highly significantP = .008
An analysis of covariance (ANCOVA) showed that freezing during the last three trials of extinction does not predict freezing at 24-h recall, F (1,37) = 0.96, P = .34, and the effect of phenotype on 24-h recall remains highly significant even after adjusting for freezing differences during extinction, F (1,37) = 7.81, P = .008.