Exercise: Compare the AIC Scores
Compare the AIC scores of the models with and without the random effects.
We'll cover the following
Let’s test what we’ve learned about AIC scores.
Assignment
Compare the AIC scores of the models with and without the random effects. Note that this isn’t a comparison of versions of the models from the More Linear Models in Rchapter (which used tank means) and our new model. There’s no way to compare those because they have different response variables—one being each individual tadpole and the other being the tank means of those individuals. Instead, we’re making a model like that in the previous lesson titled More About Model Selection, where we include all the data but without the random effects.
There are a couple things we should note:
- We’ll need to check how both the tail length at metamorphosis,
Tail.initial
, and the resources the tadpoles had access to during the larval period affected the rate at which they resorbed the tails,Resorb.days
, into their bodies. - We may need to log-transform the data and use the
filter()
function to remove any troublesome data.
Get hands-on with 1200+ tech skills courses.