An example of the importance of experimental design and statistical analysis in dental research.

In this paper a specific dental investigation was used to demonstrate the value of several statistical principles. It was concluded that: Blind evaluation of tissues is important in preventing bias in interpreting the histomorphologic results. The loss of replication makes it impossible to evaluate the effect of sacrifice time on radioautographic labeling. A probability value P = .12 can be of interest if one understands the data to which it is applied (i.e. P = .01 or P = .05 is not sacred). Investigator inconsistency exists and should he considered. In this case evaluating the tissues in a random order averages out the effect of counting fewer labeled cells as the evaluation progressed. Random allocation of animals to control and experimental groups eliminates the possibility of bias in allocation. The smaller size of the experimental animals is not, therefore, a function of allocation.