ChristianKl comments on 2014 Less Wrong Census/Survey - Call For Critiques/Questions - Less Wrong

18 Post author: Yvain 11 October 2014 06:39AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (269)

You are viewing a single comment's thread. Show more comments above.

Comment author: Vaniver 12 October 2014 08:32:03PM *  1 point [-]

Before the professional intelligence measurement question, please add something like:

Please ESTIMATE your IQ, using a standard average of 100 and stdev of 15.

It might make sense to add on something like:

You may want to check a normal distribution calculator (such as http://stattrek.com/online-calculator/normal.aspx ) to ensure you aren't overestimating.

Or, alternatively, we could get at the same sort of question by asking this:

Please estimate how many people would have to be in a group of adults, chosen uniformly at random, for you to expect the group to contain one (and only one) person smarter than you.

But it's not obvious to me that this is a thing people would be good at estimating.

Comment author: ChristianKl 13 October 2014 10:58:30PM *  4 points [-]

Please estimate how many people would have to be in a group of adults, chosen uniformly at random, for you to expect the group to contain one (and only one) person smarter than you.

That depends a lot on the group of adults you use as reference. World population? The population of the country in which you are living? City? Your facebook friends?

I think IQ is much better because there are objective standards. "Smart" is also a word that different people interpret slightly differently. IQ is much more precise.

Comment author: Vaniver 13 October 2014 11:48:34PM 0 points [-]

That depends a lot on the group of adults you use as reference.

Ideally, ethnic Brits, since I believe that's what's typically done for IQ reference distributions.

I think IQ is much better because there are objective standards.

Yes, but people may be better at imagining groups than normal distributions. Just how much unlikelier is 150 than 145?

Comment author: ChristianKl 14 October 2014 01:56:04PM 0 points [-]

Yes, but people may be better at imagining groups than normal distributions.

That would also be interesting to test. Ask both for the IQ and ask for the relative intelligence.

Do well calibrated people perform better on this task?