126 / 218
124
Chapter 6
Table 6.2
Experimental design: test
Task cue
Compound stimuli
Correct lever
Reward context
Low [RC-A]
High [RC-B]
8 x AUD
4 x A1-V2
Lever 1
1 pellet
3 pellets
4 x A2-V1
Lever 2
1 pellet
3 pellets
8 x VIS
4 x A1-V2
Lever 2
1 pellet
3 pellets
4 x A2-V1
Lever 1
1 pellet
3 pellets
Task cues (noise and house light): AUD (auditory) and VIS (visual) signal the relevant modality; compound
stimuli are composed of one of two auditory stimuli (A1 or A2; clicker and tone) and one of two visual
stimuli (V2 or V1; flashing or steady panel lights), creating a response incongruent compound; In case of
AUD
A1-V2 a left lever press is rewarded, whereas in case of VIS
A1-A2 a right lever press is
rewarded with 1 or 3 pellets depending on the reward context (RC).
response
TRIAL 4
task switch
le lever right lever
TRIAL 3
task repeat
le lever right lever
TRIAL 2
task switch
le lever right lever
TRIAL 1
le lever right lever
task cue
target
low reward block
high reward block
feedback
Figure 6.1
Cued task-switching paradigm with reward manipulation
On test, animals were presented with response-incongruent combinations of the auditory and visual
stimuli (e.g. tone and house light), inducing a response conflict. Resolving this conflict correctly required
the use of the task cue. For example, when a tone and flashing panel lights (associated with a left and
right lever press, respectively) were preceded by a task cue signalling the auditory task (e.g. white noise),
a left lever press would lead to a reward. However, when the same combination of stimuli was preceded
by the presentation of the house light, a discrimination based on the visual stimulus was required
(i.e. a right lever press was correct). Crucially, this paradigm allowed the trial-by-trial manipulation
of a repetition or switch of the task set, allowing a direct comparison of performance on repeat (i.e.
auditory -> auditory and visual -> visual) and switch (auditory -> visual and visual -> auditory) trials.
In addition, animals were trained and tested in two distinct reward contexts, allowing the assessment of
task-switching under different reward circumstances.