The influence of signal variation, bias, noise and effect size on statistical significance in treatment studies of the common cold.

Many groups are working on new and improved methods of common cold treatment that include antivirals, synthetic viral receptor, compounds which block symptom pathways, and combinations of these approaches. Because the common cold syndrome is in large part subjective, symptom measurement remains an important parameter in evaluating the effectiveness of cold treatments. This review examines the features of the experimental and natural cold testing methods that effect recognition of illness signal and influence its variance and strength. Also, the importance of changes in signal variance and in the magnitude of therapeutic effect size as they relate to statistical probability were compared using a symptom data set from young adults with experimental rhinovirus colds.