Normally Google can answer all of one's research questions, but this one is exceeding my skills. Does anyone have any idea how many people in the world are, say, seven feet tall or more? Or, relatedly, know the standard deviation of human male height?
« Simple Answers to Silly Questions | Main | Early Exit Polls »
Heights Bleg
07 Nov 2006 03:12 pm
Comments (33)
I suspect Matt wants to know what percentage of 7-foot tall men aged 20-35 are already playing NBA basketball.
This Google query is as close as I could get; some articles if you are willing to pay for them. But none of these cover extreme edges of the spectrum, so it's hard to estimate.
Here's another possibility from Google Scholar. (Hope that links correctly; if not, search terms are human adult male stature variance.) The first couple of articles look promising, though I didn't dive into the details.
Whatever the number is add 1 for Bill Walton. Always claimed to be 6' 11" because he thought 7 footers had a bias against them. Yes, this is odd, but so id Bill.
Check the wikipedia entry for Homo Sapiens?
Note all the links in the wikipedia article. This one for example:
http://www.cdc.gov/nchs/pressroom/04news/americans.htm
man, didn't we already get into this recently?
Like at Crooked Timber or somewhere?
Somehow it's all mixed up with the Lancet report in my mind--you know, a lot of ill-informed bloggers speculating about statistics.
which reminds me that one of the best slams on Coulter I've read is that she looks like Bill Walton in a blond wig.
Kind of harsh on Bill, but harsher on Coulter, so okay.
Didn't Steve Sailer ask the same question a month ago? This is a re-bleg.
I can't give you a reference, but generally, human height within a single country, for single sex, has a standard deviation of 3-3.5 inches.
For the answer to any question anyone might have in the future I recommend posting on metafilter.
http://ask.metafilter.com/
They'll never steer you wrong.
This might be a good question for Andrew Gelman, professor of statistics at Columbia. He recently had an interesting post that largely concerned the distribution of height in human populations:
http://www.stat.columbia.edu/~cook/movabletype/archives/2006/10/galton_was_a_he.html
You know what they say about tall men? They are good at seeing over walls.
The Dutch are the tallest people in the world.
If American male height is normally distributed with a mean of 69.4 inches and a variance of 2.4 inches, as this website (http://www.kuro5hin.org/tag/height) suggests, then only 1 in 1.7 billion people is over 7 feet tall. Thus, we can be fairly confident that there are no males in the U.S. over 7 feet tall.
NHANES does periodic surveys of vital stats; here's a report broken down by ethnicity and year of survey:
http://www.cdc.gov/nchs/data/ad/ad347.pdf
Note that human height (like several other characteristics, apparently including IQ) is not quite normally distributed; it is a fat-tailed distribution (more extreme cases than a normal).
If the mean is 5'10" and one SD above adds 3.2 inches, my spreadsheet shows about 900 men at or above 7' for a normal distribution and about 3400 if the distribution is log-normal (my favorite distribution, 'cause height can't be negative, for one thing!) These numbers feel about right to me. For a SD of 2.4, the numbers are very small -- 1/2 a man, or 5 men for normal and log-normal respectively. Those feel wrong. But I think you need a lot more data on the true distributions: clearly, dwarfs don't lie on the same curve, and I doubt Masai do either.
Note that human height (like several other characteristics, apparently including IQ) is not quite normally distributed; it is a fat-tailed distribution (more extreme cases than a normal).
So it seems . . . my methods, apparently, are unsound.
Matt asked a very similar question on 10/11/06.
Why are you excluding women in the second part of your question? I know some pretty tall women. Don't know about seven feet, but maybe.
Here's something I didn't know until recently: Andre the Giant wasn't really 7-4. He was more like 6-10, but the WWF chose that height because they wanted him to market him as taller than Kareem Abdul-Jabbar. So between Andre and Bill Walton, we've got a wash.
Matt asked a very similar question on 10/11/06.
Right. I knew there was something in the basketblogging archives about this.
Another way to think about the problem of what are the odds a 7 footer will play college basketball or make it to the NBA is to think about how many 7 footers you went to college with. Were they all on the basketball team? Did they play on the NBA?
When I was at Rice, there were two 6-11 guys. They both played for the college team, neither made the NBA. (The second stringer outraged the coaches because he was secretly at Rice to get an engineering degree and put more effort into his studies than into playing.)
When I was UCLA, there were two seven footers out of 35,000 students. Both played for the Bruins. Both went to the NBA. The Bruin's second stringer, Mark Eaton, became something of a legend as a pro.
Why don't you ask the real question directly? Extrapolation from the statistics for the height is prone to errors.
How many men have it >= 11 inches?
SELECT * From Population WHERE Height > 84
> 10000 Rows returned. Buffer not big enough.
This blog has a great discussion of this very question (only a few days ago)...
http://www.stat.columbia.edu/~cook/movabletype/archives/2006/10/galton_was_a_he.html
The Mark Eaton story is real weird:
http://en.wikipedia.org/wiki/Mark_Eaton_(basketball)
Apparently, he started in his rookie year at utah after being on the bench the year before at ucla.
Here's something that may help -- The Association for Professional Basketball Research, http://hometown.aol.com/bradleyrd/apbr.html
has a database that you can download (click on "STATISTICAL DATABASE" at the top of the page). It includes heights of NBA players beginning with the '37-38 season through current, and has a sort function so that you can sort the players by height. Although you won't get your data based on world population, you can get accurate data based on professional basketball players.
Although you won't get your data based on world population, you can get accurate data based on professional basketball players.
And if I do the statistics correctly, I'll get a result just as valid as the Lancet study! After all, if the statistics are done correctly, how could it be wrong?
And if I do the statistics correctly, I'll get a result just as valid as the Lancet study! After all, if the statistics are done correctly, how could it be wrong?
Uh, right. And if you do the addition of 2 and 2 correctly, you'll get 4. You do realize that statistics is math, don't you?
Comments closed November 21, 2006.
| 
Heads in the Sand
What research question is this related to?
Osama is only 6ft 4'.
Posted by otto | November 7, 2006 3:17 PM