[R-sig-ME] intra-class correlation coeff

Andrew J Tyre atyre2 at unlnotes.unl.edu
Fri Jan 23 15:27:50 CET 2009


The model you are estimating assumes that the differences among markets 
(on a log scale) are multivariate normal with mean 0 and variance sigma^2 
= 0.07 - and zero correlation among markets. Another way of interpreting 
the magnitude of that value is to compare the standard deviation of the 
market random effect with the size of the intercept - so you have a CV 
there of 0.27 / |-1.37| = 20% not an insignificant amount of variation 
among markets. This model allows the mean responses to be similar WITHIN 
markets - accounting for the correlations between traders at the same 


Drew Tyre

School of Natural Resources
University of Nebraska-Lincoln
416 Hardin Hall, East Campus
3310 Holdrege Street
Lincoln, NE 68583-0974

phone: +1 402 472 4054 
fax: +1 402 472 2946
email: atyre2 at unl.edu

More information about the R-sig-mixed-models mailing list