If you register for free, you will be able to post threads, vote on polls and lots more. If you have problems with the registration or logging in, please contact the administrator.
SizeUSA - The National Sizing Survey conducted by [TC]² (TC Squared) has been released with more body measurement data than has previously been available. It will take time for [TC]² and the SizeUSA sponsors to analyze this data and decide what actions, if any, they need to take in their product development.
The following table summarizes SizeUSA “Average Body” data. Some preliminary observations from this data are:
• People in this survey get larger as they get older.
• Black women are larger than White and Hispanic women of similar ages.
• Waists increase the most with age.
• Women's hips are larger than their bust, so they are more pear shaped.
• Men's chests are larger than their hips.
These are general observations. Significant factors not shown here, such as height and weight will enhance the understanding of body shapes and sizes.
Yes, this data is old and I haven't found current data on it - and it's not really what I was looking for but I was jsut amazed to find that for my age, I'm not the average woman -I'm smaller - average being whatever sample they took this survey from. .. and I know that numbers can be tweaked to find anything and this doesn't take into account health of the sample participants... but it was interesting
What I was really looking for was how sizing has changed over the years... because I'm not sure that such an animal as a size 0 existedin the 70s and 80s and I know people aren't getting all that much thinner so adjustments were made to create that size somehow
Last edited by maleficent; October 26th, 2007 at 05:54 AM.
Reason: men don'tt have busts
there've been countless articles over the years about how we're getting bigger as a population and the "average" woman is a size 12 or 14... so an average is gotten by takig the sum total and dividing it among the participants.. so it's possible based on the sample they took...
It seems like it IS average, in the USA. I remember reading a Glamour Magazine when I was 15. (in 1997 I think). It said the average American woman was like 5'4" and 140 lbs. I read in Glamour now that the average woman is 167lbs (no height was mentioned) with a 32.5 inch waist. It said the average woman's goal weight was 133 lbs.
I wonder what the numbers would be like if a survey was with MILLIONS of people instead of thousands?????
As long as you have a random sample, accuracy doesn't improve significantly with a larger sample size. Assuming an infinitely large population, you can get a ~3% margin of error with a sample of 1,000, and a 1% margin of error with a sample of 10,000.
Curvie Girlie, as a social scientist I deal with surveys all the time. The size of the sample, as for selection bias, are two of the main problems researchers face. Usually, a 1-2% sample is considered a reliable measure of the whole population given that appropriate procedures are followed to avoid selection bias (for an example, they must select a sample that reflects the actual make-up of the population).
So is a sample of 10000 individuals enough to draw accurate conclusions on a population of 300 million? No, no nearly enough. But comparing similar surveys with similar sample size can give us an indication of certain patterns.
So I believe that we are getting fatter, but the size and height numbers should only be considered a vague approximation and not accurate representation.
A size 41 waist for a man would be pretty overweight, no... well, depends on his height.
I'm a size 29 and I find 36 is a bit big so... then again I'm not American.
Usually, a 1-2% sample is considered a reliable measure of the whole population
If you're dealing with populations in the 50,000 range. For a population of 100 (or even 1,000), it's not nearly enough. For a population of 1 million or more, it's far too many.
Quote:
Originally Posted by siromar
So is a sample of 10000 individuals enough to draw accurate conclusions on a population of 300 million?
It gives you a 99% confidence level and less than 2% margin of error. In most circles, that would be considered pretty accurate.
(My degree is in psych; I don't have enough of a gut feel for the math to have gone straight stats, but I got 9 hours. When I went back to school, it was for accounting, and I did another 3 hours of stats for that. Neither in social sciences nor in auditing does your sample size increase proportionately to your population size.)
You are correct. Technically, a sample of 1000 should represent any population with a margin of error of 3% and confidence level at 95 or so percent. However, for a population of 300 millions with a vast number of variation, you realistically need a larger number to draw accurate conclusions. It is difficult to avoid selection bias as obesity is not constant throughout the entire United States (e.g. people of Vermont are a lot thinner than people of Georgia). Some areas will always be over-represented, and some under-represented.
Surveys (not all of statistics), like economics, are not an exact science. I was not suggesting that those surveys should be dismissed, but that they should only be taken as an approximation.
Linear Mode
