Type II error and statistical power in reports of small animal clinical trials.