It is common to attempt to estimate how variable the arithmetic mean, variance, and standard deviation of a set of data are.
A simple, but effective, method is called jackknife resampling (refer to http://en.wikipedia.org/wiki/Jackknife_resampling). The idea behind jackknife resampling is to create datasets from the original data by leaving out one value each time. In essence, we are attempting to estimate what will occur if at least one of the values is incorrect. For every new dataset, we recalculate the statistical estimator we are interested in. This helps us understand how the estimator varies.
We will apply jackknife resampling to random data. We will skip every array element once by setting it to NaN (Not a Number). The nanmean()
, nanvar()
, and nanstd()
can then be used to compute the arithmetic mean, variance, and standard deviation:
First initialize a 30 x 3 array for the estimates, as follows:
estimates...