Bayesian Precision in High-Stakes Decision Making

追加: 2026-05-17

[00:00:00]Well, I mean, I just feel like I almost want to do like a PSA really quick for all the grad students out there listening, which is you can use Bayesian methods to augment small samples because this is there's no other higher stake than, you know, what you're doing with astronauts. And if it applies with what Suzanne is working on, it probably applies for your silly study on graduate students as well. So, Well, hopefully not silly. I mean, you know, ask ask good questions about problems and it's not a silly study.

[00:00:30]Find out what the problem is and provide a solution. But yeah, no, it's um what I one of the thing I like about a Bayesian approach to is is we can quantify [music] and express our uncertainty. So, so the other thing to know is sometimes you don't know, >> [music] >> right? We always want to race to conclusions and be so proud of our like big effect we [music] found in something. But it could be that we have what we think will happen [music] in this data and then we update that with the Artemis 2 data and we just say, you know what, with what we're seeing, we actually don't know yet. But but that's important for decision makers, [music] too. So, to to I I can I like Bayesian frameworks sometimes because you're you're expressing [music] your uncertainty, which in high-stakes environment like spaceflight is absolutely critical. I can't go in and make [music] a judgment call just because P is less than 0.05, you know, [music] and and change the history of spaceflight. What I need to do is express what I found and then express [music] my uncertainty around it. It's absolutely critical in an environment like this.