So my father took me the the shrink for one of those IQ test

[Ziggy]

That's not what I meant. It is my understanding that 100 being the average is kind of like what a dollar is worth and inflation. If the general population gets much stupider then your score of 140 isn't worth as much any more, because 100 is still the average. At least, this is what was explained to me, I could be wrong. But IQ tests are pretty bogus any way.

[lordofduct]

Watch this:

http://www.openculture.com/2009/04/tale ... _luck.html


... who cares if you're intelligent, if you don't utilize it and learn stuff you'll still be dumb. You need ambition and dedication motha fucka.

[Limewater]

Even if general intelligence is increasing (I'm not sure that's true), 100 is still the average IQ score by definition because the IQ score is a measure of your intelligence in relation to the average person. It is a standardized score so the top of the bell curve falls right at the 100 mark.

100 is supposed to be the average. However, they have to renormalize the test periodically because the mean drifts away from 100. For the last several decades, the trend has been strongly upwards. The average is only ever right at 100 when they renormalize. It drifts substantially in the meantime.

[lordofduct]

Even if general intelligence is increasing (I'm not sure that's true), 100 is still the average IQ score by definition because the IQ score is a measure of your intelligence in relation to the average person. It is a standardized score so the top of the bell curve falls right at the 100 mark.

100 is supposed to be the average. However, they have to renormalize the test periodically because the mean drifts away from 100. For the last several decades, the trend has been strongly upwards. The average is only ever right at 100 when they renormalize. It drifts substantially in the meantime.

Yeah, that's what he said basically... inferring the drifting part. ::face palm::

[Limewater]

Yeah, that's what he said basically... inferring the drifting part. ::face palm::

No, that's not what he said at all. He said that 100 is by definition the mean, making no mention of drift or temporal sampling.

The whole point of my post was to point out coarse granularity of sampling in time. They don't really renormalize it very often, so substantial drift takes place before they "correct" it. I thought that was pretty clear from my post.

[lordofduct]

Yeah, it was... crystal clear. As it was for "me" in the original.

"It is a standardized score so the top of the bell curve falls right at the 100 mark."

standardized tests implies the idea that it is reformed every so often. And if the 100 mark is standardized it can be inferred easily your exact comment. I'm just saying, you repeated a statement already made under the assumption everyone reading is a lamen or some crap like that. How rude...


wait...

I'm rude too.

[J T]

You are correct that there is drift over time, but because it is a standardized score based on a comparison of your raw score against the currently known population mean multiplied by 100, the average IQ score is still going to be 100 by definition.

The exact mapping to reality of that average is only as good as the estimation of the current population mean, but still there isn't that much drift to really throw it off.

[Limewater]

Yeah, it was... crystal clear. As it was for "me" in the original.

"It is a standardized score so the top of the bell curve falls right at the 100 mark."

standardized tests implies the idea that it is reformed every so often. And if the 100 mark is standardized it can be inferred easily your exact comment. I'm just saying, you repeated a statement already made under the assumption everyone reading is a lamen or some crap like that. How rude...


wait...

I'm rude too.

Once again you're completely missing the point.

The ENTIRE point of my post is that the IQ scores are NOT renormalized often enough. The top of the bell curve VERY RARELY falls right at the 100 mark. The ONLY TIME the top of the bell curve falls "right at the 100 mark" is once every couple of decades or so. I do not see that statement made in the thread prior to my making it.

Also, I write under the assumption that everyone is a magical pendant?

But, upon further review, it does look like they have renormalized the scores since I last paid the test any attention eight years ago, so they're probably now only off by a couple of points.

[Limewater]

You are correct that there is drift over time, but because it is a standardized score based on a comparison of your raw score against the currently known population mean multiplied by 100, the average IQ score is still going to be 100 by definition.

The problem here is that new data is being added to the population mean every time the test is given. The average IQ score is not "by definition" 100. The estimated mean at the time the test was renormalized is "by definition" 100. They aren't continuously updating the scale because that is not really practical.

The exact mapping to reality of that average is only as good as the estimation of the current population mean, but still there isn't that much drift to really throw it off.

How much is "that much?" Half of a standard deviation? A quarter of a standard deviation?

I don't know. Maybe I'm just being too nitpicky and demanding. That's what I get for spending too much time in Mathland.

[J T]

You are acting as though the mean should be the actual population mean with every score for every person added into it. That's really impractical and almost never happens in real world data analysis, which is why the means are calculated with estimation.

But anyway... how did this thread turn into a statistics lesson?

Maybe we should get back on topic.