Repeated measurements, bilateral observations and pseudoreplicates, why does it matter?

A common requirement of statistical methods, critical to the interpretation of the data, is that the analyzed observations are independent. This is not always the case in experiments and clinical studies, a mistake which can be expected to lead to erroneous study results. The phenomenon is explained, its consequences described, and suggestions to avoid the problems presented. (C) 2012 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.