In my current research the sample size was calculated to be 500 people. I have gathered the data now I want to cosider my exclusion criteria and exclude 80 of them. therefore only 420 samples would be analysed. From a statistical point of view, Is there any problem or I shouldn't have had counted them among the calculated sample size?
In the context of blinded comparative clinical trials, ICH-E9 specifies that the rules need to be specified in the protocol. In some cases, changes to these rules may be presented in a statistical analysis plan, typically prior to the collection of any data, but a blind data review may legitimately result in other revisions. If you have a single-arm study or if the blind has been broken, there should be no changes, but a supportive analysis can be performed if the need arises.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size