analysis, and constructed tasks in expert-knowledge work, and their Critical Decision Method study shows repeated probes around actual incidents yielding decision requirements and timelines. Elicitation is not equivalent to an entire research program or to the resulting fact claim. A captured answer can be shaped by recall, question wording, role, setting, and sampling. Document analysis is an adjacent supporting activity: reading existing text without a human encounter is indirect acquisition, while discussing that text with its author can be part of elicitation.
Structural Signature¶
Sig role-phrases:
- Inquiry target — States what knowledge, judgment, preference, or requirement is sought and for what decision. It is constitutive. Counterfactual: Unfocused conversation cannot be evaluated as a defined elicitation method.
- Human source and context — Identifies participants whose experience, choices, or task behavior bear on that target. It is constitutive. Counterfactual: A desk review of documents alone lacks the human source encounter.
- Elicitation setup — Uses questions, an assigned task, or a declared observation protocol to make relevant responses or naturally occurring behavior available. It is constitutive. Counterfactual: Unplanned incidental watching with no inquiry or recording rule does not establish a technique.
- Captured response or behavior — Records statements, choices, actions, or task traces under the chosen method. It is constitutive. Counterfactual: Without an observable contribution there is nothing to analyze.
- Interpretation and validity limit — Connects the elicited material to the inquiry while retaining sampling, framing, recall, and observer limits. It is boundary. Counterfactual: One elicited account cannot be treated as an exhaustive or automatically verified truth.
What It Is Not¶
- Not a document review alone. Existing text is an indirect source unless a person responds to or performs with it.
- Not necessarily an interview. Planned observation of ordinary work or a contrived task can elicit material without complete verbal self-report.
- Not an interaction technique. A user-interface gesture changes software state; elicitation gathers material for an inquiry.
- Not validated truth by default. Responses and observations have framing, recall, and sampling limits.
- Closest near-miss. Document analysis is the closest near miss: text reading alone supplies indirect evidence; asking its author to explain a marked passage would add a human elicitation encounter.
Scope of Application¶
- Expert knowledge. Draw out decision cues and task structure for analysis.
- Requirements inquiry. Ask stakeholders or observe work to identify needs and constraints.
- Preference research. Use declared prompts to surface judgments or choices.
- Field observation. Examine task behavior when explicit accounts omit tacit steps.
Clarity¶
Ask what the analyst is trying to learn, which people can supply it, which prompt, task, or observation protocol makes it visible, and what response or behavior is captured. Reading a document alone is the near miss: it is indirect acquisition. An interview is one technique; planned observation of naturally occurring behavior also qualifies without asking questions. Neither route guarantees the interpretation is complete or unbiased.
Manages Complexity¶
The method turns otherwise tacit or dispersed human knowledge into examinable evidence. Its apparent simplicity hides choices about participant selection, prompt or observation framing, recording, interpretation, and validation. A richly detailed incident interview may reveal cues that a checklist misses, yet can also suffer retrospective reconstruction; naturalistic observation may show action but not the participant's reasons.
Abstract Reasoning¶
- Define the information, knowledge, or preference being sought.
- Identify participants and the context in which they hold or express it.
- Select an interview prompt, assigned scenario/task, or planned naturalistic observation protocol suited to the target.
- Capture the resulting statements or behavior without silently replacing evidence with interpretation.
- Assess framing, recall, sampling, and cross-checks before making claims from the record.
Knowledge Transfer¶
The target–human source–elicitation setup–captured contribution pattern travels from cognitive task analysis to requirements gathering or field research only after participants, evidence type, and validity checks are redefined. Critical Decision Method probes appropriate to an incident cannot be copied unchanged into naturalistic observation or a short preference survey. A UI gesture and a static document may share interaction or information but do not become elicitation until they contribute to a designed human-source inquiry.
Examples¶
Canonical¶
An analyst wants to learn how an experienced operator recognized an unusual event. After the operator selects a real past incident, the analyst returns to key decision points with probes about noticed cues, alternatives, and uncertainty, recording the resulting account. The method elicits situated judgment; the recollection still needs comparison with other evidence rather than being treated as complete access to cognition. A different inquiry could instead observe the operator's naturally occurring work under a declared field protocol without assigning any task.
Mapped back: Inquiry target → recognition and choice during an unusual event; Human source and context → experienced operator recalling a concrete incident; Elicitation setup → repeated decision-point probes; field observation is a distinct allowed setup; Captured response or behavior → recorded cues, alternatives, and uncertainty; Interpretation and validity limit → retrospective account with recall and framing limits.
