Please help me! I am stuck choosing which model to use! :(


New Member
Hello! I have been reading so much about stats that I am a bit confused now.. I have samples from an experiment in which I have counted cells. So I took 2 pictures of each kidney. I did this for two kidneys (per individual). And then each individual belongs to a group (infection). My hypothesis is that I will have different cell numbers per group. My problem in that I can't do one-way ANOVA, because the values I have are not independent. Doing the average per individual would not be correct. Then I though about doing nested, but in the end, the picture/kidney are not really important (I don't want to know if choosing one or the other will have any effects)... I just took the different pictures and chose both kidneys to low the error that I could be adding (for example by choosing a specific kidney or area of the kidney) Can someone please help me?

This is how my data set looks like (roughly)

Individual Photo kidney treatment N. cells

1 1 1 Control 11
1 2 1 Control 9
1 1 2 Control 21
1 2 2 Control 25
2 1 1 Infected 37
2 2 1 Infected 33
2 1 2 Infected 25
2 2 2 Infected 35


TS Contributor
this looks like a clear case of a nested model to me. You would need to use a nested model, not because kidney or photo is of interest but simply because a nested model describes your data best.