2 ms·
This reminds me of some study that showed something like 90% of drivers rated themselves as "above average". It seems like self-evaluations are actually measuri
by bumby 4y ago
This reminds me of some study that showed something like 90% of drivers rated themselves as "above average". It seems like self-evaluations are actually measuring confidence and not ability (which aren't always well correlated).
- Jensson 4y agoConfidence and ability has high but imperfect correlation. You see it clearly in the Dunning Kruger experiment, performance is way higher among the more confident groups. Just don't assume displayed confidence is a good measure of their actual confidence, many confident people act humble since it is socially advantageous to do so. Edit: Also note that confidence and performance could be 100% correlated even if everyone rated themselves as higher than average. Correlation doesn't mean accuracy, it means that we can calculate their performance based on their confidence.
- bumby 4y ago>Dunning Kruger experiment Does this not show a J-shaped relationship between confidence and ability to accurately assess skill? [1] shows a weak negative correlation (strongest between multiple studies was -0.36). [1] https://www.sciencedirect.com/science/article/abs/pii/S0167487013000949?via%3Dihub https://www.sciencedirect.com/science/article/abs/pii/S01674...