3 ms·
Experimental evidence and logic counters your point. It is wishful thinking to believe that most people are average but this is not the case. There are more pe
by Geminidog 6y ago
Experimental evidence and logic counters your point.
It is wishful thinking to believe that most people are average but this is not the case. There are more people on the right and left side of a bell curve than in the center.
When you account for all bins most people ARE not average though the "average" represents the biggest bin. (It also depends on how big you make your bins but in general if you make your bins a reasonable resolution more people live to the side of the average than in the center)
What this means is that there are people all over the spectrum, tons of people who are bad, and tons of people who excel and are better.
It is convenient to think otherwise but it does not mesh with reality. There are people who are better than you in all things just through sheer talent and there are people who are worse than you in all things also because of sheer talent.
Your uni anecdotal example also is biased. It also depends on the university you attend. Universities filter off of cognitive ability through grades and testing. Students within the same university are more or less roughly equivalent in ability due to going through the same filtering to attend the same university.
There is, however, a very evident difference in skill between students who go to different universities. If you ever transferred from a lower tier university to a higher one, you can feel this difference anecdotally. The difference in ability between students on average in say Stanford is markedly different then those in a community college.
- SoSoRoCoCo 6y ago> There are more people on the right and left side of a bell curve than in the center. That's almost tautology they way you worded it. Statistics has a term to describe "left and right", it is called standard deviation, and one standard deviation covers ~70% of the sample population. What is true is that half of any population is below average by whatever metric you choose to apply. (EDIT: for normal/Gaussian distributions)
- Geminidog 6y ago>What is true is that half of any population is below average by whatever metric you choose to apply. Not exactly, this is sort of true in practice but not true logically. Imagine a population of [1, 100, 100, 100] or a population of [1,1,1,1,999999]. Two examples where the population that is below average exceeds half the population and is below half the population. But this is just semantics. If the population follows a bell curve and at high resolution standard deviations, what you said is generally true.
- SoSoRoCoCo 6y agoYou are correct for non Gaussian distributions. IQ tests appear to be a normal dist. I was not specific enough. That's the old joke, "If Jeff Bezos walks into a dinner party, suddenly everyone is on average a billionaire!" Which used to be "Bill Gates", and before that "Warren Buffet". I wonder who will be the next person to fill in this joke. :)