jamovi

Good evening everyone. My name's Paulo and I'm a Master Student here in Brazil. I'm looking for a question regarding small sample analysis, and how to minimize background variations between experiments. I use Flow Cytometry to collect my data, and because of the high variability found in most Cytometers, analysis between experiments from different days is often considered non-viable.



for example, here is an experiment i've made on three different days of the week. even though isolated analysis show no crossing bars between groups (therefore showing some significance), the effect size observed on each experiment is different, and when I analyze the whole data I end up losing the effect. How can I get the best approach to this situation?

Some extra information for context:
- The experiment in question uses "technical replicates" and each bar on the graph shows all data from that day's triplicate.
- Flow cytometry uses High-troughput analysis and measures mean population fluorescence. Each observation is merely the median from the distribution of all cells in that cell plate.
- Given the criteria from homogeneity and normality met, the experiment in question would use an ANOVA.
- I am well aware that non-paramatric analysis can be a solution. However, I would like to check for every possibility that doesn't rank or normalize the data.

Thanks in advance!

hmm, good Q. let's see if anyone else has any thoughts.

jonathon