This is the EPLI research that you mentioned earlier. This is EPLI.

Now, here's where things take an interesting twist. And that is when we're first meeting someone and they're making a first impression, we're actually surprisingly accurate. About 70% of the time, we size people's global personality traits up pretty accurately and they would match that person. They would match what others think. But again, 70% if your kid brought home all these, you wouldn't be too thrilled. So 30% of the time we're wrong. So we're good. But here's the double-edged sword of first impressions. And that is, we are right, we're right more often than we're wrong, you know, a lot. However, once we're wrong, confirmation bias kicks in. Once we've made up our first impression, now our brains are not objective scientists looking for the truth. Our brains are lawyers vigorously defending the positions we hold, looking for things to confirm what we already believe. So if somebody makes a bad first impression, they can get stuck because the store that updates our beliefs is not open for business. And it's very, very difficult. And that leads to continuing problems because for us, if we meet somebody for the first time, they don't make a good impression. If we have the option, what might we do? Not hang out with them anymore. So our negative impressions are always going to be less accurate than our positive impressions. Your positive impressions, you see them again, you get the chance maybe to update your beliefs. So you're getting a bigger sample size versus negative impressions. It might be somebody who's having a bad day and they're not going to get a chance.