Figure 3.

Schematic overview of modeling analysis. Left Panel: Perceptual and response model of modeling analysis. The perceptual model computes the means μ_a, μ_p and μ_s and precisions π_a, π_p, π_s of the prior distributions “associative learning,” “priming,” and “sensory memory.” These are computed into a joint prior distribution with the mean μ_prior and π_prior. Taking the stereodisparity information of the likelihood distribution into account (μ_stereo and π_stereo), θ is derived to predict the responses y. π_a, π_p, π_s and π_stereo constitute the free parameters of the model. All other parameters are fixed to optimal values. Right Panel: Exemplary distributions for the prior distributions, stereodisparity information of the likelihood and resulting posterior distribution.