Module 4 · ~20 min · Structured Interviewer Certification
Behavioral & Situational Interviewing in Practice
Module 3 gave you what to score. This module gives you how to get the evidence: behavioral and situational questions, the probing that turns a rehearsed story into real signal, and the discipline of separating what a candidate did from what you assume about them.
AI Instructor: "Most interviews fail here. The questions are fine — but the interviewer accepts the first polished answer, never probes, and writes down a conclusion ('strong communicator') instead of evidence ('used teach-back, checked understanding twice'). Today we fix that."
What you'll be able to do
- Distinguish behavioral from situational questions and know when each is the right tool.
- Use STAR/SOAR to structure a candidate's answer and spot where it's thin.
- Probe a vague or rehearsed answer down to real, scorable evidence.
- Separate evidence (what was said/done) from inference (your interpretation) in your notes.
- Run a clinical case / chart-talk station that assesses judgment without becoming a pop quiz.
Key concept 1 — Two tools: behavioral and situational
| Behavioral | Situational | |
|---|---|---|
| The ask | "Tell me about a time you…" (past) | "What would you do if…" (hypothetical) |
| Premise | Past behavior predicts future behavior. | Judgment/intent under a scenario they may not have lived yet. |
| Best for | Coachability, teamwork, professionalism, values — things they've done. | Autonomy, surge, escalation, ambiguity — UC scenarios, and for early-career candidates without a long track record. |
| Risk | Rehearsed/borrowed stories; "we" instead of "I." | Aspirational answers ("I'd always…"); tests knowledge, not habit. |
Use both: behavioral for what they've actually done, situational for UC-specific judgment they may not have a story for yet. Pair them on the same competency to cross-check (a candidate who describes good escalation should also have done it).
"Situational questions are especially fair to new grads and to APPs moving into UC — they haven't lived every scenario, but they can still show you how they'd reason through one."
Key concept 2 — STAR / SOAR: a frame for the answer
A complete behavioral answer has four parts. If a part is missing, that's exactly where you probe.
| STAR | SOAR (variant) | What you're listening for |
|---|---|---|
| Situation | Situation | Real context, recent, relevant — not generic. |
| Task | Obstacle / Objective | What was actually at stake; their specific responsibility. |
| Action | Action | Their action ("I"), the decisions, the reasoning. This is the meat. |
| Result | Result | Outcome, what they learned, what they'd change. |
The "I vs. we" tell: candidates who narrate in "we" are often describing the team's work, not theirs. Probe: "What was your specific part?"
Key concept 3 — Probing: where the real signal lives
The first answer is the rehearsed one. Probes get you the rest. Four probe types:
| Probe | Use it when… | Sounds like |
|---|---|---|
| Depth | The answer is high-level. | "Walk me through exactly what you did next." |
| Evidence | You hear a claim, not a fact. | "What specifically did you say to the nurse?" |
| Reasoning | You need the why, not just the what. | "What made you decide to transfer rather than observe?" |
| Contrary | The story is suspiciously clean. | "What part didn't go well? What would you change?" |
Silence is a probe. After an answer, a 2–3 second pause often surfaces the real detail the candidate was about to leave out.
"You don't need more questions. You need to finish the one you asked. One well-probed question beats five surface questions."
Key concept 4 — Evidence vs. inference (the note-taking discipline)
Evidence = what the candidate said or did, captured close to verbatim. Inference = your interpretation of it. You score evidence; you flag inference to test it.
| You wrote… | That's… | Better |
|---|---|---|
| "Great communicator" | Inference | "Used teach-back; checked understanding twice; slowed down for the anxious parent." |
| "Seems arrogant" | Inference (bias risk) | "When asked about an error, said 'I don't really make those.'" |
| "Strong team player" | Inference | "Pulled the charge nurse aside, owned the delay, reworked room assignments." |
The rule: write the quote, not the verdict. The verdict comes later, when you match evidence to the BARS anchor. Inference written as fact is how halo and affinity bias sneak into a scorecard.
