how to study the influence of personality (repeated measures) on a trait

#1
Hi all,

I am trying to investigate how frogs level of aggressiveness influence their plastic responses to a more or less big opponent.

To do so, I presented an individual with a simulated opponent, representing the average type of the population, and I measured how quickly the focal individual reacted to the opponent, jumped towards it and reached it. I know from previous results that these variables are repeatable and that a latent variable ("aggressiveness") explains the covariance between them. I measured these three variables on each individual 4 times (40 individuals total).

I then proceeded with the same test, but this time presenting the focal individual with either a small or a big simulated opponent. I again measured how quickly the focal individual reacted to, jumped towards and reached the opponent. So for each variable, I have two values per individual on two different 'treatment': small or big opponent.

My initial idea was to built one MCMCglmm comparing the latency to react to the opponent in the 'normal' test and in the 'small/big' test, then another one comparing the latency to jump towards the opponent in the 'normal' test and in the 'small/big' test and finally one model comparing the speed to approach the opponent in the 'normal' test and in the 'small/big' test.

Models would look like this:
Code:
prior <- list(R = list(V = diag(1), nu = 0.002),

              G = list(G1 = list(V = 1, nu = 0.002), G2 = list(V = 1, nu = 0.002)))


mod1 <- MCMCglmm(latency to react in the small/big test~latency to react in the normal test*treatment, random = ~ ID + repetition,

                 family = "gaussian", prior = prior,

                 nitt = 1000000, burnin = 50000, thin = 500,

                 data = freq, verbose = FALSE, pl = TRUE)
Repetition represents whether the latency to react in the small/big test was first measure in the 'small opponent' treatment and second in the 'big opponent' treatment, or the opposite.

However, since I have four values for the normal test and only two values for the small/big test, this does not work. I would like to avoid calculating a mean for aggressiveness, as it feels to me like doing stats on stats...but I don't know what else to do.

Would anyone have an idea how I should built my model?

My data look like this:

ID_repetition in normal test_repetition in small/big test_treatment_latency to react_latency to jump_speed
1 _ 1 _ NA _ NA _ 28 _ 33 _ 1.30
1 _ 2 _ NA _ NA _ 28 _ 33 _ 1.30
1 _ 3 _ NA _ NA _ 28 _ 33 _ 1.30
1 _ 4 _ NA _ NA _ 28 _ 33 _ 1.30
1 _ NA _ 1 _ small _ 28 _ 33 _ 1.30
1 _ NA _ 2 _ big _ 28 _ 33 _ 1.30
2 _ 1 _ NA _ NA _ 28 _ 33 _ 1.30
2 _ 2 _ NA _ NA _ 28 _ 33 _ 1.30
2 _ 3 _ NA _ NA _ 28 _ 33 _ 1.30
2 _ 4 _ NA _ NA _ 28 _ 33 _ 1.30
2 _ NA _ 2 _ small _ 28 _ 33 _ 1.30
2 _ NA _ 1 _ big _ 28 _ 33 _ 1.30

Thank you very much in advance! :)
 

Karabiner

TS Contributor
#2
I would like to avoid calculating a mean for aggressiveness, as it feels to me like doing stats on stats...but I don't know what else to do.
If you present them with the average stimulus 4 times, then the 3rd and 4rth presentation
might introduce a variation not present in the presentations #1 and #2 (fatigue, habituation,
or whatever). So couldn't you just leave out the presentations #3 and #4, and work with
a 3*2 repeated measures design, either with 1 dependent variable, or with all three at a time
(repeated-measures MANOVA)?

With kind regards

Karabiner
 
#3
Thank you for your answer. I forgot to say that in the previous study I mentioned, I already checked that there was no habituation effect across the repeated measures. But of course, it might still be that some variation is introduced, just not to a level that it would show a significant difference between trials.

I will try to run my model with only the first two trials, however if anybody has an information on how to run the model with including the four trials in the normal test, I would be really interested in learning how to do that. It might always be useful at some point :)

Kind regards,
Mélissa