Topic: POLL: What's the Average IQ of a bitcointalk.org poster (Read 2470 times)

Yes, I've often thought the IQs here are Cauchy distributed.

And that's not even an IQ score.

i'm glad i can be of service

i was actually surprised at how deep the mathematics discussion got in the first page of the thread

omg, thank you for this thread.  rofl, I can count two potatoes!

I'm going to have to assume you're either an idiot or a troll, if you still don't understand why I chose those examples after I explained it three times, and especially if you seriously think that a "random" sample means excluding only the most extreme members of a set. Either way, it's not worth my time to continue trying to educate you.

missing this as an option.

Standard deviation is calculated by the square root of the average of the squared differences of the mean.

And I've been using my iphone, which doesnt have a standard deviation button as far as I know.


Actually, I did do the standard deviation. 78.71 is the population standard deviation. But that number is what should be used, so thank you for correcting my mistake, even if you did it in a condescending way.



Ok, my mistake. I should not have added numbers.


Why do you choose which numbers are removed? Say the entire population of Earth is 10 people: 1, 2, 80, 90, 100, 110, 120, 130, 299, 300. Population Standard Deviation is 97.74, like you said. The Standard Deviation is 103.02. 8 people use bitcoin talk. The genius and idiot don't. 2, 80, 90, 100, 110, 120, 130, 299. Population Standard Deviation (even though it should be disregarded) is 78.23 . Standard Deviation is 83.63 . Both are lower, which is to be expected. Why would you remove the average members? It's cool that you think that bitcointalk has "fewer near-average members", but that's unrealistic. There will be people on bitcointalk with average IQs, and they will make up the majority of the population. There will not be people with IQs of 1 or 300. In fact, if anything, it makes sense to remove the 2 and 299 as well.

Of course, even if you don't agree with that don't call me an idiot because I am doing the exact same thing that you are doing- removing numbers to suit our arguments. But the numbers I remove actually make sense. There are average members on bitcointalk. There are not people with IQs of 1 or 300, so those should be removed.


The old "US schools suck".Too I actually took a IB course in high school because my school didn't offer American courses at higher levels. Blame Europe.

Here's the thing. If I have 1, 2, 3, 4, 5 as my population numbers I have a range of 4, and a population standard deviation of 1.41 . If I want to change the range, I have to remove the outliers. If I remove the middle numbers like you did that does jack shit to the range. So if I remove 1, so that my range is 3, I have a new standard deviation (not population) of 1.29.

You're right, a smaller range does not necessarily imply a smaller standard deviation. But if the only numbers you remove are outliers so that you can change the range (which would make sense in bitcointalk) then the standard deviation will go down.

Also, stop being so condescending. Really? Your first set of numbers were pulled out of your ass. You picked easy to calculate numbers that were irrelevant just to make a point. I used numbers that make some sense, and you called me out for simple mistakes.

Your ability to press a single button on a calculator neither impresses me nor convinces me that you know what that button does.

Your ability to press the correct button on your calculator doesn't impress me either. The standard deviation of this population is actually 78.71.

That is not a sample. A sample is a subset of the population. That means that every member of the sample must also be a member of the population. Let's start by taking the superset of your group and the original population: 1, 2, 80, 90, 100, 110, 120, 130, 299, 300. The standard deviation is 97.74.

Now, my point, in case your reading skills are as bad as your math skills, was that the forum contains fewer near-average members. Of course the standard deviation is going to be lower if you remove the extremes but don't remove the near-average values. So let's try that: 2, 80, 90, 120, 130, 299. We've removed the minimum and maximum, and also a couple near the average. The standard deviation of this sample is 98.56, higher than the population. It is higher than the population despite having a smaller range because it has fewer near-average members. If you can't understand that, then there's no hope for you.

Just out of curiosity, where did you get this idea that a smaller range necessarily implies a lower standard deviation? Are schools in the U.S. really that bad?

