Effect of sample size on reproducibility of behavioral teratological study results: a computer simulation experiment using data from the Collaborative Behavioral Teratology Study of the National Center for Toxicological Research.

A computer simulation experiment which attempted to examine the effect of sample size on reproducibility of the effect of treatment was performed on the basis of actual data obtained from the Collaborative Behavioral Teratology Study of the National Center for Toxicological Research. The degree of the treatment effect was assessed in terms of the strength of the association (eta square). The results indicate that sample size has a large effect on the reproducibility of results which are assessed with the magnitude of SD for eta squares obtained from replication experiments. Suitable sample sizes to obtain relatively consistent results across studies were discussed, pointing out that not enough attention has been paid to the effect of sample size in the issue of reproducibility of results in some behavioral teratology studies.