BEHAVIOR OF SAMPLE MEANS AND PARAMETRIC TIME SERIES ESTIMATION.
Conference Paper
Overview
Identity
Additional Document Info
View All
Overview
abstract
The standard error of the sample mean for autocorrelated data is directly proportional to the value of the spectral density evaluated at zero frequency of the process being sampled. Thus, confidence intervals for the true mean using traditional formulas can be greatly in error. The author describes both parametric and nonparametric methods of spectral density estimation and illustrates numerically the basic results using a microcomputer program called TIMESLAB which acts as a laboratory for studying such problems. The results indicate that in many situations, adjusting for autocorrelation is easily performed.