Key concept 5 — The approved question bank (role-calibrated)
A starter bank mapped to the eight competencies. Use the same core questions per role for fairness; calibrate expectations to MD/DO vs. NP vs. PA scope.
| Competency | Behavioral | Situational |
|---|---|---|
| #2 Communication | "Tell me about explaining a scary or 'nothing's wrong' result to an anxious patient." | "A patient is angry about a long wait and a normal work-up. What do you do?" |
| #3 Professionalism | "Tell me about a clinical error or near-miss you were part of." | "You realize after discharge you misread an X-ray. Next steps?" |
| #4 Teamwork | "Describe a disagreement with a nurse or collaborating MD/APP." | "A nurse pushes back on your plan in front of the patient. What now?" |
| #5 Coachability | "Tell me about feedback that was hard to hear — what changed?" | "Your medical director flags your documentation. How do you respond?" |
| #6 Autonomy | "A time you had to decide alone with incomplete information." | "9pm, sole provider, transfer-borderline patient, no one to curbside." |
| #7 Adaptability/surge | "Your busiest, most chaotic shift — how did you keep it safe?" | "Three high-acuity patients arrive at once with one nurse. Prioritize." |
| #8 Values | "A time you spoke up about something unsafe or unfair." | "You see a colleague cutting a corner that risks a patient. What do you do?" |
"Competency #1 — clinical judgment — isn't on this list on purpose. A conversation can't validly measure clinical reasoning. That's what the case station below is for."
Clinical Case / Chart-Talk Station: Assessing Judgment Without Turning It Into a Pop Quiz
Who runs this: a qualified clinician (or the AI clinical-reasoning simulator in this course) — never TA. Competency #1 is the one competency that requires a clinician rater. TA screens logistics, licensure, and interest; clinical judgment is scored here.
What this station measures (and what it doesn't)
The station assesses reasoning, not recall:
- Clinical reasoning — how they build and narrow a differential.
- Disposition — treat / observe / refer / transfer / 911, and why.
- Safety judgment — red-flag screening, safety-netting, return precautions.
- Escalation — when and how they reach for help/transfer.
- Ambiguity tolerance — acting reasonably with incomplete information.
- Scope awareness — knowing the edge of their training and collaboration model.
It is not a board exam. Don't reward memorized doses, zebra diagnoses, or trivia. A candidate who says "I'd look up the exact dose" while reasoning safely scores higher than one who rattles off facts but misses a red flag.
How to run it
- Present a common UC case, not an exotic one. Give realistic, incomplete information (vitals, brief history) and let them ask for more.
- Think-aloud: "Walk me through your thinking out loud — what's on your differential, what you'd do, and your disposition."
- Add a wrinkle once they commit: an abnormal vital, a worsening trend, a resource constraint ("the CT is down"), or a social barrier. Watch whether they re-reason or anchor.
- Probe reasoning, not facts: "What would change your mind?" "What's the worst thing this could be?" "What's your safety net if you discharge?"
- Score with the BARS below — on reasoning quality, not on whether they "got the answer."
Role calibration (MD/DO vs. NP vs. PA)
| Role | What "Meets Standard" looks like |
|---|---|
| MD/DO | Independent breadth across acuity; owns complex/ambiguous cases; clear escalation when truly indicated. |
| NP | Sound autonomous reasoning within scope and the state collaboration model; appropriate, non-excessive consultation. |
| PA | Sound reasoning within the physician-collaboration framework; consults appropriately and can act safely while awaiting input. |
Same case, role-calibrated bar: appropriate consultation is a green flag for NP/PA, not a deduction. The deduction is inability to reason or act safely — at any level.
Scoring clinical reasoning — BARS 1–5
| 1 | Unsafe or guideline-discordant; anchors and won't revise; misses red flags; no safety-net; can't justify disposition. |
| 2 | Knows facts but reasoning wobbles under the wrinkle; over- or under-triages; weak safety-netting; thin escalation logic. |
| 3 | Sound, guideline-concordant approach to a common presentation; appropriate work-up and disposition; adequate safety net; escalates when needed. |
| 4 | Strong structured reasoning across the wrinkle; explicit differential and red-flag screen; resource-aware; clear escalation/return precautions. |
| 5 | Expert judgment with incomplete info; anticipates deterioration; re-reasons cleanly when the picture changes; scope-calibrated; teaches their thinking. |
Worked example — atypical chest pain, 58yo
Case: 58-year-old, vague chest discomfort 2 days, normal initial vitals, no prior records. Then the wrinkle: repeat BP 150/95, diaphoretic.
