Learning environment-specific learning rates.

As can be seen in Fig 5B , participants indeed showed a numerical trend to switch faster in the formerly volatile versus formerly stable casino, a difference that turned out to be significant in the third (χ 2 = 4.551, p = .033) and fourth (χ 2 = 4.01, p = .045) trial after the first reward contingency switch.

also in 1,012 other papers

Thirdly, also model fitting results based on the entire test phase showed a trend towards higher negative learning rates in the formerly volatile compared to the formerly stable environment.

also in 53,322 other papers