Investigators are designing an observational study comparing fracture risk in chronic glucocorticoid users and nonusers. They are particularly concerned that age and sex differ substantially between groups and may distort the results. Which of the following study methods is most appropriate to reduce the effect of these variables before analysis?
- A)Performing only univariate analysis
- B)Using a larger alpha level
- C)Increasing the p value threshold
- D)Matching participants on age and sexGABARITO
- E)Lowering the confidence interval width
Explicação
Matching is a design stage method for reducing confounding by ensuring that exposed and unexposed groups have similar distributions of important variables such as age and sex. It is especially useful in observational studies when certain confounders are known ... Ver explicação completa e trilha adaptativa →