Figure 1.
Summary of the four-quadrant map for understanding different domains of human decision making, based on whether a decision is made independently or socially and the transparency of options and payoffs [19].
Table 1.
Parameters and variables of the model.
Figure 2.
Simulations of time steps, with group size
and
agents per group.
From equation 7, we set , and
. The noise component,
has mean 0 and
. Shown are proportions of the 10,000 total agents who have made one of the two possible two choices, one shown as red and one shown as blue, through time. The different plots show simulations with varying starting points for each proportion. The payoffs
and
are chosen from from time-varying normal distributions
, and
and
are both chosen from time-varying normal distributions
. Gray bounds show 95% quantiles for sample paths over group and red/blue curves show the mean paths for proportion over groups.
Figure 3.
Simulations of binary choice model with varying choice intensity and social influence intensity
.
For clarity the plots only show the proportion of agents making one of the two choices (e.g., non-parent). The panels show 16 different combinations of and
, with
for all. Each panel shows results of simulation with 30 time steps, 100 groups and 200 agents per group, noise component
with mean 0 and
, and starting proportion 80% for the choice shown (so the choice not shown starts at 20%). The payoffs
and
are chosen from
, and
and
are both chosen from
. Gray bounds show 95% quantiles for sample paths over group for the proportion of non-parents over groups.
Figure 4.
Map of the chosen values of and
used in Figure 3.
The red dots represent the true parameter values and are linked to their corresponding nonlinear least squares estimates in blue.
Figure 5.
From equation 9, showing how the probability that option (versus option
) is chosen depends on this payoff difference
, two values of
.