Build a trial-level design for a learning task

One row per subject and trial, with the payoff schedule of EVERY option fixed in advance and a placeholder response. Fixing the whole schedule ahead of the choices is what makes a draw from these models coherent: a simulated subject that takes the arm the real one did not still needs to be paid.

Usage

frm_task_design(
  task = c("bandit2arm", "reversal", "bandit4arm_restless", "igt", "twostep"),
  n_subject = 30L,
  n_trial = 100L,
  p = c(0.7, 0.3),
  seed = NULL,
  ...
)

Arguments

task

One of "bandit2arm" (a stationary two-armed bandit), "reversal" (the same with the contingency reversed at the halfway point, and an after_reversal factor to model it with), "bandit4arm_restless" (four arms whose payoffs follow the decaying Gaussian random walk of Daw et al. 2006), "igt" (the Iowa gambling task's four decks) or "twostep" (the two-stage Markov task of Daw et al. 2011).

n_subject, n_trial

Subjects, and trials per subject.

p

Two payoff probabilities, for the two-armed tasks.

seed

Passed to set.seed() when not NULL.

...

Task-specific settings: decay and center and sd_walk and sd_obs for the restless bandit, p_reward for the two-step task's stage-two options.

Value

A data frame, ready for frm_task_simulate().

Details

The column names are canonical, because frm_task_simulate() reads them: id, trial, choice, and pay1 to pay2 or pay4. A fit to your own data needs none of this; the family reads whatever columns your formula names.

See also

frm_task_simulate(), bandit2arm_delta()

Examples

d <- frm_task_design("reversal", n_subject = 4, n_trial = 20, seed = 1)
head(d)
#>   id trial choice after_reversal pay1 pay2
#> 1  1     1      1         before    1    0
#> 2  1     2      1         before    1    1
#> 3  1     3      1         before    1    0
#> 4  1     4      1         before    0    0
#> 5  1     5      1         before    1    1
#> 6  1     6      1         before    0    0
table(d$after_reversal)
#> 
#> before  after 
#>     40     40