Trial sequence

[emirvine]

With the 4-arm layout we need to decide how many trials of each type we want and how to organize those trials.

One proposal is for each session to have the following numbers of trials.
Low: 5
Medium: 25
Medium-control: 25
High: 50

If we are limited to no two trials in a row then the majority of trials will be alternating with every other trial along the "High" trajectory. This seems like it may pose a concern because it reduces the complexity of the task. But maybe that's not a huge issue.

Alternatively, we could allow for two of the same trials in a row but make it so the rat has to go back to the central port before each trial. So, in the case of two in a row he would go down the arm to the feeder then come back to the center to turn around and go back down that arm. This would allow for a more diverse sequence of trials.

[mvdm]

The medium and medium-control should have the same number of trials I think, so experience is equal but we are manipulating preference. That way we can test if any effect that seems like it's due to experience is actually due to preference.

Something that Travis found helpful is to think of the trials in two pairs (North-South and East-West). The rat alternates trials taken from these pairs. That way the rat never needs to turn around, and it's easier to keep track of for the experimenter.

For instance:
Low (North): 5
High (South): 35
Medium-predictable (East): 20
Medium-random (West): 20

[mvdm]

(I reduced the total number to 80 but could be higher: High 45, Medium 25 for instance)