Hello Respected seniors!
İ am doing a research on COVİD-19. İ want to determine the effect of Physiotherapy in COVİD-19 patents. İ made two groups.İn group one we included patients with ventilatory support and other group contains patients without ventilatory support. Both group underwent same treatment intervention. 6MWT is my primary outcome measure. A two way repeated measures ANOVA will be test of choice to determine effect within group(time) and between group.
How can İ calculate sample size for each group?
İs my test of choice is good enough for my hypothesis?
Thanks in advance
You could use the free program G*Power. Since you have a within-between design and you seem to be interested in the interaction you should choose this option. But you need an estimate for the intercorrelation of the repeated measures and you need to have an idea what how large the minimal/smallest effect size of interest is, which is relevant for you to detect. Of course alpha and beta error you are willing to accept need to be specified.
For sample size estimation, you may just use t test to compare the change from baseline between two groups, but you will gain additional efficiency if you use mixed model repeated measures analysis later. One thing you need to consider is the difference in patients' characteristics between to comparison groups. Patients who are on vent vs those who not are very different. The difference you see at end of the study may not be the result of vent yes or no.
The G*Power software is suitable program to estimate sample size. Please refer to the following link:
https://www.researchgate.net/publication/339553439_Quantifying_lower_limb_inter-joint_coordination_and_coordination_variability_after_four-month_wearing_arch_support_foot_orthoses_in_children_with_flexible_flat_feet
Alternatively to G*Power, you could simulate your study in R for example and assess under which circumstances desired power can be achieved. This can be more realistic, since G*Power simply assumes that all assumptions are met.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size