. Author manuscript; available in PMC: 2018 Nov 1.

Published in final edited form as: Behav Processes. 2017 Aug 19;144:20–32. doi: 10.1016/j.beproc.2017.08.004

Table 1.

Summary of models.

Model	Acquisition rule	Absent Beta	Absent Alpha
RWM	ΔV_Stim1-Stim2 = α_Stim1 * β_Stim2 * (λ - ΣV_i-_Stim2)	β_Stim2 = S_Stim2 * k1	Null
WCM	ΔV_Stim1-Stim2 = α_Stim1 * β_Stim2 * (λ - ΣV_i-_Stim2)	β_Stim2 = S_Stim2 * k1	α_Stim1= k2 * S_Stim1 * ΣV_i_-Stim1
CEM	ΔV_Stim1-Stim2 = β_Stim1 * (λ_i - ΣV_i_-Stim1) * β_Stim2* (λ_j - ΣV_i-_Stim2)	β = S * k	NA

Note: 0 < k1 < 1 and −1 < k2 < 0. By convention, when a stimulus was presented (omitted) λ = 1 (0). S = salience (0 < S < 1). RWM = Rescorla-Wagner (1972) model. WCM = Within Compound Model (as described in Witnauer & Miller, 2011). CEM = Conjoint Error Model. NA = Not applicable. All models assumed that Vⁿ = ΔVⁿ + Vⁿ⁺¹. In simulations of CEM, cues and outcomes were represented by separate values of k. Thus, k1 modeled the reduced associability of absent outcomes and k2 represented the reduced associability of absent cues.