I have inputted my occupancy data which generated an AIC of 764.46. I then inputted my site co-variate data and ran the model, but it did not change at all. I suspect that it is not reading my site co-variate data and that I am maybe inputting the data incorrectly. Could someone please help?
Many Thanks
Hi Freya, which model and program are you using? The AIC does not have to vary but it's unlikely that nothing changes at all. What about other model fitting measures ? (-2LL, BIC, etc..).
Hi Nuno,
thanks for taking the time to look over my query. I am using PRSENSE and I have tried a custom model looking at the psi and also a constant p model. The AIC dod not change at all which is what puzzled me, as even a slight variation is s result. It was as if the program was not identifying my covariates at all. I have not looked at those models as my supervisor recormended the presence program, but I'll have a look.
Hi Freya.
Did the number of estimated parameters change? If it changed, it seems that the model converges, and thus that Presence is 'reading' your covariate. If it did not change, it means that your model did not converge because you either made something wrong or you have few data to run such model.
Thank you so much for all your replies. The number of the covariate has finally changed I used a custom model which seems to detect my covariates of fire age. I was wondering if there is any easy way of getting presence to read a word value such as 'woodland' 'heathland' 'shrubland' or will it be best to convert them into numbers in order to input them into the model as a covariates?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling