I am going to study about cell population data generated by automated hematology analyzers related to Neutrophils in sepsis.Study will compare medians of several parameters between normal control group and sepsis patent group.So what would be the sample size determination formula? study is case unmached type and conventional formula needs "Percent of controls exposed"
Using the zoom factor and the radius of the lens, you can count the number of cells in a given region and then either multiply to cover the volume, or do a few more counts and average them. Just report on which technique you used in the project, so you can come back to it with the same data and have someone else suggest a technique and calculate the new population.
As far as I know there is no clear cut answer. Laith G. Shareef has given an answer for the situation where the outcome of interest is the frequency of occurrence of something. There are formulas for calculating the sample size for a continuous interval scale data, where you typically do an unpaired t-test to test the difference between means.
When you are comparing medians, there is no definite answer. In this case you typically do a Mann-Whitney U test. You need to do a sample size calculation assuming you are testing a difference in means, figure out the "Asymptotic Relative Efficiency" (ARE) of the t-test compared to the Mann-Whitney and multiply the sample size you obtained earlier to get a typically larger sample size.
The problem is, the ARE is usually known for cases where the Null Hypothesis is true. If there is a real difference between the groups, ARE is not well-defined. You have to figure out a way, most probably with the help of a statistician, who is at least somewhat familiar with your domain.
Sorry I haven't been a better help.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size