Exercise: Comparing means, 3

Question 3. What test of significance would be better in this study?

Suggested answer

We should compare the two groups directly, using two sample t test for the creatinine at six months follow-up. Even better would be to use the three month data as well, perhaps by averaging the two. We should also control for the pre-treatment levels. We could compare the mean difference between pre- and post-treatment creatinine between the two groups. An even better way to do this, which we will look at later in the course, would be to use analysis of covariance of the post-treatment creatinine using the pre-treatment creatinine as a covariate.

Back to question.

This page maintained by Martin Bland.
Last updated: 27 July, 2009.

Back to top.