I am making an analysis to select the best model based on AIC, but one of my models gets a value of 1 in the AICcWt. I understand this basically means that this model has a power explanation of 100% and I am not sure if this is normal, since in most of the examples I have found the AICcWt values are often divided between different models.
Akaike weighting provides a way to compare various models based on their fit to the data and their complexity, assigning relative probabilities to each model to aid in the selection of the best model for a given data set. If you are testing several models you cannot have a value of 1, because this implies that the other models have 0 probability. One possibility is that the software you are using applies rounding.
Thanks, that's exactly what's happening: all the other models show a value of 0 in the AICcWt
- Badges
- Science topic
- Similar topics
- Correction