«

[Quintopotere]

Remember that the members of this forum have to add 20 points to the IQ test result!

I'm sure that playing oldies and write in this forum improve our intelligence! k:

[TheSmyth]

Originally posted by Quintopotere@May 13 2005, 03:09 PM
Remember that the members of this forum have to add 20 points to the IQ test result!

I'm sure that playing oldies and write in this forum improve our intelligence! k:

It might increase my IQ, but it doesn't mean my work gets done...
*Mumbles to self* "Damn stupid, having to work for a living........."

[Quintopotere]

Originally posted by TheSmyth@May 13 2005, 03:12 PM
It might increase my IQ, but it doesn't mean my work gets done...

Some genius die poors... maybe i have to worry about this...

[xcom freak]

Originally posted by Evad@May 13 2005, 05:00 AM
Hi guys. I have scored quite well on a few IQ tests lately, and was wondering if they are just a sales ploy or actually accurate. They always seem to offer a more complete version that sub-catagorises the info, and elaborates on your intelligence for 9.95. Are these free tests somewhat accurate?
Just to narrow it down, the tests took about 20 minutes and comprised of about 40 questions.
there were no really difficult equations, just pattern recognition and word association, and a few picture based questions.

I studied these in my stat course. If you get different questions everytime you start your test. Youll have to take the test 10 times then calculate the 'ecart-type' i think it is the same in english as in french. YOu have to divide ecart-type by rout3 then multiply it by 1.28.


You calculate the average then you do average (of 10) + - the result you got.
Youll be sure at 95% that your IQ is situated between average-result and average+result.

Reading it again dont make much sense but if you want to do the test 10 times and send me your 10 values ill tell you what's your iq.

[Quintopotere]

Originally posted by xcom freak@May 13 2005, 03:54 PM
'ecart-type'

This have to be the "standard deviation" or "experimental variance"

[xcom freak]

thats it thank you!

Anyway all this math was meant to show that your 'one' result doesn't meen anything and is not accurate at all.

[a1s]

Originally posted by xcom freak+May 13 2005, 03:54 PM--></div><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td>QUOTE (xcom freak @ May 13 2005, 03:54 PM)</td></tr><tr><td id='QUOTE'>YOu have to divide ecart-type by rout3 then multiply it by 1.28.
[/b]

since the resutlt isn't very precise to begin with, why not replace that with a singe operation of division by 1.35?

<!--QuoteBegin-xcom freak@May 13 2005, 03:54 PM
do the test 10 times and send me your 10 values ill tell you what's your iq.[/quote]
we're not falling for that one

[Sebatianos]

A word about the IQ tests...

They are quite inacurate. The thing is that they basically test if you posses the typical knowlage of the sorroundings you're in. A person from let's say East Timor will undoubtably have different general knowledge then a person from North Dakota.

So if you want to make an accurate test it has to be addopted to a certain social, age and geographically defined group.

Such general tests can only show if you're good at comprehanding some logical problems and the thing is, that most of this can be thought. Real IQ test should also show if you have more knowledge then you should have aquired by learning (so if you do three similar tests you'll be learnd on how to do the fourth similar test so the resoult will be better).

But you can't be all that bad if you score high...

[A. J. Raffles]

Originally posted by Sebatianos@May 13 2005, 09:28 PM
But you can't be all that bad if you score high...

That's probably most people's attitude towards IQ tests in a nutshell.

[xcom freak]

Originally posted by a1s+May 13 2005, 09:04 PM--></div><table border='0' align='center' width='95%' cellpadding='3' cellspacing='1'><tr><td>QUOTE (a1s @ May 13 2005, 09:04 PM)</td></tr><tr><td id='QUOTE'>

since the resutlt isn't very precise to begin with, why not replace that with a singe operation of division by 1.35?

<!--QuoteBegin-xcom freak@May 13 2005, 03:54 PM
do the test 10 times and send me your 10 values ill tell you what's your iq.

we're not falling for that one [/b][/quote]
It is an idea but you multiply by 1.28 only if you want to be at 95% sure that your iq is in this interval. You have to multiply by different values if you want different pourcentages.

It doen't take a lot of time so i don't think i would trick anybody if they sent me their 10 results except :evil: