Intuition Check

Your win is real, but one correct call doesn't validate the mechanism. Let's separate signal from noise using the Kahneman-Klein test. First, environment validity: the ER is a high-regularity setting—aortic dissections present with consistent cues (tearing pain, radiation, pulse deficits, risk factors). Second, feedback quality: do you consistently learn the outcome of every CT you order or decline? If you only get feedback on the positives, your pattern library is skewed.

Here's the sharper test: can you articulate which specific cues triggered your recognition? If you say "something felt off," that's bias. If you say "the pain quality was atypical for MI, plus radiation to the back and a subtle BP differential," that's recognition. Write those cues down after each case.

When the senior doc disagrees, don't default to hierarchy or your gut—ask what pattern they see. Their disagreement may be based on a different, equally valid pattern. Use your intuition as a hypothesis generator, not a verdict. Run it through protocol as a checklist. If your cue list is specific and your feedback loop is tight, trust it enough to advocate—but let the CT be your arbiter, not your conviction.

Stop asking what the feeling "means" and start asking where it comes from. Your gut is not a verdict—it's a cue that points to a pattern. The useful move is to decompose it.

First, name the specific cue. Is it his eye contact, his speed of speech, a detail in the term sheet, or the way he deflected a question? Vague unease is noise; a concrete observation is signal. Write down the exact moment you felt the shift.

Second, run the RPD test: Have you closed many partnerships in this exact industry, with this deal structure, against this type of counterpart? If yes, your pattern library is valid—trust the recognition. If this is your third deal ever, your "burned before" memory is a sample size of one, not a pattern.

Third, separate fatigue from data. Re-read the contract tomorrow morning, rested, and check: does the specific cue still bother you? If it survives a fresh read, it's signal. If it evaporates, it was projection.

Finally, don't decide on intuition alone. Use it to generate hypotheses—"he might be hiding something about X"—then verify with a direct question and a reference call. Intuition tells you where to look, not what to conclude.

Your instinct isn't a guess—it's pattern recognition built from years of watching kids. But here's the key: you're in a high-validity environment for noticing change, not for diagnosing cause. The recognition that something shifted is trustworthy. The inference that it's home-related is a leap. Use Klein's RPD: recognize the pattern (sudden behavioral shift in a quiet student), then simulate—have you seen this exact pattern before, and what followed? If yes, act on the recognition, not the explanation. The safe move is low-cost, high-information: a private, non-accusatory check-in. "I've noticed you seem off lately—anything you want to talk about?" That's acting on intuition without overcommitting to a story. Delay only if the behavior is ambiguous and you lack a baseline. But you have a baseline—you said "usually quiet." That contrast is your cue. Trust the cue, verify the cause. Awkwardness comes from asserting a diagnosis, not from asking a caring question.