Evaluating the medical literature part II: Statistical analysis

摘要: Summary We have attempted to provide sufficient information enable the reader verify that investigator has used an appropriate statistical test for evaluation of his study data. not discussed actual calculation tests presented. With wide availability computers and programmable calculators, it is safe assume performed necessary mathematics accurately. Instead assessing accuracy these calculations, should correct was chosen in first place. The astute may be surprised at how frequently incorrect used. Merely achieving significance does characterize author's data as clinically important. Neither it, itself, prove one agent superior another nor a cause-and-effect relationship between two variables. When appropriately interpreted, however, analysis can very useful powerful tool helping arrive "truth."