Abstract
Using open-field test data of male rats, a correlation provided an estimate of the magnitude of dependency among litter mates. A simulation experiment was performed to estimate the magnitudes of Type I error rate distortions when individual scores instead of litter mean scores were employed as the unit of statistical analysis. The results showed a rather large correlation among litter mates in defecation, latency of defecation, and latency of urination, suggesting that it is better to avoid treating the individual animal as the unit of statistical analysis. However, the magnitude of correlations in the remaining measures was negligible. As to Type I error rate, except for defecation, latency of defecation, and latency of urination, the obtained values obtained were not seriously affected. A very seriously inflated distortion in the error rate was found for body weight.
Get full access to this article
View all access options for this article.