Take the 2-minute tour ×
Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It's 100% free, no registration required.

How should I calculate sample size for a clinical trial when I need a power of 80% and I predict an effect size to be at lest 0.5? do I need to estimate the mean of outcome as well?

share|improve this question
1  
You have to provide much more details. What does your effect size suppose to detect? What test will you conduct in order to detect it? What will be the type I error rate? You will also need to provide some measure of variation. Anyway, I am afraid your question cannot be answered as it stands. Also, note that one can find online sample size calculators for most common tests... –  ocram Mar 20 '12 at 18:53
    
I will measure the severity of pain with an analogue scale. there is a new intervention and a standard intervention. the type of this new intervention has made a noticeable change in previous studies, but with different measures. I will use repeated measures. alpha will be 0.05. I am confused, whether these are enough to estimate a sample size. –  Sara Mar 20 '12 at 19:20
    
Also relevant is whether the VAS scores will be normally distributed (unlikely) or skewed (likely). –  pmgjones Mar 20 '12 at 19:39
add comment

2 Answers

There is a free program called Gpower that is specifically designed to help you answer these kinds of questions. One of its features is that it computes "sample sizes for given effect sizes, alpha levels, and power values (a priori power analyses)".

share|improve this answer
add comment

If you were looking at the difference of two means you have enough but not for a linear model. The SAS proc glmpower will solve the problem for you when you have a general linear model. I think it can be used for mixed models as well as fixed effects models. I don't know whether the free software will do this. If you don't have SAS nQuery Advisor and Power and Precision are inexpensive sample size/power determination tools that you can use.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.