Hi,
I have infestation of dynamics of two different groups of hosts. Total of 20 points of measurement per individual, 10 individuals per group.
I want to compare the dynamics of the 2 groups over time. I guess I had to use GEE (Generalized Estimating Equation).However, I cannot find any reference using this analyses in a similar way.
Any recommendation?
Thanks!
First plot the data. Then fit an appropriate polynomial regression (infestation (y) vs time (x) for each group). Then compare the two polynomial regrssions (for which there are methods)
You can try with panel data regression and binary independent variable to distinguish between two groups. This would also allow to include subjects' characteristics if you have any.
You may want to consider random effects panel analysis
her are some resources
http://www.bristol.ac.uk/cmm/software/mlwin/mlwin-resources.html#multipleresp
and this one is a straightfoward worked example
http://www.bristol.ac.uk/media-library/sites/cmm/migrated/documents/repeated-measures.pdf
Dear Mario:
¿What do you mean by dynamics?
¿What is your respond (dependent) variables and what is its level of measurement?
As I understood, you have made several measurements on the same study subject, being so you are violating the basic assumption of independence.
Thus, you have to use the repeated measurement procedures, but at this point in time, I can not recommend you any of these procedures without knowing your research question and the level of measurement of your respond variable.
Dear Gus,
my dependent variable is the count on bacterial cells on blood at different time points on the same individuals (repeated measures over time, longitudinal study).
Thanks
In addition to what others have written, you could use Trend analysis, Survival analysis etc etc
Thank you
The key issue here is that you have two groups, and several measures over time on the same individual. Being so, you should consider using the repeated statement on ANOVA procedure. Remember this statement handles repeated measures designs with one repeated response variable. The variables on the left-hand side of the equation in the model statement represent the repeated response variable. It does not matter if you have only two groups. You can also use ANCOVA procedure, if you are interesting in controlling some external (confounders) factors.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology