Animal models for osteoarthritis--ensuring experimental validity

Confounding factors, bias factors and hidden variables affect the design of experiments involving animal models. In a frequently used dog model for osteoarthritis these can arise, for example, because of the influence of age, sex and breed of dog. Controls are required to investigate the progression of osteoarthritis even in experiments forming time series. However, there are no true controls in animal models for joint degeneration. Analysis of data collected from experiments involving animal models depends on the number of factors being varied. The number of animals influences the level of confidence associated with a result. Ignoring statistics can invalidate conclusions drawn from the experiments. Statistical considerations are also important in the presentation of results.