Applied / In Practice¶
Hoffman, Crandall, and Shadbolt's published Critical Decision Method case study uses multiple-pass retrospective interviews to elicit expert knowledge from critical incidents. It reports products such as time lines, situation-assessment records, and decision requirements for cognitive task analysis and system or instructional design. This is a documented methodological application; it does not show that every interview produces complete or unbiased expert knowledge.
Mapped back: Inquiry target → expert decision requirements in critical incidents; Human source and context → experienced participants described in the case study; Elicitation setup → Critical Decision Method multiple-pass probes; Captured response or behavior → event timeline and situation-assessment material; Interpretation and validity limit → method's quality and reliability concerns retained.
Structural Tensions¶
T1 — Structured Prompts versus Unanticipated Knowledge. A fixed interview can compare answers but may miss categories that an open incident account reveals.
Diagnostic: Does the chosen method leave room for the source's own distinctions?
T2 — Reported Reasoning versus Observed Performance. People may describe decisions differently from how they act in a task; each source has its own validity limits.
Diagnostic: Which evidence is being elicited and what can it support?
Structural–Framed Character¶
The skeleton is designed access to information held or expressed by another person. An elicitation technique uses prompts, tasks, interviews, or observation to draw knowledge, beliefs, preferences, or requirements from human responses or behavior. It is an approved unparented root because measurement is not present in every open narrative and interface interaction changes UI state rather than necessarily gathering evidence.
Evaluative weight: Captured answers need interpretation and validity checks; a prompt alone does not guarantee truth.
Human-practice-bound: Participants, inquiry goals, social context, and task design matter.
Institutional origin: Research and requirements practices supply named protocols with different evidence limits.
Vocabulary travels: Data extraction from a static database lacks a human-source inquiry unless people’s responses are being elicited.
Import versus recognize: The target–source–setup–contribution pattern transfers across methods only after those roles are redefined.
Its character: A deliberate human-inquiry method, not a prime for obtaining any information.
Structural Core vs. Domain Accent¶
Skeletal core. An investigator can structure conditions that make otherwise unavailable information observable.
Domain-bound accent. Elicitation uses a human source and a chosen interview, prompt, assigned task, or naturalistic observation protocol, recording a response or behavior relevant to a specified inquiry.
Why not prime. A UI gesture or database query may generate data without a designed human-source inquiry. Different elicitation methods also require their own validity checks.
Instantiates / Related Primes¶
-
Related — interaction technique. Both may involve prompts or actions, but the latter mediates an elementary user-interface task rather than gathering inquiry evidence.
-
Related — observation. Planned observation can be an eliciting route even when the human source performs only naturally occurring work.
Neighborhood in Abstraction Space¶
Elicitation technique sits in a crowded region of the domain-specific corpus (32nd percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.
Family — Communication, Learning & Information Practices (15 abstractions)
Nearest neighbors
- Appreciative inquiry in education — 0.89
- Rhizomatic learning — 0.89
- Problem-posing education — 0.89
- Content Analysis — 0.88
- Informational listening — 0.88
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
- Document analysis. Tell: Was a human participant asked or observed, or was existing text merely read?
- Interview. Tell: Is verbal questioning the whole method, or one of several elicitation forms?
- Interaction technique. Tell: Does the action reveal inquiry material or merely operate software?
- Requirement validation. Tell: Was information drawn out, or was an already stated requirement checked?
References¶
- Hoffman et al., Eliciting Knowledge from Experts: A Methodological Analysis (1995), author manuscript: https://eprints.soton.ac.uk/252301/1/El_20Know_20from_20Exp_20OBHD.pdf
- Hoffman, Crandall, and Shadbolt, Use of the Critical Decision Method to Elicit Expert Knowledge (1998): https://journals.sagepub.com/doi/10.1518/001872098779480442
- Frozen Wikipedia discovery revision: https://en.wikipedia.org/wiki/Elicitation_technique (revision 1367151192).
- Preserved source candidate: https://cdn.ymaws.com/www.scip.org/resource/collection/472B204C-541E-45B4-BEB3-445E51942C37/Elicitation-Applying-the-Art-and-Science-of-HUMINT-to-Business.pdf
- Preserved source candidate: https://scholarship.law.bu.edu/faculty_scholarship/781
- Preserved source candidate: https://www.informatika.web.id/2013/07/elisitasi-proses-dan-metode-untuk.html
- Preserved source candidate: https://web.archive.org/web/20241208194635/https://www.informatika.web.id/2013/07/elisitasi-proses-dan-metode-untuk.html