ABSTRACT
Many competent researchers lack sufficient statistical training to determine the appropriate sample size, the validity of statistical principles employed in an experimental design, and the statistical methods for data analysis. This is particularly true when sophisticated analysis is required. Consultation with a trained statistician is strongly advised. A statistician should be consulted, however, to verify the accuracy of the estimation before the experiment is started or before a manuscript is submitted for publication. In some cases bilateral models are not appropriate. When major procedures are performed, especially those involving a joint, they create some degree of disability. If using a bilateral model, this may cause ethically unacceptable disability to the experimental animal. Examples of commonly used descriptive statistical terms include: measures of central tendency (mean, median and mode) and measures of variability (minimum, maximum, range, standard deviation, standard error of the mean, and coefficient of variation).
