4 ms·
Using numbers for probabilities is not a panacea either, when those numbers are not the result of a calculation. One person might say "99% of the time" to mean
by twiss 8y ago
Using numbers for probabilities is not a panacea either, when those numbers are not the result of a calculation. One person might say "99% of the time" to mean "almost always", and another person might say "9 out of 10 times" to mean exactly the same thing.
The problem is that it's hard to estimate the probability of something that you know happens "almost always" without doing the counting, which most people don't do. And hence it's difficult to gain an intuition for what 99% means exactly.
Furthermore, there is also the possibility that two people share the same intuition about a given word/phrase, but then disagree about which number describes that intuition. This could account for part of the results of this paper.
- c3534l 8y agoA good example of this is ratings on Amazon. You can give up to five stars, but most people just give 1 or 5 depending on if they were satisfied or dissatisfied. Additionally, people confuse precision for certainty and will perceive an estimate of 89.1% as more accurate than 90%.
- Izkata 8y agoYour examples are kinda funny, because that could easily be true. Now, if the second one was 90.0% on the other hand...
- taeric 8y agoWorse, many people that say "almost always" really just mean "I notice it when it happens, not when it doesn't."
- ska 8y agoThis is somewhere that common usage and mathematical usage diverge strongly. To anyone versed in measure theory talking about "almost always" wrt probability will mean that there is zero probability of the event not happening (even thought there may be other outcomes in the space).