Is it possible to include repeated measures (permanent environment effect) inside the age(time) classes when utilizing random regression models?
The idea is trying to maintain the "time classes" interval length, so I won`t lose too much information in "not so big" data files.
Anyone have any experience with that?
Does that even have any support on random regression models theory?
Thanks in advance.
Hi Bruno,
I am not clear exactly on what you're asking. I assume that when you write "random regression models" you are referring to "random effects" regression model? I am also not clear on what you mean by "permanent environment effect"? Are these your X variables or your Y variable? Can I assume that age is an X variable that represents the time periods in which you are taking measurements?
The more detailed information you provide, the greater likelihood that you will get a helpful response.
Ariel
Hi Ariel,
Sorry, but I really meant "random regression models", not just random effects. These models asume that, if individuals can be treated as random effects, the regression coeficients associated to each individual can also be treated as random effects.
If you associate that with a matrix of genetic relationship(expectations) between individuals in the analysis, we get plenty of usage for that modelling strategy in animal breeding and genetics, specially when studying variables that can be measured in different ages/times on the animal.
That will result in a different approach to repeated measure models, in wich we can generate a curve(or a line) wich tries to explain the genetic variance behavior through the course of a "time oriented" variable studied.
To answer this question, I probably need someone with vast experience in random regression applied to animal breeding and genetics.
Fair enough. I am interested in hearing how others with experience in this area respond.
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Genetic Analysis