| Level | Response | Why |
|---|---|---|
| Strong (4–5) | "ACS is top of my differential until excluded — also PE, dissection, GI. ECG and troponin now, IV access, aspirin if no contraindication. With the diaphoresis and rising BP I'm treating this as high-risk and arranging transfer rather than discharging. If discharged anything, strict return precautions and a documented shared decision." | Structured differential, red-flag driven, re-reasons on the wrinkle, clear safe disposition + safety net. |
| Average (3) | "I'd get an ECG and troponin and probably send him out to the ED to be safe. Chest pain at his age I don't want to sit on." | Safe disposition and appropriate caution; reasoning is correct but shallow — light on differential and the why. |
| Concerning (1–2) | "Vitals are basically normal and it's been two days, so probably musculoskeletal. I'd give him something for the pain and have him follow up with his PCP." | Anchors on normal vitals, dismisses ACS red flags, ignores the diaphoresis/BP wrinkle, unsafe disposition, no safety net. |
"I'll present the case, reveal information only when the candidate asks, drop the wrinkle once they commit, and hand the transcript to Scoring/Calibration. I never coach mid-case or hint the answer."
Interactive — finish the probe
A candidate gives this answer to "Tell me about a disagreement with a nurse":
"Yeah, we had a situation once where there was some tension, but we worked it out and it was fine."
This is all surface. Write your next two probes before revealing the model probes.
Model probes: (Depth) "What was the disagreement actually about?" → (Evidence) "What did you specifically say or do?" → (Reasoning) "Why did you handle it that way?" → (Contrary) "Looking back, what would you do differently?"
Interactive — evidence-vs-inference sort
Tag each note as Evidence (E) or Inference (I):
- "Said 'I'd call my SP and wait' for the transfer scenario." → (E)
- "Not a confident decision-maker." → (I — test it; what did they say/do?)
- "Named teach-back and gave a return-precaution example." → (E)
- "Reminds me of our best APP." → (I — affinity flag; not scorable.)
- "Couldn't name a case they'd handle differently." → (E)
Score-the-answer exercise (Coachability, #5)
Question: "Tell me about feedback that was hard to hear — what did you change?" Candidate (an MD) answers:
"My director once said I was too slow on documentation. I disagreed honestly — I'd rather be thorough — but I sped up a bit to keep them happy. I think my notes were always fine."
Assign a BARS 1–5 for Coachability and justify with the candidate's words. What would you probe?
Expert key (~2 Developing): accepts feedback grudgingly, reframes the change as appeasement ("to keep them happy") not growth, and dismisses the premise ("my notes were always fine") — little evidence of genuine reflection or durable change. Evidence: "I disagreed honestly"; "to keep them happy"; "always fine." Not a 1 (no outright rejection/defensiveness), not a 3 (a 3 needs a real growth area they acted on). Probe: "What specifically changed in your documentation, and is it still in place?" — to test whether any real change occurred.
Reflection
- Think of your last interview: did you accept the first answer, or probe? Where would one more probe have changed your read?
- Look at how you take notes today. Are you writing quotes (evidence) or verdicts (inference)?
Knowledge check
- When is a situational question the better tool than a behavioral one?
- A candidate narrates a whole story in "we." What do you probe?
- Name two of the four probe types and when you'd use each.
- "Strong communicator" in your notes — what's wrong with it, and what should you write instead?
- Who should run the clinical case station, and what does it measure that a behavioral question can't?
- In the case station, why can "I'd look up the exact dose" score higher than reciting the dose from memory?
Module summary
Behavioral questions get what they've done; situational questions get how they'd reason; STAR/SOAR frames the answer and shows you where it's thin. Probing — depth, evidence, reasoning, contrary — turns a rehearsed story into scorable evidence, and the evidence-vs-inference discipline keeps bias out of your notes. The clinical case / chart-talk station, run by a clinician, scores reasoning (not recall) for Competency #1.
Key takeaways
- Use both behavioral and situational; pair them to cross-check.
- The first answer is rehearsed — probe to the real evidence.
- Write the quote, not the verdict.
- "I vs. we" reveals individual contribution.
- The case station scores reasoning and safety, not trivia — and a clinician runs it, never TA.
Readiness checkpoint (1–5)
The AI scores your probe drill, evidence-vs-inference sort, and score-the-answer, returning a Module 4 proficiency score with evidence, your probing/note-taking tendencies, and one targeted improvement. No pass/fail.