Hi there,
I probably know the correct answer, however, I would appreciate if you guys share with me your opinion whether is accurate to do any statistic when less than 3 donors are available per group?
Just to notice that more HE slides are analyzed per patient, thus indeed I have more than 3 values in a group, however, the source of these values are less than 3 patient!
No more patients can be analyzed, therefore the problem with number of patients cannot be solved!
Thank you in advance for taking time reading this question! Any suggestions are welcome!
Best
Mariya
Dear Mariya
on instinct, I would go with: No. Assuming you have 2-3 people per group, even if you run an ANOVA with Gabriel or Games-Howell post-hoc Test and the sample-sized adjusted Cohen's to determine the effect size of the different means, you will still only have 2-3 people per group - not to mention all requirements for ANOVA would possible be violated, e.g. normal distribution, homogenity of variances, significant overall ANOVA.
Note though that I am far from being a statistician, so a more capaple person might shed some more light onto your question.
All the best
Marcel
Thank you Marcel for your answer! I knew it, however, i wanted to get more opinions in that matter! The problem is that for rare medical conditions the patient’s number is limited! Thanks again for your reply
In my opinion you should treat the 3 cases as 3 ideographic single cases in a within design and avoid nomothetic conclusions.
- Badges
- Science topic
- Similar topics
- Correction