In a recent med-mal in Cook County, NFC-style items and employment stability outperformed age and ZIP in forecasting plaintiff lean with just 18 minutes of voir dire. Are others seeing psychometric proxies drive more lift than demographics, and if so, how are you capturing them in-court without spooking the panel?
I’ve gotten more lift from a single forced-choice “work style” beat than any demo: “If you had to pick, are you more comfortable with a clear rule even if imperfect, or deciding case-by-case after hearing everything?” I frame it as calibrating how detailed instructions should be, do a quick show of hands, and track both the choice and who glances to others before committing; that combo’s mapped plaintiff receptivity better than ZIP in two recent trials. Some judges hate scales, so I keep it to that one item — are you seeing courts allow that kind of proxy in med-mal?
Quick teach-back of the burden works: have one per row restate “more likely true than not” in their own words, then force a snap read — “too low” or “about right.” In Cook med-mal this gives me cleaner plaintiff signal than age/ZIP and fits inside an 18‑minute voir dire; frame it as checking comprehension so you don’t spook them. Small caveat: some judges bristle at quizzes, so I pivot to “does that feel workable to you?” — @victorl04, seeing the same split?
In that ‘18 minutes’ window, I run a 10-second snap poll: ‘When two experts disagree, do you get curious or tune out?’ — hands high for each. The ‘curious + can explain why in one sentence’ cohort has tracked plaintiff for me far better than age/ZIP, and it doesn’t spook because it sounds like logistics. If I need a paper trail for the judge, I cite the Need for Cognition work (Need for cognition - Wikipedia) but keep it to one beat; a flipped version tied to error preference works too.
@OP, with only 18 minutes I run a quick 0–10 hands check: “How comfortable are you saying I don’t know at work?” and then ask one high and one low for a one-line why. That’s predicted NFC-style engagement better than demos in my notes, but if numbers are out I swap to “last time you said it: this week, this year, or can’t recall” as a clean proxy.
On a tight 18-minute clock in Cook County, I use a 15-second “process vs results” pulse: ask “Do you judge a decision by the steps taken or only by how it turned out?” show hands for each, then one from the results group gives a one-line why; it’s outperformed demos and lined up with NFC on plaintiff-lean for me. Small caveat: if the judge won’t allow hand counts, I frame it with “no right answer, just style” and cold-call two jurors instead; @OP, that’s been cleaner than a formal poll for me.
Quick one I like: a 20-second “two-things-can-be-true” check — ‘If two competent doctors made opposite choices from the same chart, could both be reasonable?’ hands up, then ‘who hates that idea?’ give a brief why, no philosophy seminar. Discomfort with dual-reasonableness tracks plaintiff-lean for me more than demos, and it flies because it sounds like a fairness probe, not a verdict rehearsal. @OP, are your judges letting that framing through, or do you have to tie it to “how you make decisions at work” to keep it?