# Thread: Standard Deviation from Percentage groups

1. ## Standard Deviation from Percentage groups

I have data that I can use to indicate what percentage of the US population falls with different groups. IE

under 100 lbs .2%
100-110 lbs. 4%

The whole will equal 100%. But I want to calculate the standard deviation of the data, but have no idea where to start. Any help would be greatly appreciated. Here is the data I am working with:

http://www.census.gov/compendia/stat...es/12s0210.pdf

2. ## Re: Standard Deviation from Percentage groups

Hey badams.

To calculate the standard deviation (or sample error), you need to have either a population distribution or a sample of observations from that distribution.

Calculate the sample standard error of a sample is basically 1/(n-1) * sum(x_i - [x])^2 where [x] is the mean of the sample and x_i is the ith observation summed over all observations (basically sum of squares divided by n - 1).

This can be used to estimate the standard deviation of a sample, but you still need to decide whether you have an assumed population or whether you don't and these are two very distinct things using very different kinds of statistics (but the same core principles do apply).