I am trying to analyse confocal images of coronal brain sections to compare non-injected (contra) vs injected (Ipsi) hemispheres of control (pbs) animals and treated (agent) animals.
The resulting comparisons are:
PBS: contra - Ipsi
Agent: contra- Ipsi
PBS contra - Agent contra
PBS Ipsi - Agent Ipsi
I just can’t shake the feeling I’m missing something by not factoring in matched pairs for PBS: contra - Ipsi and Agent: contra- Ipsi. This was intended as an internal control for each animal, but each hemisphere could be seen as a distinct treatment (I.e no intervention). I am going in circles with two-way ANOVA, with or without repeated measures, and/or mixed effects models.
In practice I am only sampling and imaging once, but am comparing two sides of the same brain within each treatment group.
Any advice/guidance would be very helpful. Thanks.
It seems as though its repeated measures (which is essentially equivalent to mixed effect models in a balanced design with two levels on the repeated measures factor). The way this might not be is if you are deriving a difference between the two sides and hence have one measurement (the difference) per individual.
- Badges
- Science topic