Exercise: Comparing means, 4

Question 4. What method would be used to calculate the confidence interval, and why?

Suggested answer

What condition should the data meet for this method? This is a comparison of two samples, so we would use an unpaired or two sample t method. The data should be from Normal distributions with the same variance for the standard two sample t method. However, in this study the standard deviations in the two samples are very different, 48.7 Ámol/L in study patients compared with 14.3 Ámol/L in controls. We should therefore use the two sample t method for unequal variances, the Satterthwaite correction to the degrees of freedom.

Back to question.

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

Back to top.