Objective measurements, such as the level of functioning of a patient, can be analysed and give reproducible data and are therefore superior in
terms of absolute analysis.
The double-blind randomised controlled trial is the gold standard for obtaining data for statistical analysis. A ‘P’ value of below 0.05 is taken to be the threshold for indicating significant validity for a finding. However, there is, by definition, a 1:20 probability that this finding can happen by chance even though there is no difference between the groups in reality. Furthermore, 0.05 is only valid if
two groups are being compared. If larger numbers of groups are included in the analysis the P value must be further reduced (e.g. to 0.016 for three groups.
This excerpt is reproduced with the permission of the publisher.