Use of repeated measures to interpret genetic and environmental correlations in animal research

A method for partitioning environmental correlations into two distinct sources of covariation--lifetime rearing effects and idiosyncratic stimulus events occurring during testing ("noise")--is presented. The method, which is based on structural equation modeling of repeated tests, is demonstrated using correlations obtained from pairs of sessions in an Open Field Test and in a Light-Dark Test. Heritabilities of most behaviors are low, but genetic correlations between- and within-test sessions are high and thus substantively influence phenotypic correlations, including test-retest reliability. Testing "noise" is usually the primary source of environmental covariance among pairs of measures, although some instances of rearing environment being the sole source of E correlations were observed. Effects of Session 1 testing and/or the additional experience between S1 and S2 test sessions produce some significant differences between S1 and S2 within-session correlations, but these are usually not large. Although varying in size, the genetic, the rearing environment and the test environment correlations between a pair of variables were always consistent in sign. The analysis demonstrates the value of incorporating some of the contemporary research and analytic strategies used in the human individual differences field into animal studies.