I have a problem regarding pooling data from replicate experiments.
I have done an experiment where I collected multiple tissue samples from 5 mice, then the samples were allocated to treatments and analysed after 5 days. I repeated this experiment 4 times.
The data was pooled and analysed by General Linear Model Univariate with the treatment as a factor and replicate as a random factor.
Now my paper has gone out for review, and one reviewer has commented: "Pooling, while valid, with replicates as you have done, is valid if the replicates were not different from each other - where do you address differences between replicates in your data?"
My questions are:
Is the reviewer correct?
Is there anything wrong with my analysis?
Should I report the effect of the replicate on the model (eg P value), or the interaction with between the replicate (random factor) and the treatment (factor)?
There is a similar previous question here, but the answer didn't address the questions I have: Pooling data from replicate experiments
Also this question, but I didn't understand the answer! Independent replication experiments yielding contrasting results; how to combine them?
