I'm using a multinomial logistic regression analysis to examine deer behavioural responses to camera traps in terms of 7 predictors (both singly and their interactions).
I have found that the model with the lowest AIC value is season and vegetation (AIC = 1005.023). However, the AIC value for season as predictor variable on its own is 1005.103. From what I have read I gather that this means that these 2 models are nominally equivalent in indicating the processes influencing reactions as the AIC value difference is less than 2 or 3. I'm wondering if it is okay to report both AIC models and say that the model season and vegetation indicates that these variables when combined have a strong influence on the underlying processes which influence behavioural responses to camera traps even though the AIC values are nearly the same? The AIC value for vegetation as an isolated predictor variable is 1008. 289.
Also, when I run various models based on my hypothses, the best models seem to centre around the 1008 AIC value mark. As the difference in AIC value between the model season and vegetation and these models is approximately 3, does this means that the models with an AIC value of ~1008 are equally as good at indicating the underlying processes influencing behavioural responses as the model season and vegetation?
Thanks a mil!