Heh. Been there, have had others done that.

I think I do understand what standard deviation is... After all, somehow I was able to calculate it. Let's say you have this set of numbers, which represents IQ in the total population. You have someone who's braindead, a bunch of people in the middle, and a genius.

1, 80, 90, 100, 110, 120, 130, 300

There is a big difference between the highest and lowest. The mean is 116.375 and the deviation is 84.14. The overall difference is obviously 299.

Take Bitcointalk. We have no genius, and no idiot.

80, 90, 100, 110, 120, 130

The average drops to 105. The deviation drops to 18.70.

Let's say bitcointalk has an idiot, and has a genius. But they cannot possibly be the dumbest or the smartest in the total population. This is a sample.

2, 80, 90, 100, 110, 120, 130, 299

Average is 116.375 and the deviation is 83.63. It is a tiny bit lower. The difference is 297 instead of 299. The average is the same, but the deviation is lower.

Standard deviation is obviously going to be lower when you have less of a range. And bitcointalk cannot possibly have a lower range of values than the population. That is impossible.

Any sample, whether its bitcointalk, NASA, your town, whatever will always have a lower deviation than the general population. Your mistake was assuming that the dumbest person and the smartest are on this forum. They aren't.

I don't think you understand what standard deviation means. It is not the difference between the minimum and maximum values. It is simply a measure of the average variance from the mean. A random sample of a population will theoretically have exactly the same standard deviation as the population, even though it will have a smaller range between the minimum and maximum, since you're removing just as many near-average members as extreme ones, so it balances out. But this forum is not a random sample. Nor does it require the smartest and dumbest people on the planet in order to have a higher standard deviation than average. A higher standard deviation just means there is more variation than average, not that the variation is the greatest range possible.

But in the actual population, there is a greater difference between the lowest and the highest. In your example, the second set of numbers is the bitcointalk population, correct? You're wrong to assume that there is a "1" here or a "49". The smartest person on Earth and the dumbest are not on these forums. Bitcointalk is more "average" than the entire population, simply because it will lack extremes. It would make more sense to not include the 1 and 49. Instead the set should be 4, 9, 25, 36 which has a deviation of 14.71 (rounded). Of course, these are just random numbers, and that's not how the actual population is distributed. But you should still see the main point of how bitcointalk cannot possibly have a greater difference between the highest and lowest than any other population.

Wrong. Consider a population consisting of square numbers less than 50: 1, 4, 9, 16, 25, 36, 49. The mean is 20 and the standard deviation is 16.37.

Now let's take a sample consisting of the odd numbers in this population: 1, 9, 25, 49. The mean is 21 and the standard deviation is 21.17. That's higher than the standard deviation of the population! Why? Because this particular sample contains the most extreme numbers of the population (1 and 49) but the number closest to the mean of the population (16) is not present. By containing more "extreme" members and fewer "average" ones, the standard deviation is higher.

The point I was making in my post is that this forum contains fewer "average" members than the rest of the population, ie, most people here are either really smart, or really stupid. I wouldn't even be particularly surprised if IQs on this forum follow a bimodal distribution.

I don't think the standard deviation of a sample group can be higher than that of a population group. The population would cover many more people, while the sample only covers a portion. The sample would be equal to or less than the population standard deviation.

The people that actually mine using a setup they put together are clearly more likely to be more intelligent versus someone that is just an investor or buys an ASIC. Most computer users cannot even find files. Really annoying. But having said the above, maybe the investor-only types are smarter.  lol

I'd say "self proclaimed genius" is pretty accurate. In the other thread about intelligence on Bitcointalk people were claiming to have IQs of over 140...

Agree, was thinking the same thing but I don't know if couldn't have said it "Nicer" than that!

 

They are considered geniuses when create the IQ tests for money?

That's not absurd?

I think the mean IQ of bitcoin.org posters is about the same as the rest of the population, but the standard deviation is much higher.