Which statistical method is used to compare two independent groups while adjusting for a covariate?

• A. ANCOVA
• B. Two-way ANOVA
• C. Paired t-test
• D. Chi-square test

Answer: (A)

When you want to compare two independent groups while accounting for a continuous covariate, the method used is analysis of covariance (ANCOVA). ANCOVA adds the covariate into the comparison as a predictor in the model, so you can separate the effect of the group from the effect of the covariate. This yields an adjusted comparison of the group means at a common value of the covariate, reducing the influence of the covariate on the outcome.

In practice, you fit a linear model with the group indicator and the covariate. The group effect tests whether the outcome differs between groups after adjusting for the covariate. This approach also provides adjusted means and can improve precision by decreasing residual variance when the covariate is related to the outcome.

Key assumptions include a linear relationship between the covariate and the outcome within each group (with roughly equal slopes across groups), normally distributed residuals, and homogeneity of variances.

For context, imagine comparing systolic blood pressure between a treatment and control group while adjusting for age, since age can influence blood pressure. ANCOVA would estimate the treatment difference in blood pressure after holding age constant.

Other options don’t fit this scenario as well: a two-way ANOVA handles two categorical factors and their interaction but not a continuous covariate, a paired t-test is for related or matched samples, and a chi-square test is for categorical data without adjusting for covariates.