Procedure
In the MRI scanner, participants performed a two-choice decision-making
task by deciding between two options (see Figure 1). Each option had a
probability p (between 5% and 95%) for winning and a
corresponding probability 1-p for losing. The EV of all options
was kept constant at 0.10 Euro. Each option was either of high or low
risk according to the variance of the monetary outcome per option (as
compared to all options used in the study). A trial was defined as
high-risk trial if both alternatives were high-risk options (both
gambles with large variance in outcome). The range of wins was 15 to 100
Cents and that of losses between -15 to -100 Cents with a difference
between winning and losing outcome of 100 and 200 Cents for high-risk
options. A low-risk trial consisted of two low-risk options, which had a
low variance in outcome. The range of wins here was 11 to 57 Cents and
that for losses between -5 and -60 Cents and the difference between the
winning and losing option ranged between 16 and 75 Cents. Finally, in
mixed-risk trials one option of high risk was paired with one option of
low risk.
The task included three scenario blocks of 99 trials each (see
supplementary Table 4). In the scenario block with pure risk options
every trial had one of the aforementioned risk levels (33 trials each).
A second scenario block consisted of 99 trials with these three
different risk levels transformed into trials with high or low ambiguity
where the probabilities p and 1-p were either unknown
(high ambiguity, 48 trials, see Figure 1 for an example) or given as a
range of the probabilities (e.g. between 30% and 50%; low ambiguity,
51 trials, presented as “30-50%”) in half of the trials,
respectively. In the latter case, the range of outcomes was presented
and indicated the level of risk to participants. And finally, in a third
scenario block, risk and ambiguity trials were mixed (results for this
block will be presented elsewhere). The order of the three scenario
blocks was counterbalanced across participants.
Every trial started with an initial fixation cross that was presented on
a monitor screen for 525 ms plus an individual period for each trial
varying between 10 and 25 ms (timeline shown in Figure 1a). After this
initial phase, the trial options to choose were presented on the screen
for up to 3000 ms and participants were requested to decide for one of
the options. If participants did not choose during this time period they
lost the gamble. Then visual feedback with a colored frame around the
chosen option was presented upon button press until 3000 ms elapsed
since presentation of the options screen. Subsequently a screen informed
whether they won or lost the gamble (for additional 1250 ms). Finally, a
last screen showed the total amount of money they won so far (for 1250
ms).