I am working on a Parkinsons disease model in mice. The parkinsonian behaviour is scored before treatment starts. Each mouse has a specific score, which is based on the extend of the lesion (more lesion = higher score). There are several time points after the treatment has started in order to see if a treatment wears off over time. The pre score for each individual mouse is important as the treatment effect needs to be compared to it.
I am not sure what analysis to use for it.
Anova - I can only analyse the post treatment data
Ancova - I can only compare 2 time points.
What do I do?
As Daniela said, a repeated measures ANOVA is possible, but whether it is valid depends on what the distribution of the mouse scores are. What is the range of possible scores values, and what does a histogram of those values look like across your study population? Depending on the properties of the score and the sample size, likely the best approach is a longitudinal model (either a GEE or a GLMM).
Anything that takes into account the repeated measures would be fine. Repeated measures ANOVA, linear mixed effects, etc. Using mixed effects allows you to treat timepoint as an ordinal variable, which I'm not sure if you can do in ANOVA.
- Badges
- Science topic
- Similar topics
- Medicine
- Disease
- Disease Models