I collected activity budgets over a course of a whole breeding season for individually followed paired mother and pup grey seals. I want to look at how their behaviours are impacted by different variables.
I have 5 different behaviours for mothers and 3 behaviours for the pups. different variables (location (2 sites) tide state (4 states) lactation stage (3 states) breeding season (3 states) and time (3 states).
I have already done basic analysis looking at how individual variables effect behaviours but now want to try and put it into a model and look it as a whole. The data does not conform to normality but want to know whether there is a model I can use or whether I should transform it or rank it in some way.
Thank you very much, Bob
-Can attach or forward data if that helps.
Dear Robert,
this sounds like a non-linear multivariate classification approach (e.g. Random Forst) is the appropiate choice, since it can handle categorial and continuous variables (available e.g. in R). In case you have further question do not hestitate to contact me directly.
Best,
Ron
You might consider permutational MANOVA (also called nonparametric MANOVA) using adonis in R. See this page "Permutational Multivariate Analysis of Variance Using Distance Matrices" (http://cc.oulu.fi/~jarioksa/softhelp/vegan/html/adonis.html). These types of analyses are often used for data sets in ecological studies. I've never done this analysis, so sorry I can't be more help. Marti Anderson (https://www.researchgate.net/profile/M_Anderson2) does a lot of this type of work.
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology