6 ms·

any sampling method is 'biased', thats why its a 'sample'

bentona 7y ago
This study not only selected for people who go out of their way to join a society based on high IQ, it also selected for people who open & complete email surveys. Yes, many practical sampling methods are biased, but it is not difficult for me to imagine that any result here is caused entirely by these biases.
- anewguy9000 7y ago
  i agree with your criticism in general -- actually it applies to virtually all psychological and medical studies. but if you think the physical disabilities are better explained by something else to do with this population sample than their iq, you could very clearly publish this theory
scarmig 7y ago
Mensa members are... Not representative of people with high IQs.
mindblock 7y ago
Yeah I don’t think smart people pay a membership fee to an organization to be certified as smart.
mlyle 7y ago
This is what my 4th grader just learned is called a "convenience sample" and all the ways it can be biased... :P If he can understand the shortcomings, then so can the rest of us.
richdougherty 7y ago
If you take a look at the definition and you'll see it's possible to have an unbiased sample. https://en.m.wikipedia.org/wiki/Sample_(statistics) https://en.m.wikipedia.org/wiki/Sample_(statistics)
- anewguy9000 7y ago
  how do you randomly sample high iq? by definition the population would have had to have taken a test. so already, even if the study looked at every known high iq individual in the world, it would still be biased -- for those with academic backgrounds, psychological histories, criminal backgrounds, and whatever else leads one to take an iq test. nb i am not advocating for this particular study nor their methodology.
  - zingermc 7y ago
    You would have to sample a lot of people (yes, there is a selection bias in the people that participate) and give all of them IQ tests.
  - dahart 7y ago
    What you meant to say is that IQ tests are biased, not that “any sampling method is biased”. IQ tests are biased, but statistical sampling methods can be unbiased.
  - Mountain_Skies 7y ago
    Are there any widespread test given in schools that are proxy IQ tests? Mensa accepts older ACT and SAT scores as they consider the current versions to be knowledge tests but older versions to be intelligence tests. Not sure if this is still the case but at one time all high school students in the State of New York were required to take the SAT prior to graduation. While not representative of the entire population, it would have been a good representation of the high school student slice of the population across the entire state. Perhaps that would have been good enough for some uses.

This study not only selected for people who go out of their way to join a society based on high IQ, it also selected for people who open & complete email surveys. Yes, many practical sampling methods are biased, but it is not difficult for me to imagine that any result here is caused entirely by these biases.

anewguy9000 7y ago
i agree with your criticism in general -- actually it applies to virtually all psychological and medical studies. but if you think the physical disabilities are better explained by something else to do with this population sample than their iq, you could very clearly publish this theory

scarmig 7y ago

Mensa members are... Not representative of people with high IQs.

mindblock 7y ago

Yeah I don’t think smart people pay a membership fee to an organization to be certified as smart.

mlyle 7y ago

This is what my 4th grader just learned is called a "convenience sample" and all the ways it can be biased... :P If he can understand the shortcomings, then so can the rest of us.

richdougherty 7y ago

If you take a look at the definition and you'll see it's possible to have an unbiased sample. https://en.m.wikipedia.org/wiki/Sample_(statistics) https://en.m.wikipedia.org/wiki/Sample_(statistics)

anewguy9000 7y ago
how do you randomly sample high iq? by definition the population would have had to have taken a test. so already, even if the study looked at every known high iq individual in the world, it would still be biased -- for those with academic backgrounds, psychological histories, criminal backgrounds, and whatever else leads one to take an iq test. nb i am not advocating for this particular study nor their methodology.
- zingermc 7y ago
  You would have to sample a lot of people (yes, there is a selection bias in the people that participate) and give all of them IQ tests.
- dahart 7y ago
  What you meant to say is that IQ tests are biased, not that “any sampling method is biased”. IQ tests are biased, but statistical sampling methods can be unbiased.
- Mountain_Skies 7y ago
  Are there any widespread test given in schools that are proxy IQ tests? Mensa accepts older ACT and SAT scores as they consider the current versions to be knowledge tests but older versions to be intelligence tests. Not sure if this is still the case but at one time all high school students in the State of New York were required to take the SAT prior to graduation. While not representative of the entire population, it would have been a good representation of the high school student slice of the population across the entire state. Perhaps that would have been good enough for some uses.