I'm sorry if this may seem a simple question but the analysis I have been doing up till now has been the non-parametric paired t-test (Wilcoxon) for primary chondrocytes treated with or without IL1a. To illustrate the variation in gene expression between patients I have expressed the data graphically as line graphs for n=15 patients. However, for the stats analysis of the same data I have arranged the data in columns for control and treated ensuring each row corresponds to the same donor and done the above analysis. Is this correct?
If donors are expected to constitute a potential source of variation, it may be reasonable to block the experiment by donors. This will make paired comparison of gene expression by donors reasonable. If not, the testing should be unpaired. You are the medical expert and you should know whether donors are a source of variability or not.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs