Statistics: sample median, means, s.d. vs sample size

AI Thread Summary
The discussion centers on the relationship between sample size and the standard deviation of sample statistics, specifically means, medians, and standard deviations. The user conducted an experiment with 10,000 data points, taking 1,000 samples of varying sizes and graphing the standard deviations of the medians, means, and standard deviations against sample size. They noted that the graphs for the standard deviations of medians and standard deviations appeared identical, leading to the hypothesis that the formula for standard deviation of sample means might also apply to medians and standard deviations. The inquiry seeks confirmation on whether the formula SD of sample statistics = population SD / sqrt(sample size) is valid for medians and standard deviations, similar to its established application for means. The discussion emphasizes the need for clarity on the applicability of this statistical relationship across different sample statistics.
jhson114
Messages
82
Reaction score
0
i have a set containing 10000 data. i took 1000 samples of size 4, 16, 64, and 1024 and took the medians, means, and stadard deviations of each size. i graphed them sd of medians vs sample size, sd of mean vs sample size, and sd of s.d. vs sample size. for sample mean, i know from a textbook that:
SD of sample means = population SD / sqrt(sample size n).
But it seems from the graph i created using R, sd of medians and s.d. vs sample size all have the exact same looking graph, which to me suggests that:
SD of sample means, medians, and s.d. = population SD / sqrt(sample size n).
Is this right? it seems a little awkward. any input will be very helpful. thank you
 
Physics news on Phys.org
graph i created using R
What is R?

More generally, is your question whether the "sample vs. population" formula (sample stat = population stat/sqrt(n)) applies to median and std. dev. in addition to the mean?
 
R is just a program language kind of like matlab.

"More generally, is your question whether the "sample vs. population" formula (sample stat = population stat/sqrt(n)) applies to median and std. dev. in addition to the mean?"

This is exact what I'm asking.
 
Back
Top