|
Post by rs on Mar 20, 2014 23:39:58 GMT
Hi, what would be the right interpretation of the standard error of a variance component? I saw the thread about how the AI matrix diagonal is used to get SE, but what is the layman interpretation?
Thanks
|
|
|
Post by Jian Yang on Mar 21, 2014 7:10:46 GMT
Any parameter is estimated with error. In classical statistics, the point estimate of a parameter is regarded as a random variable following a distribution, e.g. normal distribution if the sample size is large. SE is just the standard deviation of this distribution, measuring the accuracy of the an estimate.
|
|