Table 2.

EVL and PVL model equations for estimating parameters. Model-fitting selects parameter values that maximize the likelihood of the decision maker’s responses, given the model.

Concept	Model	Model Equation	Free Parameter(s)
Valuing a card	EVL	u(t) = w · win(t) − (1 − w) · loss(t)	w = Attention to Gains
	PVL	$u(t)=\{\begin{array}{rr}\hfill x{(t)}^a& \hfill \text{if}x(t)\ge 0\\ \hfill -\lambda \mid x(t){\mid }^a& \hfill \text{if}x(t)<0\end{array}$	λ = Loss Aversion α = Utility Shape
Creating a deck expectancy, E, for decky j on trial t	EVL & PVL	Ej = Ej(t − l) + A · δj(t) · [u(t) − Ej(t−1)]	A = Recency
Probability of choosing Deck j	EVL & PVL	$Pr[D(t+1)=j]=\frac{e^{\theta (t)·E_j(t)}}{{\displaystyle \sum _{j=1}^4}{e}^{\theta (t)·E_j(t)}}$
Consistency between choices and expectancies	EVL	$\theta (t)={\left(\frac{t}{10}\right)}^c$	c = Consistency
	PVL	θ(t) = 3c − 1

Note. j refers to deck A, B, C, or D. The variable δ_jt is a dummy variable equal to 1 if deck j was chosen on trial t, otherwise 0.