Uploaded by umneem
800x2153 JPG 325 kBInterested in advertising on Derpibooru? Click here for information!
![Techy Cutie Pony Collection!](https://derpicdn.net/spns/2022/11/21/1669059816166133097826956.gif)
Help fund the $15 daily operational cost of Derpibooru - support us financially!
Description
Information bias
Tags
+-SH safe2221063 +-SH artist:umneem127 +-SH apple bloom61177 +-SH dinky hooves4731 +-SH scootaloo59699 +-SH sweetie belle57683 +-SH earth pony524742 +-SH pegasus517632 +-SH pony1651622 +-SH unicorn560501 +-SH comic:my rational pony88 +-SH g42081519 +-SH 4koma747 +-SH bandage7998 +-SH comic138137 +-SH cutie mark crusaders22796 +-SH doctor899
Loading...
Loading...
I believe Dinky thinks about that right now…
It doesn’t depend on anyone’s personal experience, and that’s the point of this strip. To know that it’s 80% likely to be disease 1, the relevant tests are already done. Figuring out which of the two runners up is second most likely doesn’t change that you’ve already done the tests that matter, and you already know it’s 80% likely to be disease 1. It’s straight math; doing the tests makes you feel like you did something useful, but all you did is waste money and time doing an unnecessary test. What you’re mistaking it for is a totally different problem in which you have a choice to do a better test for disease 1. If you solve the wrong problem because you don’t like the problem reality gives you (no medical test is ever 100%), you get an F on the final for Reality 101. But if it makes you feel better to say it’s 96% likely to be disease 1 and only 2% likely to be disease 2, and 2% likely to be disease 3, try it with those numbers. That’s the same problem.
The symptoms could be disease 1 (96% of all cases of those symptoms are disease 1), disease 2 (2%), or disease 3 (2%). The only test that’s available can tell disease 2 from disease 3. You do the test. It says it’s not disease 2. What you’re left with then is it’s 96% likely to be disease 1 and 4% likely to be disease 3. You still treat for disease 1, just as if you’d never done the test. Let’s say it says it’s not disease 3. Then you’re left with 96% chance of disease 1, and 4% chance of disease 3. So you also still treat for disease 1, just as if you’d never done the test. So if the test is expensive, you don’t want to do it. But even if it’s free, you also don’t want to do it. Doing it at all serves no actual purpose. Regardless of what the test says, your only reasonable option is to treat for disease 1. The math gives the same answer with 80%, 96%, and really anything above 50%.
Where it does depend is if the cost function isn’t equal. Let’s complicate it a bit. If disease 1 will clear up on its own and treating it will only ease the symptoms, but disease 3 is fatal without treatment, then it can be worthwhile to rule out disease 3 even if it’s only 10% likely. If you wind up ruling out disease 3, you’re left with 80% chance of disease 1 and 20% chance of disease 2. You treat for disease 1, and 80% likely the patient feels better. On the other hoof, if you wind up with an 80% chance of disease 1 and a 20% chance of fatal disease 3, you may consider treating for disease 3 instead, even though it’s very likely to cause the patient considerable discomfort and do no actual good; assuming the rare chance that it really is disease 3, the alternative is so much worse. But as long as each of the three diseases is about equally bad and has its own treatment, knowing whether it’s 20% disease 2 or 20% disease 3 makes absolutely no difference in what you’re going to ultimately do. Unless you’re completely daft, you’re still going to go with the 80% chance that it’s disease 1, not the 20% chance of whichever other disease it is. (Another interesting variant where the test might be worthwhile, but only if it’s not too expensive, is if disease 3 is fatal, but untreatable. Play with that one yourself.)
Someone’s personal experience is called an anecdote, and that’s a totally different logical fallacy. You remember the 2 weird cases where the doctor did an unnecessary X-ray, and just by luck happened to find some rare, odd thing, and knowing that helped to save the patient’s life. You remember those because they were immediate and interesting. You don’t remember the 100,000 people whose X-rays showed nothing, because that’s not interesting. You don’t remember the 1,000 of them who needlessly died of cancer because of the unnecessary X-rays, because that happened years later in the middle of another 5,000 who would’ve gotten cancer anyway, so you don’t know who they are, just that 6,000 people got cancer when only 5,000 would be expected to get cancer. You saved 2 lives and took 1,000. Just because you remember the 2 better than the 1,000 doesn’t make that a good trade-off.
Because it comes form the horses mouth.
And is then filtered through Applebloom.
That would require intelligence and actual studies instead of this flawed garbage.
nah man she’s like
“Shit I am so high
I don’t even know what you bitches are going on about
Man that light is just the most amazing thing”
Exactly why I’m not a fan of this bias. Problems like that can arise and makes things a nightmare, exactly why the saying “Better safe then sorry” exists.
To many variables to worry about, sometimes. Ofcourse it would be stupid to apply this sort of thinking to everything, hence why it’s intelligent to think differently about different things.
That’s mostly only an issue when diseases/disorders that present with similar symptoms have wildly different causes and treatments; you don’t want to give someone high octane antibiotics for viral meningitis, for example.
Well if it explained it better I might of understood the bloody thing…
Though something about that bothers me, diseases can act similarly like different ones thus making it difficult to tell which is which. Eliminating possible options when there is doubt is the smart move IMHO. Most cases, this varies from one person’s experience from one another, it likes to turn out that, yeah, doing that extra bit of work payed off because you found out that the obvious answer was the wrong one.
This is the majority of my experience anyway.
@Rpground
Actually, it’s the opposite. It’s, “make sure you stop gathering information, when you already have enough information to make a decision.” The classic statement of this problem has three remaining diagnoses. It’s 80% likely to be disease 1, but if it’s not, a very expensive test will tell if it’s disease 2 or disease 3 (both of which are equally likely). Do you do the test?
The answer is no. The cost of the test is irrelevant; even if you can do the test for free, it will only rule out a 10% chance of disease 2 or a 10% chance of disease 3. You’re still left with the fact you already have: that it’s 80% likely to be disease 1, and you don’t need any further tests to know that. So you do no further tests and treat for disease 1. (That happens to be true even if it’s only 50.0000…0001% likely to be disease 1. It just psychologically feels even worse. However, no matter what the test says or how cheap it is, the test cannot at that point change the fact that it’s most likely to be disease 1.)
I really like the three-eyed martian ponies. I’m assuming Apple Bloom drew them.
It’s actually a typo, thank you for noticing.
That sounds painful, you should go to the CMC Clinic.
THESE COMICS MAKE ME WANT TO SLAM MY HEAD AGAINST A WALL.
Uhh…ok…but, fucking wat?
I think I got “Make sure you have enough information to proceed.” But it’s so poorly explained I’m not even sure if it’s trying to make that statement or the opposite. Fuck everything just bang your head against a wall until it breaks, you’ll get it eventually…